Why you're losing the AI battle

Aden: Next up is Why you're losing the AI battle, all about the habits holding firms back from getting value out of AI. So please welcome Amy Holdsworth, Director of Clarity Street. Amy, welcome and thanks so much for being here. We'll let you share your screen and let you go.

Amy Holdsworth: Thank you so much, Aden, I really appreciate it, and thank you so much to everyone you just mentioned, without re-mentioning them all again, for inviting me on and allowing me to chat today. So let me just share my screen. Dear God, I hope I have the right one. Is that the right one, Aden?

Aden: Yeah, that looks good. Awesome.

Amy Holdsworth: Yay! Okay, great. So I think I've got about half an hour, 45 minutes give or take, depending on how many questions there are, to talk through today's session on why you're losing the AI battle, and I guess why AI won't fix an integration problem. I did put a little hello into the chat to everyone, so feel free to keep the chat active. I am always keen for that. I'll try and answer things as I go as well, if I notice them. But otherwise, thank you so much for joining. All right, let's kick things off.

Amy Holdsworth: This is obviously deliberately provocative, right? Why you're losing the AI battle. But we will be unpacking this, because there's lots to unpack when it comes to AI. It's the newest, shiniest toy on the market. I was around when the cloud wave happened, from server-based to cloud technology, and that was huge. And I feel like it's the exact same thing. AI is like the new cloud, if you get where I'm going with that, from the tech evolution perspective in the accounting industry space. So, a little bit about me, for those who don't know me. My name is Amy Holdsworth. I am an ex-practice manager. I have been in practice for the last 20 or so years, either running or helping other practices, basically. Sorry, Cecile, that is hilarious. What is cloud? So cute. I've been implementing accounting tech for the last 15-plus years.

Amy Holdsworth: Close to 20 years, actually, looking at firms of all shapes and sizes. I run and own Clarity Street. I'm the founder of a cloud integration business. We started about 10, 11 years ago, helping firms implement technology, apps, etc. I'm also a podcast co-host of the All Aussie Accounting Adventures Tech Edition, so if anybody loves a podcast, feel free to geek out on that one. And in my spare time, I'm either off decorating cakes, feel free to follow me on Instagram, Redcoatcake, or playing with my little fluff button featured in my photo there. Her name is Agatha. She gets mentioned quite a bit internally. So that's a little bit about me.

Amy Holdsworth: But also, who is Clarity Street? We are a cloud-based business. We design and deliver dynamic, tailored best-practice infrastructures for accounting firms, but what does that actually mean? We spend a hell of a lot of our time doing projects and consulting work with accounting practices. We have all come from practice in some way, shape or form, so the bulk of what we do is either help accounting practices transition to a different app stack than they're already on, typically from a server-based app stack to the cloud-based apps out there at the moment, or, for the clients that have already done all of that, we help them optimise what they've got. And now we're starting to have many conversations about AI as well.

Amy Holdsworth: Those projects and that consulting work are powered by our Clarity Campus, which we designed: online coursework. So self-paced online learning courses. Watch a bit, read a little bit, do your knowledge check as well. It's powered by, I think, about 1,200 different knowledge-based articles on all the different processes and procedures you could ever possibly think about when it comes to running your practice from an operational standpoint. And then my little love child is Project 95. There is a heap of support out there in the accounting industry for accountants and business owners, but there appears to be very little for the administration role, the operations role, the practice managers out there. So Project 95 is small group coaching for our practice managers and practice ops. Accountability, monthly catch-up calls, submit your figures, things like that, and I love it. It's great, it's very empowering.

Amy Holdsworth: So anyway, that's a little bit about Clarity Street. We have spent a lot of time in the accounting industry. We've all come from practice in some way, shape or form, and that means we're not just some random IT people who woke up one day and went, let's support and service the accounting industry. We've lived and breathed it quite a bit. So, enough about us. Let's move on to why we're actually here. Because there is a massive conversation happening in every firm right now, and it's usually with a practice owner or someone who's just spent time doing something they hated doing, right? And the issue we're currently facing is that it's really challenging to diagnose things, honestly. Is it the people? Is it the processes? Is it the technology? And understanding what the fix looks like for each one.

Amy Holdsworth: We are seeing a lot of disconnect between workflows, and a bit of desperation for wanting those workflows connected, genuinely connected. Engagement, signing, job management, your document storage, all the rest of that kind of stuff, your onboarding. Onboarding is a very hot topic at the moment. We're being asked a lot about where AI actually sits in a firm, where it adds real value, when you should use it, and when you should consider that maybe right now isn't the right time to adopt it, which we'll unpack. And sometimes, when you've turned it on without thinking about those things, how it just exacerbates and creates absolute chaos. So the other thing we're hoping to bring you today is first steps, some practical moves on how you can actually build some resilience in practice around using AI as well. So that's why we're here today.

