The modern AI firm series

I wanted to bring in Imbol to take over this session here and start presenting her content. And hopefully, it's an excellent launch platform for the fact that seventy percent of us don't know what this is and where to start. Thanks, Imbol. Fantastic, Ethan. Let's do it. So it's totally okay. You're in the right place. MCP is emerging. Most of us are hearing about it for the first time. Very few are already using it and for a good reason, because it's mostly immature. By the end of the next thirty minutes, my promise to you is that you'll know whether you're on the right track or you're falling behind. And we'll talk about agents and MCP. I want you to know a bit about what other firms are doing with AI agents, and you'll know what MCP is, and you'll know what to do next. You've seen me before, Ethan, you're saying, if you've seen me talk about AI in twenty twenty three, four, probably most of five as well, you've heard me say that if you're in the mainstream with the majority, you're not an innovator that's trying to jump on every new shiny thing, you should be focusing on AI literacy, on selecting secure AI tools for your team, on creating a good AI policy for your team so they're safe to experiment. That was what was right to do in AI in those years. But now it's shifting because if you've heard me then say that AI agents are emerging and don't worry about it, it's the land of innovators, now it's for everyone. We're all starting to play with agents with MCT. So my goal today is not to show you the exciting things that others are doing and automating their entire firm with AI and every employee's AI and they don't have to do anything anymore, because that is the land of innovators. That's not the land of the majority. And that is noise and that is scary. And that just makes us feel like we're behind when we're not. So my focus is on us majority who like using what's already mature, what's already proven, what's not going to try and fail. And we're not here to experiment. We're here to do our work. So we'll focus on those things. I do a lot of public speaking. I run programmes inside Accounting Fence and I run an AI club that now has one hundred accountants in it where we meet regularly and we implement together. So what I'm telling you are the things that are actually working today, not the big promises of vendors. I am not interested in that. So we'll start with talking a bit about agents and you can use the chat and sort of tell me if you've used an AI agent or what agent you used, or maybe you don't know anything about them and you never use them. And when I ask this question, there's usually going to be a few people that tell me exciting stories about the agents they have built in their businesses, often larger businesses that have an IT department that sets up agents for them. And sometimes very small organizations like a one person team that is an explorer and tries everything, has got agents working. And the vast majority in between go, no, not sure, no idea. And I could ask you questions like saying, if I used Outlook, the AI inside Outlook to draft me an email, is that an agent? And if I ask it to do that automatically every morning for me, is that an agent? And if I ask it to summarize my emails every day and give me a brief, is that an agent? What is an agent? And what I want to tell you today is not to worry about it because whether you're building agents or not, the AI tools that you're already using are becoming more and more agentic. And all you need to do, all we need to do is just start using those capabilities. So the first stage of AI, how we use the AI was just chat. It was off to the side and I did my work and I went to AI and I asked a question. I came back to my work and I copied the results or the answer and so on. But that is very twenty twenty four. Then we started focusing on playbooks because when we just tell the AI, do a thing for me, it's quite generic and it's not really what I need. What I want is I want to give it a playbook. I want to tell you this is how we do the thing. And I'll give you specific examples in a moment. This is how we do the thing in our business. These are the steps. These are the guardrails. And then it starts doing things a lot more like I like them. If you go in your Microsoft Copilot, in Copilot chat or the agents on the side, you can create these playbooks. They're basically saved prompts. That's all they are. They're very safe to use because they don't do anything. They're just a saved prompt, maybe with some knowledge that will make sure you get the task done the same way again and again. We then start giving our AI knowledge. So we start seeing our AIs getting access to context about our projects, about our clients, about our world, and then the work that they do is so much more relevant and so much more accurate and they can connect the dots. And the next step, we give our AI tools. So now it's not just living in its chat box and that's it. We're allowing it to touch our spreadsheets and to go into our emails to read and also to create, to summarise and also to draft and to create a file and to update the row and to do things in my systems. Next, we give them triggers. So we say, you know, all of these things that I taught you how to do the playbook exactly how I like it I give you access to my email so you can do it and I give you access to my data so you can work. I actually want you to do it every morning or every quarter or whenever a certain thing happens when there's an email coming or a file saved into this folder. I just want things to happen. I don't want to have to tell you all the time. And finally, and this is where we get totally into innovator and early adopter emerging world is those AI employees. The difference there is that I give my AI a goal rather than a task. Rather than tell it, do this, I tell it, I need you to make sure that this is always done and it's done properly. You figure out how and it will improve itself. So this is very theoretical. Let's look at a specific example. So I'll show you a running example first and then I'll break it down to these steps and what we actually did. So this is my AI environment and what you'll see me do here, it's a prerecorded video from my screen. I'm running one of my tasks that is scheduled to write nightly, but now I'm just running it to show. And I've got this folder here that has some invoices or receipts all in the names they came with in their messy names that came by email or when I took photos with my phone. And I ask it to organize this for me. The first thing it does, it goes and reads all of them, as you can see, and it renames them. Now it renames all of them the way I like them with a date and then the supplier and then the suffix. And then it goes and puts them into Xero. And we can see that it created the spend money transaction and attached a PDF there, the original receipt or invoice. And one of the amazing things that happened when I did that that I didn't even expect to happen is that when it did this receipt that you'll see right now, you'll see that I had some handwritten notes on it. I wrote here Chateau Airport to Hotel in Auckland, and it literally copied that into the description, which I often write on my receipts so that my team knows what it was. And it literally copied that and used that, which I didn't even expect. So what we saw here, I had a messy folder of invoices and receipts that we saved for emails and took snaps with my phone. And my AI renamed them nicely and pushed them into Xero. And what it also does, it creates a spreadsheet and it gives me comments if anything there needs to be looked at, like something didn't make sense, foreign currency with GST, whatever, and then it asks me about that. So let's look at this example on this agency continuum. Is this an agent? Is this not an agent? Doesn't matter. But let's think of where it is and every step of the way is fine. It's just we're giving it more agency. So step one, if I automated my receipts, would be just when I go to my chat and I upload an image or a PDF and I say, get me the details from this one, it would tell me this is the supplier, this is the date, this is the line item, you should put it into that account, here's the GST, and then I would go and copy that into my system. So this is not very agentic. This is what we used to do in twenty twenty four and what many of us still do, but we can move on. If you're here, we need to think about step two. Step two is where I go, here is a playbook, Dear AI, here are my rules for coding. Here is a spreadsheet or here is a document or whatever it is. Here are my rules for how I like the coding to be done. So from now on, whenever I give you a receipt or an invoice, I want you to look at that and keep that in mind so that you do it properly. In Microsoft Copilot, this can be what I said in chat, those agents, just a prompt that has all my rules in it. In Claude, it may be a skill. In Copilot co work, it's a skill as well. So they can be different. It doesn't matter, though, how you do it. But that's the next step of agency. My rules now it will work a bit better for me. If you are here, this is great. This is what most firms are doing right now. They're creating playbooks and calling them agents so that next time we ask it to quote an invoice, it knows exactly how to do it my way. Next, I give it knowledge. So this is where I give it access maybe to my calendar. And so it knows that I was on a trip to a client and it can use that context when it codes my invoice. It knows what I was doing or it knows what projects I'm working on. Next, I give it tools. This is how it could actually rename my invoices because I give it access to my files. This is already percent of the agentic AI that firms do today is at this stage. What happens next is emerging. If you're not doing that, you're fine. So this is giving me tools. So that's how it could rename my files and how it could push them into Xero. Next, I give it triggers, like saying, I want you to do this every night or every hour or every quarter or whenever there's a new file. And next, I give it a goal. Like, I just need you to keep my data clean and my receipts clean. You figure out how to do it. You improve over time. You monitor the issues that come up and improve the playbook, etcetera. So this is the agentic spectrum. Let's look at another example. The most popular example right now is people are creating GST reviewer agents. So level one would be I'm going to normal chat. I upload some last year's or last quarter's GST. I upload the work paper. I upload some relevant documents and I ask the AI review my work before I give it to my superior to review. And that's it. That's level one, just chat. Level two is I would have a playbook. I would tell it, here is the checklist. Here is what my reviewer, human reviewers like reviewing. You review it for me. Here's the checklist. So that's the next level. And you do that, as I mentioned before, with those agents in Copilot or skills or whatever that is. Next, I could give it knowledge so I could give it some context about my clients and what businesses, industries they're in. So it knows that for this client, Bunnings is business expenses because they're a trade person. But for this client, it's personal and whatever it may be. So I want to give it more context. Next, I'll give it tools like the ability to fetch the reports or to prepare work papers. Next, I'll give it triggers. I want you to do this every quarter. And next, I'll just give it a go. I'm not telling you exactly how to do it. You're a senior employee. Your job is to keep our data clean. So that's sort of, again, most firms are here right now, but this is what we're seeing emerging and where we're seeing it going. I want to show you another example now that will bring us into MCP world. So let's look at the example first and then I'll explain. So in this one, I'm asking my AI, what's my next meeting today? And it says, well, your next meeting is with Who is with? It's looking through