Amy Holdsworth: All right. Technology was supposed to fix this, surely, right? So AI is here. But wasn't tech meant to fix all of those problems? The cloud apps with open APIs that talk to each other, weren't they all meant to fix all the problems we currently have in practice? Because from my perspective, that's what most accounting practices and businesses we deal with actually think. The problem is that it hasn't. It just hasn't. Or it has, but then the rules change. The ATO dictates something, or AUSTRAC dictates something, or ASIC dictates something. You get where I'm going with this. The rules change, which means the apps then need to evolve and adapt and change as well. So maybe they did solve the problems. In my opinion, I don't think any one app solves every single problem out there.

Amy Holdsworth: I also don't necessarily believe that a combination of apps solves all the problems. I think there will always be a need for some human intervention. I don't think humans are going anytime soon. We definitely still need to have the knowledge and the advisory and the relationships, if nothing else. But the point here is that technology was supposed to fix the problems and it actually hasn't. But it was never really a tools problem, right? The tools actually exist. They were just never designed to deeply talk with each other. What I mean by that is, if everybody understands the concept of cloud technology and having an open API, the APIs are there for a lot of apps, but that API language is 10-to-20-year-old language to a certain degree, and newer as well. But there is actually a new language model out there, the AI language model, which has obviously been out for a little bit, but more so, it is now accessible.

Amy Holdsworth: So instead of the terminology API, you'll start hearing the terminology MCP. Where I'm going with this is, APIs allow you to go from this point to this point. It's a very direct connection. Whereas an MCP, the new language model that a lot of the apps and the AI world are being built on, allows for a much more dynamic conversation and connection point between apps. Which is exciting, right? But the issue is the disconnected systems: the best-of-breed apps that still have, call it, traditional APIs. I think a lot of them are in the process of scrambling to change their language model to an MCP, or update it. The best-of-breed apps are good, and they have been good for quite some time, but they've never actually shared all of the information you needed them to.

Amy Holdsworth: The data sometimes moves by hand from one system to another, because that's the other issue: where the data doesn't connect or push and pull between apps, you need someone to physically key that information in, relying on a human. And that means that usually the best people are filling that gap. And forgive me for saying this, but accountants are really good at throwing a human being at a problem, as opposed to thinking about how to make the apps or the technology work better for themselves. In a traditional mindset, we have just gone, Betty will do that. Sorry if there are any Bettys on the call. Or Amy will do that, she can just do that little process. But is that Amy's best skill set? What if she's a qualified accountant, or what if she's been running all the operations and the administration for years? That's probably her role as well, but is that the best use of her time and skill set? The point is, the tech can actually do a lot of this for us.

Amy Holdsworth: Okay. All right, next point. The other thing to know is that the someone doing the manual fill is probably doing work that adds nothing to the client relationship and nothing to the bottom line. So we want to be able to flip that on its head a little bit. Okay? Don't worry, AI will fix it. How many times have you heard that today? For all those people that have been logged on all day, if you haven't heard it multiple times today, you've probably heard it multiple times very recently. For some things, it will. Definitely. AI will fix some things. But it matters enormously what you put AI on top of. For example, AI on connected systems means clean data flows end to end, based upon processes. It means you will have people freed up for client-facing work, which is what you obviously want, and you then get compounding efficiency.

Amy Holdsworth: Okay. However, AI on disconnected systems, or on poorly defined processes, just speeds up the inefficiency. Errors multiply faster than before. There's a heap of manual work just to fix or move around errors, and you get chaos. You don't get efficiency, you get absolute chaos. We have a lot of accounting practices that come to us and say, all right, this is my tech stack: my practice management, my document management, my engagement software, for example. They're all talking together, and we've pretty much got it mapped. Or they haven't. They think they have it mapped, but they actually haven't. I want to put AI over the top of this and streamline this process. I'm like, okay, cool, that's wonderful. Have you got the process written down? Well, no. Okay, do you know what the process is? Oh yeah, I've got a pretty good idea. I'm like, okay, great.