my calendar. It's thinking about it and it finds my next meeting, looking through my calendar and it's with seven group, whatever, whatever at three pm. And there's an open invoice there, an outstanding invoice. So I copy that and I ask it, tell me more about find me information about this invoice. What is it about? And now it goes to my accounting system and it looks for information there. It does a bit of a zero find invoice and it comes back in a minute. Let's jump it forward a bit. It comes back with information about this invoice. So what actually happened here is that I used one AI and it looked through my calendar and then it looks through my accounting system. That's what I want you to understand from this example. And that is what MCP servers are all about. So this is MCP magic. MCP is simply a layer over an app like Xero, Ignition, Outlook that allows AIs to talk to it and to ask it questions or to ask it to do things. That's all it is. So ten years ago, we all learned about APIs that allowed systems to talk to each other. MCP, called model context protocol, all it is, is a layer over our apps that we're using that allows our AIs to use it. That's it. So if Ignition has an MCP server, that means that I can go to my Copilot or my Claude or my CHA GPT and ask it to talk to Ignition and ask questions about my data or ask it to do things, create something in there. If Xero has an MCP server, that means that I can have my Microsoft Copilot, my Cloud, my ChatGPT Connect and ask it to do things for me, create an invoice or ask the questions like, give me details on this invoice. That's what we need to know about MCP. That's all. The other thing we need to know about MCP is that it's a very emerging field. It's very new. So many don't have it yet. Some apps have a basic MCP server, but it's not connected to different AIs just yet, or it doesn't allow a lot of actions, just a few actions. So, for example, when we go into this is Copilot and in Copilot, I can they call them tools and in tools I can go connectors. And here if I click select all here, which I can't because it's a screenshot, I see a whole list of apps. Really, these are apps that have an MCP server that's connected to Copilot. And I could connect them and then I can go, right, Copilot, do a thing in an app. If I go through Claude, it also has connectors, all of these, and I can scroll down. You see how tiny the scroll bar is? Because I can scroll down a million years to the bottom and there'll be all these apps that I can connect through MCP, but I don't care. It's just Connector. And I can ask my Claude to do things in those apps or to get information from those apps. Claude and Xero announced, or Xero announced their MCP server connected to Claude a few weeks ago. It was during a club session in my AI club with accountants. And I said, no, there isn't yet one. And someone said, oh, it was published today. We all looked it up and we all figured it out together. What we learned about it is, one, that it connects with Claude, not with Copilot just yet. And two, that it's very basic. So it can do get cash position, get contacts, get financial position. It doesn't have any right. It can't create invoices, etcetera. And it doesn't cover everything in Xero. So it's a pretty basic one. If you go to Copilot and go to the Connectors area and search for Xero, you'll find this sure Xero Lite and this Xero Accounting Magnetism, these are not the official Xero one. These are third parties. So I wouldn't connect them before I went and checked very well who they are, what they do with my data, etcetera, because they are third parties. There isn't one yet. There is an announcement in the Xero help world that says that the Xero Agent for Microsoft three sixty five Copilot is coming. So it's coming. I don't know when. And I imagine it will be as basic as this one at this point. All right. So we know that MCP magic happens here where we use MCP to give our AI knowledge and to give it tools so it can query Xero, Ignition, Outlook, whatever it is, for information about my client, about my project, about what it needs to know. And it can do things, draft an email, draft a proposal, draft an invoice. But in terms of maturity of agent agentic AI, the more we go to the right here, the less mature it is. And as majority, not early adopters, not innovators, We want to be where it's quite mature because the other side of it is risk and the risk goes up the newer the things are. So we want to be here in my AI Club. I hope you can see my cursor when I move it. I'm not entirely sure, but I'm highlighting the playbook area in the AI Club. When we talk about these hundred accounts and accountants, we talk about what agents they're building. Ninety percent of the work is happening in this playbook area, creating simple, safe prompts that are called agents in Copilot and skills in Cload or whatever it is that you're using. They just make the work repeatable. And the early majority is now starting to give it context and tools through MCP or in other ways. When we think about what agents to create, the easiest way that I know and how I work with firms on that is to look at shapes or patterns or archetypes or whatever you want to call them, like a reviewer agents like that GST reviewer that I talked about. It's very popular or an extractor that can take messy data and find key information in it, like when we're onboarding clients or things like that or get information from meeting summaries, drafter aggregator, etc. I'll give you at the end my entire agent map that has all of these with thirty two examples of actual agents that people in club are building and sharing. So that's what accountants are actually building. But if we go just into the reviewer one. When we think about a reviewer agent, what we want there is something that the preparer can use in order to review their own work before they hand it to the human reviewer who's not allowed to use the same reviewer. Of course, they need to do a human review at the end. So the idea is that we're saving time for the human reviewer because the preparer will do all these iterations of using the AI to review their own work a few times and correcting things before giving it to the human reviewer and increasing the quality of work of the preparer. So this is what we're trying to achieve. So people are building GST reviewers. Here is my work papers review, BAS reviewers. Income tax return reviewers, very, very popular. FBT calculation, payroll review, AML risk assessment, very big right now, creating those agents that will do the risk assessment for our clients based on our checklist and taking into account all instructions that we have to follow to make it right. Bank and inter entity reconciliation, financial statements, work paper reviews and more and more. But these are the popular ones that I see people in the club build the most. What we learned from doing that. So when we were all doing a GST review or GST return for the New Zealanders, bus return for the Australians and we did a GST review prompt party. So people share their prompt and people run the prompts of others on their same data to see what we find. And we found some really interesting things. That's also in the thing that I'll give you at the end that has the map. We learned that a short prompt and a long prompt, a longer prompt is not necessarily better because the more detail you give the AI on how to do it, the more you constrain its thinking and you want to allow it to think because it's smart. We also learned that we get better results from breaking the agents to a doer and a reviewer rather than trying to do everything in one. So the more we break the agents to bits that work one after the other, the better result we end up getting. And I'll give you all the examples for the others. So next steps. If you are late majority, if you're one of those who never touched an agent, have no idea, create your first reviewer agent. The risk is very, very low. The risk with a reviewer agent is really only the risk of overreliance, and you need to manage that with your team. But it can't do anything. It can't break anything. The only risk is that people will over rely. It's a good prompt that you iterate on until it becomes really good and you save it as an agent in Copilot or if you're using Clot or ChatGPT, save it in different ways. But that's it. Very safe, but the value is very, very high. People save massive time and find things that could have slipped through that don't. So we always say nothing goes out before a human reviewed it. We also start saying nothing goes out before an AI review it, reviewed it because they do really good review. If you're already doing that in your early majority, you want to start thinking about spreading that. So you've done a GST review, great. So let's look at the process to identify more and more agents like that that you can create to automate review in more parts of your business or extraction or all the other shapes that I showed you. So you can use my map that I'll give you to do that. And if you're in the early adopters world, then we start looking at MCP connectors to really automate these things. This is early adopters still because you'll try things and you'll find that they're limited, like the zero MCP connector, you'll find that they don't connect to Copilot and they do connect to Cloud. So this is where we waste a bit of time in early adopter and innovator land. In my world, I prefer to watch them, see their experiments, learn from them, and when things become stable, that's when we want to jump in. So we're bang on time. If you scan this QR code or go to inbal dot com. Aueignition2026, you'll get my presentation slides and you'll get my agent map with the thirty two agents and you'll get your foundation diagnostic and who to follow on LinkedIn and all of that coming directly from my inbox. So if you reply, it will come to me if you have any questions. Hand over back to Itha now on time and I'll be looking at the chat and answering some questions if there are Thanks, Emboel. That's fantastic. Just one question before you do go, and that was an excellent presentation. Presenters has asked if you could talk just a little bit about security and privacy, particularly with the MCP, and we'll cover this shortly with Ben's section as well. But what would be your sort of high level comments around just the security of using an MCP with the likes of Copilot or with Claude or so on with some of these tools that you Super important because we want to make sure this is an industry based on trust and we want to make sure we keep everything safe. So MCP servers, just like using Copilot in general, uses or applies the same permissions of the user that's activating it. So if I connect the MCP server to my Copilot and I use my credentials, it will use my permissions when it goes into Ignition. Or if I use Copilot generally and ask it questions about emails and about files in SharePoint and in OneDrive, it will use my credentials and my Copilot will have my permissions. So maybe I'll get a response based on these five files And you, Ethan, would ask your Copilot the same question and we work in the same organization, but your Copilot will find five files because you've got access to different folders that I don't. Or maybe you copied on emails, other emails than me. So when we use tools like Copilot and like MCP servers, pretty good. We're pretty safe with permissions. And except then when I mentioned those third party zeros, if I don't know who created this MCP server, I'm not going to use it. If it's created by Ignition, I trust it. If it's created by a third party, I need to do some research and make sure I know what I'm doing. Some of what I showed you uses API directly rather than MCP, like creating things in Xero. That's where I'm a bit in innovator land and I need to make sure that I manage security properly. The rule of thumb that I want to leave people with to make people feel safe is if you're using Copilot, Microsoft Copilot, and if you're using vendor MCPs that they publish directly and you don't share logins, you use your login, everybody uses their own login, we're pretty safe. It means that we get things a bit later because Copilot is the conservative, Microsoft is the conservative ones, we get everything a bit later, but we get it very secure. You use things a lot, you'll have to manage a lot of it yourself. Absolutely. Thanks, Embal. That's very helpful. And yeah, really appreciate you sharing that as So there's a bit of a guide here as well. I don't know if there'll be an option to sort of request access to this, an MCP guide from Imbil here that's available as a resource. Otherwise, it will be sent as a follow-up. Think Alana's popped that in the chat as well, that all of the links will be sent to you, as well if you didn't manage to to catch that QR code on the previous screen there. So thanks again, Imbil. That was that was excellent. And the GST review party sounds like a party that I I wanna be at. Love that. Geeky party.

Ben will talk a little bit about some more practical use cases and go into a little bit more detail around what he's built for his firm as well. Ben, how are doing today, mate? Thanks for joining us. Yeah, really good, Ethan. Thanks for having me on. To share my use case and my experimentation with the audience. Excellent. Well, I'll pass it over to you and feel free to take it from here. Awesome. Thanks, mate. Alrighty. Well, hi everyone. Good morning, I think for all of us still, for all the states still mourning. Alrighty. So what I'm going to be sharing is a handful of use cases that I felt really helpful over the past few months since I've had access to the Ignition MCP. So just a reminder as well, I run Inspire Accountants. We're based in Brisbane and we've got just over a dozen team in Brissy and we've got a couple of dozen over in the Philippines as well of our team members. I've been working with AI for the past couple of years now and particularly the last six to twelve months, it's come a long way. And what I love is that we're starting to see some of our software providers open up MCPs, APIs and all these things that connect our AI with the software that we use day to day. So what I might show you is a couple of sort of more basic use cases that I've found helpful with the MCP. And I might finish with a bit of a doozy that again, I found super helpful and it saved me tens of hours, particularly on our leadership level and managers inside Ignition over the past few months with our renewal season. So, alrighty. What I'll also start with as well is security is a question that comes up a lot, particularly with AI. And then if you scroll back a couple of years ago, twenty twenty three, twenty twenty four, you heard of big corporations leaking secrets, leaking the plans to the next stealth bomber. And we don't want to be doing that particularly when we're dealing with clients' privacy. But in my opinion, I feel like the security side of things has come a long way. I'm a big user of Claude. I don't really tend into the Geminis or the Co Pilots or the Chatty PTs, although I've played around with, I'm pretty sure all of them, yeah, come away a bit disappointed in some areas. And I think the general vibe I get is that each of them have their own strengths. But for me, Claude seems to be that all around and really strong with coding, writing, and from my opinion, logic and helping us, it's not just the strategy, but it's even actually doing the work. Alrighty. But yeah, one of the recommendations I give the people I coach is to go with a core team plan. It's an agreement that's at a business level. You turn off some options and you're all inside the security settings Claude. So you're not gonna be training their LLMs or their models on your data. You don't want to be leaking the stealth bomber plans, but it does require a minimum of five users. I've had people balk at that, but at the same time, I think that's one of the small decisions we can make, pay a little bit more and you'll probably have a handful of team that would appreciate a bit of Claude access, bit of Claude love in the early days of your implementation. It's got pretty good security standards. You can also manage what people can have access to pretty granularly, which is exciting as well. And just a general statement is I would continue to do your due diligence, a bit of a tongue twister there until you're comfortable. And I hope that is at some stage, I think that there is a risk that you're so uncomfortable with the whole thing that you don't actually dive in and you miss the boat, so to speak. All right, so use cases. So what I'm gonna do is play around with my sneaky camera setup. Now, the first use case is pretty straightforward in my opinion. The Ignition MCP got released when we were in the middle or just about to start our renewals process at Inspire. So we do proposals, monthly recurring revenue subscriptions on a financial year basis. Repricing a client is one of the processes that we need to look at last year's, apply a price increase percent and then set it up with the right dates for the following financial year. So it was a very, very common thing. Now what I might do is just share my screen a little bit differently. Woah, okay. Now where's my There I am. Cool, cool. All right, so this is my Claude. I've got a bit of a pre prepared prompt here. Now it might actually take a little while. So I'll just hit the go button and then start to explain what it's doing. So it's connecting to ignition through the MCP. I've already set up the MCP to connect. So I don't have to keep doing that each time I want to use it. And you can see as it's thinking, it's showing the ignition logo, it's pulling it all together. And I'll open up my own ignition in the background. It's a bit of a test client. And I've basically said, see this proposal here. I want you to roll it over, apply a seven point five percent increase. Don't send it again. I love what Imbil said earlier about you always want that human in the loop. You don't want to be sending stuff to clients that's not checked. But yeah, we wanna roll over the proposal and apply that price increase. Now, I also added in here because I was blown away the first time I did this. I didn't ask this in the first one, but I said any third party expenses that we pass on to a client as in get reimbursed. And in the use case it picked up, it was actually an SMSF audit reimbursement. Don't just automatically apply a seven and a half percent. I've said that in the prompt up above, but again, first time I did it, I asked it to roll over an SMSF proposal and it was my own SMSF and it picked up the audit reimbursement said, Hey, you might not want a seven and a half percent increase here. I don't know what's happening with the audit reimbursements. So I found it mind blowing that it picked up that sort of level of detail in the proposal. So it's already come back here and it's picked up all the different lines in the proposal. It looks at the twenty twenty six pricing, adds the seven point five percent and it goes, well, there's an audit reimbursement here. I'm going to hold that until I get further instruction. Or essentially highlighting that to flag to me. And it's asked me, do I want to build the proposal? So yes, build it please. All right. So now it's gonna actually prepare the proposal and it should show up in the ignition account. So I'll keep that here for now, working away, but what I do wanna share next, and I'll come back to that one is the next use case. So my next use case was refreshing over three hundred family groups proposals with our new set of terms and conditions. So how this use case came out is, I mentioned we had our renewal season and I used the feature in Ignition where you can actually bulk renew proposals, apply a price uplift all within Ignition. You don't need the MCP for that. That's a feature of Ignition. And that's how I bulk rolled forward about three hundred and something family group proposals. So it was a lot of proposals and the Ignition feature, again, no AI involved in that, did it within a few minutes. It's pretty impressive. Maybe you've tried that out yourself in the software. What I didn't realize is that at the same time I was working with our lawyer to refresh our terms and conditions. And we wanted a bit of a fairly chunky update. We've got the AML requirements coming in, we've got AI and want to build in AI considerations into our terms and conditions, but I'd already rolled over these proposals after the terms and conditions were updated. And one way to do it is to go into each proposal, change the terms to something else, save it and then change it back. It's a bit of an annoying process, particularly when there's three hundred of them. And so my brain went, well, on, do I get our admin team to look after this? The potential that some clients might be missed. And that might take a couple of days depending on their availability, but we're already mid renewal season. So we don't want to fall behind or get anyone on our old terms. And then I think at the time I'd had access to the ignition MCP for a week or two. So I thought, you know what? I might give it a crack in ignition. So I basically used Claude, plugged into ignition and said, Hey, I've got three hundred or so proposals. We've just updated our terms and conditions. Can you please go and change these proposals to the new set, essentially refresh the terms. And my goodness, it took a couple of batches because of the work involved, but it did it. And I was blown away. Ethan, I do want you to thank the development team for including the update the terms in that MCP access. Was extremely helpful and saved us hours and hours of work ourselves. So that's another use case. Again, it's a pretty niche one. You may not have that exact requirement, but I just wanna demonstrate the different things that you're able to do with this ignition MCP. Now, we go back to the first use case and I'll get, there we go. Let me get my head in there. All right. So validation pass clean. It's done. The twenty twenty seven is drafted. So if I go to our ignition account, here's the old one. I'll just hit refresh. I'm in the client screen right now. Hit refresh. So what we should see, I hope, is two proposals. One for twenty twenty six, imagining it's active, although it'll be draft because this is my actual file. Here we go. And then one for twenty twenty seven. So it'll pop up over here. Can also see Well, I'll move my big noggin out the way. There we go. You can also see down the bottom in the log that you can see that this is actually created by me through the MCP. So Imbil mentioned it uses your own permissions, but what we love about the Ignition MCP is it'll even log your activity. So who did what via what MCP and all that sort of stuff. So you can actually have a bit of an audit log there. And it's pulled the pricing through. I won't go into detail here, but I guess you can see what I wanted to show you here. We've got a proposal rolled forward. The start date's correct. It's still twelve months. It's pretty cool. So again, yeah, hope that gives you a bit of an insight into repricing as an example. So I've talked about that one. We've talked about refreshing the terms of engagement. And I get that there. Cool. Another use case, I've done this a few different ways. But what I've done is I've created a process of taking a transcript or taking even a conversation with a client and turning it into a proposal. So we also use Vinyl, which is a meeting recording app by Trent McLaren. Some of you might already be users of that. Some of their friends of Ignition as well. And before Trent had built a connection between Vinyl and Ignition, I was actually using Vinyl transcripts. So downloading the transcript in a, I think it's a JSON file, dot JSON, or it's basically a text file, uploading that into Claude with a few instructions and that would then go and build a proposal off the backend. So we do a process at Inspire called the look under the hood. It's a second opinion on