Amy Holdsworth: If you don't have your process specifically mapped out and written down, and know exactly what buttons you're clicking, how it happens, where it connects and what it does, how is AI going to replicate it and remove the human being component from that? So you have to understand those kinds of things, and that's basically what happens with AI on disconnected systems. If the systems aren't talking together, if the data flows aren't working and the processes and procedures aren't actually being followed and monitored, then it's a crap-in, crap-out scenario, but the crap pile is so much worse when it comes to AI. Basically.

Amy Holdsworth: All right. The firms seeing real results have done the unglamorous work first, which is basically what I was just saying. They have 100% understood their processes. They have mapped out, as I said, every click of the button, the actual workflows, how they actually do things. You could talk about it from a robotics perspective, in a way, as well. You want to know where you are clicking on the screen, not necessarily to the nth degree when it comes to using something like Claude, but you need to know exactly where to look, where to point it, what you're actually trying to achieve as the end goal, and what those steps are. Because if your processes aren't mapped out, then how are you actually going to teach team members, but also teach your AI tool, to do that process effectively? So they have understood their processes. These are the firms that are seeing real results. And don't get me wrong, it is boring as bad, but it has to be done. Just putting it out there.

Amy Holdsworth: They were really honest about their legacy apps. And I say the word legacy, but legacy is not quite fair. I think it's more about the apps you're currently using right now. So understanding your practice management, your document management, your engagement software, your invoicing software, whatever your tech stack is internally. What can each of those platforms genuinely do and not do? Just because you don't like the user interface isn't a good reason to say it doesn't do this. Or, I just don't like using it, is also not a good reason. So you need to be mindful of those sorts of things. It's very much understanding what it can and can't do, and I say that deliberately because a lot of apps out there are consistently evolving. Most firms we speak with have only ever used about 5 to 10% of the functionality of most of the apps they use on a daily basis.

Amy Holdsworth: And there is so much more functionality. The build on a lot of apps is actually quite rapid. Some slower than others, sure, but some really fast-paced. I would agree, sometimes it is hard to keep up with all the different changes, which is why I'm suggesting you need to understand what your apps can and can't do, and where the functionality crossover is between each app. You need to actually look at that functionality. Are you using all the different functions and features? Do you have it connected? Do you have it talking? Or are you using them in siloed instances, so they're not actually talking to each other? I'll use this as an example. If you have Ignition, Xero Partner Hub, which is XPM for those who haven't quite caught up, and you're using a Xero file for your invoicing,

Amy Holdsworth: if you have those three apps, hypothetically speaking, but you're using them individually, so you haven't got Ignition connected to XPM to deploy jobs, and you haven't got it connected to your Xero file so it deploys invoices, then you're missing out on a huge amount of functionality from the combination of the three apps working together. So it's understanding how apps integrate, how they are used from a workflow perspective, and how you can apply those workflows that's actually really important. And I would suggest you do that first, before you decide to go, let's chuck AI on top of all of this. Because AI is not going to fix the fact that the apps don't necessarily talk, especially the ones that don't have all the right language models behind them.

Amy Holdsworth: Anyway. The other thing they've also done is understood their people. They have understood that the team actually makes the automation work. They're not the obstacle. They've empowered their teams, they've involved their teams, and they've helped ensure their teams have gone through the process of understanding what their roles and responsibilities are, what the processes are, and how they should be doing things. There's a lot that these firms have actually done. As I say, this is the boring work. I mean, it's the good work, I love the work, but this is the boring stuff you kind of have to do. These are the firms that have done this bit. So if you do this bit, you are setting yourself up for success for the next phases, which is adding in the AI layers on top.

Amy Holdsworth: All right, five things to get right before you scale AI. As I mentioned, AI alone will definitely not solve an integration problem. An integration problem is data and processes. Dropping AI over the top of systems that don't talk to each other, as I just discussed, is just going to exacerbate an issue. And if you're expecting it to remove a manual process, it just hides it, and it speeds up the mistakes. So you need mapped processes, so there's a defined path for the AI to follow, step by step. You need to make sure your data is connected, so information flows from one point to another and stays consistent across the systems. The biggest pain point I see in a lot of firms is inconsistency in database information. I see many other inconsistencies, just so we're clear, but one of the biggest is data. Names, addresses, salutations, email addresses, the basics.

Amy Holdsworth: Making sure all of those things are populated means all of the data can flow cleanly between your platforms and populate all the relevant information needed to run the other platforms. Tax to practice management, practice management to engagements, engagements to invoicing, invoicing to cover letters, all of those kinds of things. You need to make sure your data is clean and moving in the right direction. The other thing is ownership, so that people actually know what the AI handles and what stays human. That's a really big one, because you need to get your team to trust the process as well. You really do. You need to get your team to trust that AI is helping them, not hindering them. I think that's the best way of saying it. It's helping them be more efficient. I will tell you, though, if you do go through this process and then throw AI over the top of it, it will make human beings a lot more accountable.