your tax. And that's a really easy one to prepare. There's nothing really custom in it. But what I've also done since then is, if you're going back and billing for things like consulting around restructures or mergers and acquisition, due diligence, that sort of thing, where there's multiple different things you've helped a client with. What you wanna be able to do is, and use cases I've done is actually build essentially here's the emails and things that we've done for you. Essentially a little bit of a timesheet. You can actually push that into ignition. It selects how many hours, what's the charge rate for whatever you've helped there with. And then it's a drafted proposal ready to send to the client. So that's another awesome use case that's been quite helpful on a number of occasions. And then the doozy. So what I want to do, and this is more so for the advanced users. My goal is not to scare people here, but it's to inspire you as to what's possible with some of this stuff. So again, I rewind a couple of months ago, we're in the middle of our renewal cycle at Inspire. And in previous financial years, I'd done really manual Excel sheets to track who our current monthly recurring clients are, what their current pricing is, what the new pricing is going to be. I had links to all the proposals in there. And again, it was January, February this year and I thought, hey, how are we going to approach this with AI? Again, we had access to the Ignition MCP and we've essentially built a renewals dashboard that I did manually, but it's all done automatically through pulling on the MCP. So this is a very sanitised version. It is built off the template of what we use at Inspire, but none of the client names are real. And what it's got is the link straight into your ignition proposal. So you can go and see. And again, all the twenty twenty seven links were drafted by the ignition renewal feature. And then again, I'll move my big noggin out of the way. We've got the status of each proposal. So I asked in my build of the dashboard, I want it accepted as green, draft as orange and sent but not yet accepted. Sorry, sent but not yet accepted as orange and then draft as the yellow. Also here you can see the monthly recurring revenue change and it automatically flags if it's under our target percent. But if it's at least or above the target increase percent, it'll show as green. So you can easily see like, maybe this client sold a business, shut down an entity or we're no longer doing best for them or something like that. There should be an explanation for why the pricing goes down significantly. Maybe we lost the client, that's a red, whatever it is. But this spreadsheet was again fantastic to be updated each time by the renewals. Our process that we had to by Claude in our instance, just fantastic. And it has a summary down below. And again, if I scroll back to our process last financial year, it was me creating all these tabs by scratch, me cross checking data that I hadn't missed a handful of clients or who signed up in the year as new clients who left during the year because we don't really wanna track their renewal because there won't be one. It was all a lot of manual. And then each proposal that was accepted in our renewal season was entered essentially by hand last year. So yeah, I'm pretty sure this process has saved, and this is the leadership level stuff, has saved us tens of hours in this latest season. So again, just another use case I wanted to share to inspire. All right. Cool. Thank you, team. My slides can look a little bit funny. All right. So what I wanted to also share is, well, what then? Because Ignition MCP is cool. How are we at Inspire implementing AI to be able to help us? I feel there's a big shift into getting AI to be able to do the work for you. And so the functions that we've got AI assisting with now is things like an initial query bot or agent that goes and prepares initial queries. It looks at last year, what we asked. It goes and pre populates. I'm gonna do a bit of a shout out here for Content Snare. Content Snare has got an amazing API. It goes and pre populates all their questions into a nice questionnaire so that our team are there to just review, tweak a few things and then send those initial queries out. So that's just one example where we're finalising a work paper prep agent. We've also got a working review agent that we love. We had that working for the twenty twenty five year end, but in the twenty twenty six financial year, which has been fairly helpful. And as Inbal said earlier, think she mentioned that there's like a pre review where the team member who prepares goes and gets a pass through check. That's a fantastic idea that we've also built the ability of so that we're not submitting stuff that's got easy quick things to fix to our high level managers or partners for review. So yeah, I'll call that the year end engine. And my guess, and this is conservative, there's a higher estimate than this, but for our first full year, full financial year running on this one, my guess it's gonna save us about six thousand hours of like job hours a year across three thirty family groups. So yeah, looking forward to essentially what that adds to the bottom line. But also, I sort of feel like that the benefits of AI are not just dollars. There's also a headspace that you get back. There's less cognitive load, particularly on our senior thinkers. So, yeah, I think you also get fulfillment and enjoyment back, not just dollars on the bottom of the P and L. Alrighty. So how I wanted to wrap up before we might jump into a bit of Q and A is three ways to engage with me further from here. So I've put together a three video course on ChatchipPT to Claude for accountants. So this is helpful for you if you've just been using ChatchipPT or maybe even Copilot, you haven't really got much results from it. And you're looking to explore why I've chosen Claude as my go to model or AI assistant to be able to help me with all my development. That middle one there, the AML starter. So I've developed this particularly for Ignition compliance users. So compliance is the AML feature that Ignition released in June last month. And if you haven't got your AML sorted yet or you've got some product but you're not so happy with it and you're Ignition user, this is perfect. It guides you through step by step. I think there's about twelve or fourteen lessons in that one that guide you through step by step how we created our AML programme to be very incorporated with Ignition's compliance features. It is free if you use the code compliance. And Ethan did say he'd shout me a steak for everyone who signs up there. Thought I'd throw that one in there, Ethan. But yeah, perfect if you want some assistance around that AML side of things. And then a paid way to interact is the AI powered delivery engine. So I did that workshop in June. It goes into much more detail than I've had the opportunity to share here on how we're adopting AI for our year end process. It is a paid thing and you're welcome to explore that as a way to interact. The links will be in the follow-up email as well. Alrighty. So I think that's it for me. Hope that was helpful. Excellent. That was awesome. Thanks, Ben. That was a lot to take in, but man, some pretty cool use cases there. So thanks again. I guess a quick thing just to add on here in in our last sort of five minutes or so. Just wanted to share another use case that I've seen as part of that renewals function, which I think is really cool. So if I can share my screen, and Ben, keen to get your thoughts on this as well. I'm sure you've played around with it. But say we are creating a new proposal within Ignition, and we're now thinking, obviously, in July about how we get those proposals out to clients. What I was doing, actually, just while you were talking earlier, is getting us to look at all of the current enabled services within Ignition, finding the clients that have compliance services that don't currently have tax planning within the proposal. And, know, basically, I've said list that as an advisory opportunity for me. So it's gone across my entire Ignition account, found those clients where there might be opportunities to add more value, more services to them. It's then gone and created an artifact for me, which is a live dashboard, like an advisory upsell watch list that I can refresh, once a week or once a quarter, and it pulls all of that data in. I then went a step further and actually got it to draft the proposals here as well. So it's added on a six hundred dollars a year optional tax planning service as part of those agreements. And those are actually sitting in my demo account as this year's compliance and advisory amended as well. So it's awesome for those standard renewals, but then also identifying those like, hey, these clients don't have these services or to your point with Vinyl, bring in the transcript and do we talk about anything that's currently not in scope and not agreed to as well? So I think there's massive new fee, new revenue opportunities to be found with this as well that would otherwise have relied on people remembering to flag and service it or your own notebook and you trying to remember to draft a proposal. I think it's quite cool. Oh, sensational. I remember our team did something where we'd have all the clients listed and then what services each do, but it was manual tick boxes. And you're right, that's a fantastic use case. I want to give another little brief, just super quick demo of another AI feature that we've built into Ignition as opposed to through the MCP. And you can access this, obviously, and do your own benchmarking and stuff through the MCP. Because Ignition has access to a large sort of data set of pricing information across a variety of different services, what we released last year was a tool called Price Insights, which actually gives you a really interesting sort of benchmark information that previously you might have gone to industry reports like the Good Bad Ugly report, Ignition. We actually we do a price benchmark report that we release every year, but it is like a PDF report and it's fairly static, whereas this is actually giving you sort of live benchmark pricing information and price positions for various services that you can then feed additional prompts into. So it might be this is a VIP client, they turn over ten million, whatever it might be to try and narrow that down, but it's giving you a price estimate and recommendation based off that data as well. So that's one example of an AI tool that we've built into the platform as opposed to engaging with all of this through the MCP. Thanks, Chrissy. Love to hear that you're using that as well. Very cool. Awesome. Final thing, I just wanted to show what the end use case looks like for this, because I know there's a big chunk of you that haven't seen an Ignition, like haven't that are actually using Ignition in your firm. So I wanted to show you what clients would receive after I for example, we did that prompt earlier, hey. Draft a a new proposal, make tax planning optional, etcetera. What they get is a really slick looking engagement here that sort of combines the fee estimate, the proposal, and the engagement letter together. So you've got videos, messages, you can add brochures and brand it up nicely. You can present, multiple packages to clients as well. So you can do multiple levels of service from Silver Gold, or you can just do a single quote and scope of work. On top of that, you can then have additional optional services such as, hey, tick this little toggle option if you want to add on tax planning, or if you want to add on audit insurance, or if you want to add on quarterly advisory meetings, or if you want to add on, a quarterly stake with Ben, or whatever it might be to the engagement. Those are then all driven through underlying service templates. So these these services include descriptions. Again, you can get AI to actually help you draft these descriptions now as part of a feature we've built into Ignition. It pulls from your standard price list, whether that's on a fixed fee basis or whether that's on a time and cost basis or fee estimate or a range or a minimum fee. However you like to quote, we can help with that. And then the ability to capture clients' payment details then and there as well. So think we've done webinars on this before, Ben, but the ability of tying that all in together, where it's the sales process, outlining the scope of work, outlining the fee schedule, asking for the payment details, and then signing off on the engagement letter, and now obviously doing your AML compliance as part of this really does make it fairly all in one. You know, I think maybe we could squeeze a few more things in there, but it's it's pretty all in one, as far as I'm concerned. And then you've you've actually got all of that structured data there to do your renewals the next year. It makes it a hell of a lot easier, as you heard Ben say. Awesome. I just wanted to go through that. So hopefully that's helpful for those who haven't actually seen Ignition. A little question I had for you, Ben, which Luke flagged was around security, and he's sort of specifically talking about TFN info. How are you sort of approaching privacy and security with this and Ignition? For one, we don't expose that through MCP, so I think we're safe there, but it's a valid question that I thought might be worth throwing to you, you've got thoughts. Yeah, nothing we send to or from Ignition has TFNs or any I don't think it really has any identifiable data. Like the TFN side of things. So no, I'm not sure we need to be concerned about it. Yeah. What it does have access to obviously is destructured service data, the payment amounts and all that sort of stuff, but not the the personal client info as well. So I think we're cool on that one. Emme has asked, will will Ignition MCP let you interact and set to set up the job workflows as part of the rollover into the next tax year? The way that I mean, do you wanna answer that one, Ben, or I can? Over to you. Maybe you. I think you can talk to a little bit more a variety of integration. So I think the the key thing here is that Ignition has native integrations with workflow tools. So whether you're using Carbon or Xero Practice Manager or something else, you can prompt the creation of jobs or even FYI now as well. And I'll I'll show you briefly what that looks like. But in the automation step, when you are building a proposal here, you can actually have each of those workflows linked to the services in your proposal. And so this is FYI's example. I've got my individual services in the proposal here. Those will actually be linked into FYI jobs when I've connected those two together. And so following the Ignition proposal acceptance, it kicks off workflows automatically off the back of that. I think Carbon's just launched their MCP integration, so maybe there's something you could do there to kind of bridge the gap. But I think you're better off at this stage probably using the native integrations there where available. I presume that's your approach, Ben. It is, yeah. And even just throwing a few other ideas out there. So Ignitions had a Zapier connection for years now. I don't know when it came in, but every proposal that gets accepted, we've got a trigger Then we run a heap of stuff off the back end of it. One of them saving a copy of the PDF engagement letter, the signed one down to a SharePoint folder. So we've got two copies of it, one in Ignition, one on our own server. Then we also post into, we've got a sales channel in Slack that every new proposal, there's a bit of a celebration that goes on there. So we can keep the team motivated from that perspective too. So yeah, there's not just, I think that the Zapier opens up the door to a heap of other stuff you can do off the backend of it. Yeah, absolutely. Rob's asked a good question as well. Can agents be built for, say, a reviewer agent? And absolutely, you could build something that's automated using the MCP to review draft proposals in Ignition, if that's question, Rob. But you would be able to have that set up as a recurring thing and and, you know, review pricing and review scope if a proposal moves into that stage. Absolutely. Excellent.

Hi, everyone. Welcome. Just kicking off a minute early as I usually like to do if you've been to one of my webinars before. Before we get started, just say hi in the chat. Calling in here from very rainy and windy Lake Macquarie, New South Wales. But instead of asking where you're all calling in from, I'd like to get an idea just to break the ice. What's your favourite drink at the moment? Could be tea, coffee, could be something a bit stronger. Maybe this kind of weather is giving you, you know, hot chocolate vibes or something. I'm just a I'm a fan of the coffee. I haven't actually got my coffee today yet, so I apologize if I'm a little bit weird. I haven't had my coffee today. But, yeah, drop it in the chat. Let me know what your what your poison is. And, also, maybe if you wanna let me know where you're calling in from as well. That's always nice. But we'll get started in a minute or so once everyone's come in. Small skin cap, like that, Beck. I'm a decaf guy, and I'd probably telling him myself there. Stopped drinking caffeine a long time ago, and I don't regret it. But I do sometimes have a tactical coffee, I call it. A tactical coffee if I'm going, like, for a long drive or something. Morning. Morning, all. Yeah, so let's get started. If you want to keep chatting in the chat box, you can do that. Tell us where you're calling in from. Tell me what your favorite drink is. Irish coffee for Ruby. You can tell it's in the financial year when that's your suggestion. I like that. Maybe something a little bit stronger. Wonderful. Well, welcome everyone. Today, we've got a wonderful webinar on a very popular topic. It's all around pricing with confidence, and this is your AI pricing workshop. So, we're talking about using some of our latest AI tools and giving you some practical advice around using these AI tools to help price your services and raise your fees during this time of year, which is very timely. So let's push on through, and I've got a fair bit to get through. All of it very helpful and practical, hopefully. We're recording, if you can pick that up already. So, if you do need to duck out, we'll get the recording. The chat, you've already found some of the chat telling us what your favourite drinks are. Double shot espresso, that's the way to go. I love it. And keep chatting in there. That's good for banter and anything you want to talk with other attendees about. If you have any questions, then drop them to the Q and A, which is a separate section. And if you have time at the end, we'll get through to those. Otherwise, I'll come back to your questions after the webinar and make sure they get answered. So, keep it active. Great to see you all and let's kick in. For those of you who are not familiar with Ignition, where have you been? But seriously, we are a platform that's been around for many, many, many, many years now. We help you go from lead to cash. We transform your sales, billing and payments process, and we're trusted by over eight thousand five hundred service based businesses all around the world. There's a poll up on the screen now. I'd love to know where you're at in your journey with Ignition. If you're just sort of, you know, feeling us out, wonderful. If you know a bit more about us, let us know. But yeah, that'll help us understand how we can tailor this content and help you on your own journey. I've been with Ignition for many years now, and it's great to see that number go from, I think I remember when we had two hundred customers or less up to eight thousand five hundred, so I'm loving the journey. Just introducing our hosts. My name is Brendan Allen. I'm a manager here, senior manager actually looking after the key accounts team. And I'm joined by Beck Mahalak, who's head of accounting at Ignition and director at Business Depot. Hi, Beck. Great to have you on with us. Hi, everybody. Good morning. How are you? We'll learn a bit more about Bec. I'll save the introductions for the actual good stuff. We're going to have a little bit of a conversation shortly around how Beck's been using AI in her business and some advice on how to review your pricing. So that'll be really good. So thank you, Beck, for coming online today, and I appreciate your time. Let's start with a little bit of a question about your business and your challenges. So what's your biggest challenge right now with pricing? What are you struggling with? Probably the best time of year to ask this. I know there's been a lot of legislation changes, a lot of things coming in this year that have made us all look at our pricing, cost of living, all kinds of things. So let us know what your challenges are at the moment. I can see that poll going off already. I think we'll share those results at the end too. So, the options are knowing what to charge, pretty broad one. Communicating increases to clients, probably the other end of that, you know, even if you know what to charge, how we tell our clients that we're going to raise our fees? Consistent application of increases. And of course, the age old issue with time and admin involved around any of these items. So, I'll let that go. It looks like communication is a big one. So, maybe Bec, when we have our chat, we can talk a little bit more about communication and how to actually get that out there. But there's also probably equally in the lead is knowing what to charge, which is a big problem as well, which is why we want to help you with some of our AI tools and some of our advice in this webinar. So thank you for answering those polls. That's going to help us lead the conversation in the right direction. Right, so let's go over what we'll be talking about today. The first thing I want to talk about just briefly is Ignition's price benchmarking report, which we do every year. This is our, I think it's our third year doing this now, and it's been a really, really good resource for the industry. We want to talk about why right now is the best time to review your pricing. Again, that poll kind of shows me that that's exactly what's on everyone's mind at the moment. Of course, we'll talk about how to leverage artificial intelligence or AI for pricing guidance, using your real proposal data from within the industry. I want to talk about MCP, and I won't give away too much here, and I won't talk about this too much, because it's a fairly new thing, but it's something that's actually been really beneficial. I was actually looking at the usage stats of our MCP this morning, and it's huge across all of our customers. So we'll talk about MCP and how we can use some of these AI tools to help do your bulk pricing tasks. And of course, across the whole thing, talking about practical tips on how you can implement pricing within your firm this month, and hopefully tackle some of those worries around knowing what to bill and how to communicate with your clients. Okay, so first off, we're going to send this to you, if you haven't got this already, we'll send this out as part of the follow-up on the webinar, and it is the twenty twenty six pricing benchmark report. This is gold, this is, let me go to the next slide, I'll show you exactly what we did. So we surveyed one hundred and sixty Australian accounting firms using Ignition, here's how we broke down the different sizes of practises, so we have a revenue breakdown there. And we also made sure we got samples from all over the country, including major cities and regional. Shout out to the regions, thirty one point three percent. So we made sure we, this report actually has a bunch of great information around some of the main services that you would probably bill for as part of your practice, and what your peers are looking to increase their pricing by based on this survey. So, it's been a really good survey. So, make sure you keep an eye on that once that comes into your inbox. I wanted to pull out some highlights from that report, though. And these are actually really interesting and quite relevant to what we want to talk about today. Some of these stats are maybe surprising to you, but as of the last report that we did, seventy eight percent of firms plan to raise their fees this year. That doesn't surprise me, again, based on the amount of changes that occur, legislation changes, costs of business have all gone up. So, that's pretty accurate, I feel. Only sixty four percent, though, are somewhat confident in their pricing strategy. So, there's definitely some assistance needed there. Seventy two percent of firms plan to increase their prices by five to ten percent. So if you're looking to get a ballpark in terms of how much you should increase, that's where most firms are landing up in. Unsurprisingly, again, sixty four percent of firms are citing rising business costs as why they need to increase their prices. Only six percent of firms have no plans to raise fees. If that's you, I'd love to hear from you. Maybe drop something in the chat, tell me, you know, why that is the case. Maybe you've done it already. Maybe you've got some good insights to share around that. But I think this one's quite relevant at the moment. This time of year, we're now running into tax return season. The most common price bracket for basic individual tax returns is now over three hundred dollars whereas I remember doing this report even last year, the year before, that was probably more in the two to two fifty mark. So that gives you an indication of how much fees are going up and how much your peers are raising their prices. So something to consider as you go into this webinar and beyond and start thinking about your own fees. Right, Bec, I'd love to chat to you if that's okay. Yes. Thank you so much for joining us today. I know you've got so much experience in this. You've been running a very successful practise for many years now. I've known you for probably those ten years, if not more. So thank you for joining us, appreciate it. No worries. Thanks for having me. It's always exciting to have these conversations, particularly with you, Brendan, because you know all about me and the business. We can have some great chats, but it's really prime time for all accountants out there at the moment to still be considering at every client interaction, what are we charging and making sure that that's actually covering a cost and also not just covering a cost, but really defining your value when you are selling your services to your clients. Absolutely. No, I I learn something myself every time I chat to you, so let's get stuck in. Three areas of focus I want to talk about with you, Bec. The first one is why firms need to review pricing right now, which you just kind of touched on. Of course, using AI and price insights within Ignition and potentially other tools to make better decisions around your pricing. And of course, turning better information into action, which we can also talk about quite practically. So, let's get stuck into the first one and talk about why you'd want to do this right now. So, first question I've got for you, Bec, and I think this is kind of the sentiment from the industry right now, is what's making pricing particularly difficult for accounting firms right now at the moment? Yeah, so it's this time where it's not just one thing, it's a little bit of everything. What we do is getting harder. We have more compliance that we have to deal with. Our costs are getting a lot more higher. So we're talking about higher wages, higher costs of technology, insurance, administration, rents going up. Everything in our world is getting more expensive and it becomes more and more important for us to explain our value and just talk to our clients quite openly and honestly as well around the fact we're also a business. Are, yes, we provide a service but we're running businesses here in the background as well and as we counsel them on increasing their fees, we've got to take our own advice. And I really struggle with this whole concept as well that when accountants sometimes look at this and if we're a little bit too time poor, we just go: Let's just apply a CPI. That is a real problem because if your cost base was already miscalculated, if your price was already too low, that doesn't do anything. That really just sort of makes sure you're not losing quite as much. So a good review is probably what's needed. But again, we're always just so time poor. Yeah, no, that's huge. And that piece you said around there, communication, that was one of the strongest sentiments in that earlier poll. So, if you're struggling with communication, do reach out to us, maybe in reply to that email, we send out with the resources and we can help you with some communication methods there. Next question for you, Bek. Okay, so I think most people on this call probably recognise that, you know, yes, CPI is probably not cutting it. Now is the right time to do it. Where should firms start in relation to reviewing your pricing? What I really normally love to say is go through in lots of detail, figure out your cost of delivery, do these big huge calculations and there's stacks of stuff on the Ignition website where I've gone through this and we've got some guides about really diving into pricing which we can share with you. But realistically, if we're talking about right now when we're really busy and we're time poor, we know. We don't need this perfect world where we're looking at timesheet data and really analyzing it. You know as a business owner what services you're delivering too cheap even when we don't want to admit it, you know it in your core. And you also know what is taking a lot longer than it should, whether it's particular clients who's you know what you need to deliver for them has changed dramatically and you haven't kept up to date with your fees. So you don't need to go blank at everything. Look at those key services that you know that selling too cheap and look at those clients where absolutely there's a lot more effort going in than you're getting paid. Start there and try and get some quick wins. Yeah, I like that. I think a mistake, and I make this mistake too with my own work from time to time, is trying to overcomplicate the solution and not just choosing a small cohort of whatever it is that I'm working with. So I think that works really well with your services, just picking And anecdotally, you all know this, right? You know where the scope's blowing out. So I love that. That's a good approach. Another question in regard to fees. Should firms apply a standard annual increase, or should you, like across all your services, or should you review every client individually? Look, again, in the ideal world, you'll review every single service for every single client, but it's not realistic. So if you can look at those key services right now and then have a look at which clients are suffering. I would never ever just blanket say I'm going to increase all my fees five percent unless you've already gone through this exercise in the past where you're very very confident in your pricing model. You know that you've got it covered. Know last year you weren't losing any money on any of your services or your clients, sure, go ahead. Just increase by CPI because you've done it. But the reality is a lot of us have it. When you're looking just really quickly around the timing, this is actually a great time because when I talk about assessing your clients individually, you don't need to sit there and do that one by one in one month. But as we're actually delivering work, it's an amazing opportunity to just remember, yep, clients accounts have come in, tax returns have come in, good time to have a look at their fee because you're already focused on them. Perfect opportunity to increase your fees also while you're engaging communication with them and delivering them value. It's much easier to talk about fees than some other ad hoc time. Yeah, that alignment of value is huge. It's something I've been talking about for many years now. Just a nice comment from Justine in the chat. This is just a suggestion around timing as well. Justine reviews their pricing in November each year for work from January onwards. So doing that during that quieter time. I know it's shooting yourself on the foot by saying quiet time because it's never quiet, but yeah, I think that makes sense. That's cool. Thank you for sharing, Justine. Let's move on to the next topic, which is sort of the main component of this webinar, is AI pricing. You know, using these tools and using this data that we have at our fingertips now to make better decisions. So another question, Bec, around this topic. I know that firms have traditionally relied on experience and judgment when setting fees, and I don't think that's, you know, going away anytime soon. But what does Ignition's price insights feature, which I'll be showing you shortly, add to that process, you know, that I know that a lot of business owners already have in their mind? Price insights is fantastic, and this hand in hand with the benchmark report actually opens up that area of sort of unknown with accountants. So we can dive in and we can figure out what is our cost of delivery and we can figure out how much profit we want to make on a different service or with a different client. What we never really knew is what's the market like. There are a couple of ad hoc reports out here and there and but they're not they don't dive in enough. They don't actually sort of quantify the value of a different service or what's actually included particularly because we all package up our services a little bit differently. Price insights has been fantastic for us because it gives us that extra information about what firms like us offering similar services are charging. So when we add in that extra layer of pricing which is your competitor analysis, now we can actually do that. And that's really changed the game with our pricing and made us really consider where do we want to sit in the market? Do we want to be cheaper on this service? Do we want to be average? Do we want to be more expensive? And given us that confidence to just go forth. Love that. Sticking with that topic of you mentioned there, obviously, we have the experience, we have the knowledge as, you know, firm owners, anecdotal evidence even, you know, we know what's costing us more. How should an accountant use the result from these AI tools like Ignition's price insights? And instead of just simply accepting that suggested price, what can you do to expand on that with your own knowledge? So what I really love about using AI in this, and particularly the way that Ignition has rolled out Price Insights, is it gives you a lot of context around why it's suggesting this service. So it will sort of have high end and low end pricing in the middle where you're sitting and talk about the difference between what's the low end service look like, what does that normally include if you're offering a premium service, what are the additional items in there? And what's really important when you're looking at that is what is your service description? And this almost ties a little bit to that communication piece as well because if we are offering a stellar, for example, quarterly BAS service where we are reviewing everything in our clients' accounts, checking the profit and loss, doing like that, going really above and beyond when we're preparing that BAS compared to someone else who might be just pulling a report, filling it in and sending it off. If the descriptions that we're using in our services don't really explain what we're doing in that invoice when compared to somebody else, it's actually really, really hard to have that communication with your client because you haven't explained what that service actually entails. So when you're having a look at these suggested prices, read all the additional text because it talks to you about what is normally included in this service. If you're including more than that or potentially even less, then adjust your price accordingly to what you're actually delivering or put in some more input in your service description to get different pricing. You just reminded me of something. I wasn't going to talk about this in this webinar, but I'll talk about it briefly. And if I remember to show it, I will. Within each service in your service library in Ignition, which is your baseline services that is like your menu, your restaurant menu of services, you can use an AI tool that actually expands out on your service descriptions. So, you do something as simple as like tax return, okay, we can do better than that. Let's expand that using Ignition's AI, an inbuilt AI tool to expand on that service description and scope it out and actually start to build some of those scope items into there so you can have a better conversation. So, shout out to that AI tool as well, which again, if I remember to show you, I will. So next question for you, Bec. What does a firm need to get right for price insights to produce something useful? I think you just touched on it already there around that service description. Is there anything else? I mean, it's more about just the detail there with service description. So we both spoke about it and using AI to really elevate what that service description looks like is also an amazing tool. Wonderful. Now you mentioned the price range before. So these tools will spit out, you know, we think you should be priced between X and Y based on context. When might it be appropriate to price outside the suggested range? And I'm thinking either above or below. Yeah. So, I talked a little bit before around it might be appropriate to use a different price or have a different range based on what your service is compared to others or compared to the descriptions that's being thrown up. But it also comes down to where you want to position yourself in the market and the kind of clients you want to attract. Like if you are sort of looking for bulk engagements, you want lots and lots of clients and we're talking about volume, then you might be really, really comfortable to actually position your items at a lower range to win more work. So that is a driver outside of pricing and value. That is you going for volume or just wanting to win that extra work and you'll deal with the profitability of it later. On the other end, maybe you were just looking for very specific premium high wealth clients who are willing to come in and pay that higher amount. You want it to always be a bigger profit. So some of this stuff starts to really drive down where you need to understand who is your ideal client and what are their buying profiles and what are they comfortable with in spending. And that's additional information that you will own that not necessarily the AI is aware of. And then you can have a look and say, well, if everybody else is charging six hundred dollars for this service, I want to be premium. I want to stand out. I only want a particular kind of client. I'm going to charge seven hundred dollars Nice. Again, I feel like that's where just your knowledge of your clients and where your business is really comes into it. There's a question there which I think might be appropriate to answer now from Cameron in chat. I've seen some other accountants adding a technology fee as a line item. What's your thoughts on that, Bec? Is that something you would do or would you more just incorporate that into your overall pricing? We incorporate it into our overall pricing because we've always been a high-tech service offering. I have actually seen technology fees in other invoices from different places and even clients where we've gone out to a specialist to engage them. They always question: What is this technology fee? What am I getting for it? I think that given our client base, it doesn't work for us because we're high-tech and we pretty much only engage with high-tech people. It's a given that there's a technology fee. We do however, which is slightly off topic, charge now in some circumstances onboarding fees. It's not about technology, but about the difficulty of onboarding clients depending on what they need and making sure that that upfront gets covered. Nice, like that. Yeah, so I think it's probably depending on your situation kind of question. But I mean, me as a consumer, I'd probably rather see, wouldn't want to see a technology fee, because it would raise some questions. I'd rather just like understand the full scope of the services you're offering, including any technological output or anything from your end. So, anyway, that's just me, my opinion. Let's go into the third topic here, Bec. Thank you so much so far. This has been wonderful. Turning better information into action. So another question for you would be, how could firms incorporate price insights into their annual reengagement and pricing process, which is kind of happening right now? Yeah, and again, this means that you're not really stuck with that process of just putting everybody up by CPI, putting everyone up by five percent because it does give you those extra insights into not just what's going on in your business but what is going on with your direct competitors or however you're positioning your services. But what is really important here, again, we've spoken about bringing in that professional judgment, is remembering so what this does, it takes the heavy lift annoying parts out of it, like that assessment of your service and looking at the range and the suggestion about whether or not you're going to increase it. And that's what AI does. It's trying to take away those friction points to make sure that you can actually do this and do it quickly and roll it out on mass in your organization. However, don't forget every single time you use AI and we've spoken about this a couple of times professional judgment is still incredibly important. So you know your clients, you know your service, you know what everybody's willing to pay and where you want to take your firm. You can't just ever one hundred percent rely on the AI. It's taking out the stuff that we don't want to deal with, and then you have to apply a professional judgment over the top. That seems to be the theme, professional judgment. You've all got it. AI is going to help you get there ninety percent of the way. It's the same with any type of work, really. Know, AI can help you to an extent, but ultimately it's up to you to pull the trigger. Next question, a little bit to the side. We've just released our own AML or compliance feature here at Ignition, which is absolutely going off the shelf, which is great. How should firms approach these new costs like AML, like CTF, when reviewing their fees? I mean, this is probably the hot topic right now is I'm getting questions, multiple questions a day around this kind of stuff. So, what's your take on that, Bec, on running those reviews this time of year? Yeah, so this AMLCTF stuff is this extra layer of expense that we now have for existing clients, depending on what's happening with them and depending on designated service and also has added an actual real cost to our onboarding of new clients where it's not just good enough for us to check their driver's license if they're a new client to make sure that they are who they say they are. We are going to incur a lot more costs now in onboarding, which is why I started to talk about this onboarding fee. We've played with onboarding fees in the past if we know that we're taking on a large client and we're implementing new systems or we're implementing new processes. So we've taken that a step further and looking at that onboarding fee to cover the costs that we actually have to incur before they potentially even become a client. We might go through a bunch of checks and find out that they're not going to be a client at all. So, understanding and using AI and Ignition has been really great with this as well and the new AMLCTF features to identify really quickly what are those designated services and throw up which clients you're actually going to have to run those checks on based on you tagging your designated service. Taking out some of that, not so much the thought process, but eliminating the chance of you missing these steps and then potentially missing charging appropriately for your clients if you are incurring these additional costs and really being able to run through that system a lot more seamlessly all in the one product. Yeah, no, that's great. If anyone wants to know more about the compliance feature, AML, CTF stuff, then let us know. Again, reply to that email if that's going out afterwards or get in touch with us through your help desk and we can put you in touch with somebody to talk about the AML features, which has just been, yeah, like I said, been extremely popular and really well received. And we're always improving them too, which is great. Last question for you, Beck, again, around the turning better information into action. What's the first practical thing attendees should do after this session based on everything you've just run through? Look, pick one service that you have that maybe you charge most of your clients or if not all of your clients and just review that one service. Like what is the standard of the service that you're delivering? How does it compare? What is your average rate? You can do all of that inside Ignition and then apply then either the benchmarking or price insights and the extra information. Just that one service might make you a lot more money or help you stop losing a bunch of money. I mean we did that blanket with our ASIC annual review fees. We charge all of our clients every year for ASIC services on the tenth of July. I ran price insights. We were way undercharging for the service that we deliver, which completely shocked me because that's not normally our state of affairs. I ran using MTP price update to everybody, and in about fifteen minutes made ten thousand dollars So that was great. That's an incredible ROI. And there you go, Bec, like I said, I've known you for so long, you're one of the most tech forward practises, if not the most tech forward practises I've known. And even you still, you got caught by surprise because you, you know, there's so much going on right now. It's easy to lose sight of these things and just leave that money on the table. So I'm glad you managed to do that. And that's a good segue. Thank you, Bec. I'm going get you to stick around because there may be some other questions coming through. But thank you so much for that. I wanted to now actually go into a bit more of a practical look at some of these AI tools and maybe close the gap on some of those discussion points. So, if you have anything in the questions or the chat, then let me know. Happy to go through anything in more detail. What was the thing you were just saying, Beck? Oh, yes. If you want to know, I can show you this shortly. If you want to know what your most popular services are, if you're an Ignition user, it's actually on your dashboard. If you go in your dashboard, scroll down a touch, you'll actually see your services listed from most popular to least popular, and you can maybe start with those top one or two or three and run a price review on those ones. So, price insights, talked about price insights. If you're not familiar with price insights, I'm going to use this as an opportunity to tell you all about that. So, it's a tool that's available across all of your Ignition accounts. It's designed to give you tailored AI pricing suggestions from your services, and this can be applied to either your service menu, like I mentioned before, like your literal restaurant menu or services that you start with for every proposal, all the way through to actually doing it inside your proposal workflow, like it says there. It's going to give you an understanding of where your price sits in the expected ranges. So it's going to take the context of that service, the price you've got on there, all the deliverables and scope in that service, and then compare that against active services across all of our admission user base securely, and spit out what we think you are. You could be overpriced, could be underpriced, could be somewhere in the middle, and like Beck said, it's going to give you that range, but most importantly, it's going to give you that why, it's going to tell you why we think that you're undercharging or overcharging, and you know, potential steps to actually get towards the right pricing. It's three ninety nine pounds for a year's worth of insights and you get two hundred price insights. So, say, for example, you're sending out a proposal and that proposal has your most popular service in it and you want to just check for that client to make sure that you're not underhudging, then run one price insight and that will give you that insight and there's a potential there to spit out tens, twenties, hundreds of dollars worth of value just from that one price insight. So, it represents really good value in terms of reviewing your pricing. Maybe you can make ten grand in fifteen minutes as well. The pricing benchmark report is also important there. Bec mentioned before, pricing benchmark report really ties in well with our price insights feature. So, have a read of that when it comes out to you in the next day or two. Have a try of some of these price insights. You actually get three free price insights per account. So, I would suggest running those on maybe a proposal that you're putting together this week, or one that you've drafted up and you're still not sure on the pricing, that would be my recommendation. In terms of the prompts, let's just have a look at You know what I'm going to do? I'm going do the actual demo first, I'm going to switch it off a bit, just so you know what the actual process looks like. So this is just a little click through demo that I wanted to show you around how price insights actually works. So you can see here I've got a couple of proposals here in draft. When I go to edit one of those proposals or start a proposal from scratch, I've got each line item here, so I've got my individual tax return, for example, and you can see here next to each service, on any proposal that you create, there'll be an analyse price option here. If I click on that, it'll pop up a little item here, it'll bring me up the actual service name and then allows me to add additional context, which is what Beck was saying before. It's then going to analyse that price for you and any additional context. It's then going to give you that range. So you can see instantly here where your price range sits. You know, your current price, two twenty five bucks, We suggest two eighty dollars as a one off fee for your tax returns, potentially even more if you have additional context. And as I mentioned before, there's some reasoning here, tells you, you know, how we did the market analysis, where you're located, if you're in city or regional, industry positioning, service alignment, all other kinds of wonderful things here. So it's not just some number plucked based on what we have seen other similar services, it actually gives you some proper insights into why your pricing is where it is or where it should be. If you are happy with that, you can then use that suggested price and that will then, of course, update the price here. And that's it. That's how price insights work. So, I've mentioned before, you get three free price insights and then you can actually sign up and get those two hundred price insights for one year for three point nine nine pounds So I want to take you through some of these prompts. So you're probably sick of hearing prompts. Sorry, it's not going away. Prompts are going to be running a lot of the work that we do going forward, whether it's pricing or anything else. But here's some examples of how AI prompts can actually help you build out the context or the scope of your services, and then potentially give you money in your pocket because you've actually undercharged for so long now and now it's time to actually get that back in to the right pricing. So sticking with the tax return theme, if it's this time of year, here's some prompts that you can use on that price insights to say, hey, this is an individual tax return, but the client has multiple rentable properties, maybe even mention how many rental properties they have, mention that they have a crypto account and that can add complexity and time. So, the AI tool will take that into account and then price that up, either just for that client or across the whole board if you have lots of clients who are more inclined to have rental properties in crypto, if that's your specialty. Client's a sole trader with no bookkeeping records. Deck mentioned that onboarding fee, this is kind of where that maybe came into it. All the return is being lodged close to the extended lodgment deadline and it needs to be prioritised. And obviously, you should charge more if you need to jump the queue with somebody when it comes to things like tax returns. So, all of that can be considered as part of the scoping out for these AI pricing sites. Some other examples, I'll go through these quickly, but you can check them out in the recording if you like. Maybe the client's got some old, you know, some books that have been reconciled for a while, there's four month catch up. So, you know, you're doing your bookkeeping services, which you normally would, which is, I don't know, X amount for that particular service, but we want to add like another twenty five percent because we need to go back and catch up on something. We've then got things like, you know, more urgency, peak tax season, our team is at capacity. You know, we want to charge at a premium if people want to get our time. What client needs is turnaround within forty eight hours. I specifically said, I need this done within two days. No problem, but I'm going to charge you for it and the price insights is going to give you an idea of what you should charge for that. Client profiling and premium positioning, it's a high net worth client, we know that there's going to be a lot more involved here, there's an SMSF structure that's a bit complicated, so we want to reflect that in the pricing. Maybe it's something on your end, you're a boutique firm that positions on quality over volume, like Beck mentioned earlier, so please make sure that all our pricing is reflected in that. This may be something you might do to your service library rather than individual proposal services, because you want all of your services from scratch, its base level, to reflect the fact that you're a quality boutique firm offering high quality services. Not that you're all not offering high quality services anyway, I'm sure of it. Or even things like regional things, you know, less competition, make sure that we're not overcharging people in different regions because we're going be coming up against local competition. Margin and bundling, You know, my direct cost to deliver this is approximately one hundred and eight dollars and I need a margin. So, it will obviously take that into account. You can feed actual stats and figures into this, and that will give you the right price insight. I've already given you the tour, so that was what I mentioned before. I want to just show you quickly, there's two things, look at it, I remembered both these things. On your dashboard, if you scroll down, you can look into the Insights section and that will actually give you a breakdown of your services. So you can see here that my basic company, Trust Annual Accounting Services, accounts for twenty five percent of all my services, and here's the current month's revenue for that service. So, if you're on Ignition, scroll down to this, maybe choose that top one, run a price insight on that. I also mentioned before that if you wanted to do this in your service library, come into your services here on the main left hand side. This will bring up that. And you can see here, you can actually sort it through here, so I can click on the active services here and it'll sort it high to low, and you can see here that I can actually use price insights to analyse the price here, where I can actually get that additional context. So, if this basic company trusts annual accounting services is my most popular service, if you're using your free price insights, one of the three, I'll recommend doing it here just to get an idea of what your most popular service is in terms of fees, make sure you're not underpricing yourself, make sure you're sitting in the right area in terms of fees. So, if you're going to play around with it, I recommend having look at your service library to begin with. All right, MCP, I mentioned MCP earlier, I spend too long here because I know it's something that's fairly new to everyone. In fact, I'm not even going to sit here because this is a bit of text that's not really going to be relevant. I've built a little understanding here, a little visual on how the MCP works. So if you don't know what MCP is, it's basically a way for your Claude or your ChatGPT or some sort of AI agent to connect into Ignition and do stuff for you. That's pretty much what it is. Bek said before that she did a bunch of repricing for ASIC using NCP in fifteen minutes. She would have prompted Claude to find all the last year's proposals that had ASIC services, get them together, add an X amount increase, percentage or dollar value, and then draft them, and then we can send them out to the client. So you simply ask your chat agent, Claude, for example, what you want it to do, MCP will then connect into Ignition account securely, Ignition will then do the work based on that prompt from your MCP, from your agent and then you will still need to review and send that. So, don't worry, we're not going to send anything to your clients, we're not going to send out, you know, a hundred ASIC proposals underpriced for some reason, you'll still be in charge of the final review and send, which is where AI is right now. So, the MCP is huge. If you need more information on MCP, we can let you know. We do have a support article on our website that takes you through it, how to get it set up. So, if you are even remotely nerdy like me, or you just want to be more efficient and automate as much as possible, I recommend having a at the MCP, it's been very well received as well. This is coming out soon, we've got an MCP guide, I think it might be out now, we'll make sure you get a copy of it and taking you through exactly what I kind of talked about in a bit more detail and how to get it set up and how it relates to the real world and how you can use it on your day to day. We've also got a free AI workshop coming up soon with Inbal and Ben, two absolute pioneers when it comes to AI and things like MCP. So if you want to register for that, there's a link up on the screen now. I would recommend doing that. Like I said, if you're even remotely interested in automating all the things, then that would be a good place to start. That's on Monday, the twentieth of July at eleven am. And would you believe it or not, that's all we've got. Thank you so much for your time. Thank you, Beck. Did I miss anything, Beck? Anything else you want to add? No, you did great. Covered it all off. Thank you, Beck. No, I appreciate your time. Thank you, everyone. Thank you for your participation. Thank you for your questions and chats. Question there from Amy, just before we wrap up, is the price insights data primarily relevant to Australian firms or where were, yeah, and Z firms too. So definitely we've got that and we make sure to regionalise that. So check that out. Sure. If have any questions, Amy, though, let us know. You see anything that looks odd, let us know. But thank you, everyone. Great to have you here. Thank you for joining me. Enjoy the rest of your day. Good luck with the rest of July. Good luck with tax returns. If you need anything from us, please let us know. Talk to you soon. Bye. Bye, everybody.