Amy Holdsworth: And the irony of the accounting industry is that sometimes accountants aren't as robust with being accountable. So this will really highlight a few of those components. Forgive me in advance for that comment. It's just something I've noticed. But it will make people a lot more accountable, so you have to make sure it's clear who owns which part of the process. People, process or technology: to my point, be honest. What's the problem you actually have here? Because the fix for each one looks different. A lot of firms come to us thinking they have a technology issue. And I can tell you now that technology makes up probably 10% of any of the problems you're experiencing within your practice. 90% of them are people or process related. Either the wrong people doing the wrong thing, or not doing the right thing, or just not doing anything at all. In relation to processes, people not following them, which is ironically a people problem.

Amy Holdsworth: But most of the time, processes aren't defined for things like that. I will come back to that comment, Blake, but in relation to AI being used to manage client contact data: yeah, it can, but you need to be careful about what data it's managing, for obvious reasons. You're all accountants, you deal with sensitive information, so just be mindful of that. That's my main point there. So back to this. In terms of the people, a best-of-breed person is quietly acting as the integration layer, so we want to fix that.

Amy Holdsworth: We definitely want to clarify the roles and the ownership, so no one's value is spent re-keying. Everybody knows how challenging it is right now in the accounting industry to find really good, valuable human beings. Everybody is feeling that pain. So when we do find a really good operator, we want to ensure they're actually doing valuable work and delivering on that value, because it's hard enough to find a human being. So let's ensure they are in the right roles and adding the most value.

Amy Holdsworth: Processes: the workflow is undefined or undocumented, the gap that nobody owns. As I mentioned, are the processes actually there? Are they documented? Do people actually know what they are? Are people following them? Are people being held accountable to the processes when they do or do not follow them? The fix on that one is to map how work really flows before automating any of it. Then the final one is technology: tools that were never designed to share data cleanly with each other. Call it a desktop app with a cloud app, and there might be a little piggy in the middle, a little app that sits there and tries to make it work. Ultimately, connect what you have. Replace only where it genuinely can't stretch and can't actually connect. That's something to be mindful of. I am not a right-AI's-here-let's-bin-everything-in-our-tech-stack kind of human being.

Amy Holdsworth: I don't think we're at that stage yet. But I do see value in what I say to everybody every year: as a minimum, you should review your tech annually. Is it still providing value? Is it still doing what you want it to do? Is it still allowing you to actually process work? All of the above, basically. So I don't think you can throw it out just yet. What does a connected workflow look like? Point number three. Data captured once, flowing from engagement to storage. I'm actually going to give you a bit of an example here. What you want is no unnecessary re-keying. So hypothetically you want it to go from client engagement across to signing. Once the engagement has been created, you make sure the client is sent off for their signing. There are a multitude of different options for how you do that and what apps you use.

Amy Holdsworth: You can use an application specifically structured for it, like Ignition, just picking one. You could use a PDF version of an engagement letter and send it out through a signing solution, like FuseSign, again, just to name one. Once it's signed, it should then hopefully deploy into your job management, your workflow management. There are a raft of options out there: XPH, XPM, FYI, Karbon, AccountKit, Imperia GPS, there are so many. There are loads you can use. But make sure it's deploying into that, and then the final one: once you actually do the work, you want somewhere for document storage. Now, this is a very baseline one, but basically, where possible, all of those apps should talk to each other. Once you do the first thing, the next thing should auto-deploy, and then the next thing should auto-deploy.

Amy Holdsworth: There is a bit of a caveat to this. If a client's details are typed in more than once, the workflow isn't joined up yet. Now, I say there's a caveat deliberately, especially in the Australian market. We have incumbent applications out there, which unfortunately means it's out of our control and we're unable to ensure everything is connected. Thank you, ATO. Thank you, ASIC. For example, you can't do the onboarding process from start to end. You just physically can't. You need to get your clients to intercede at some point, to add you to the portal, hypothetically. So client-agent linking. There are a few caveats to that, and it does mean that sometimes data is entered in twice, or three times sometimes. Where possible, the fewer times you have to add or change or type data in, that's what you're ultimately looking for. In a perfect-world scenario, you should only realistically have to add a client detail once, into one of the platforms. As I said, caveat: ASIC, ATO.

Amy Holdsworth: What does that look like in reality if you had an app stack? Now, I need to put a very big caveat on what I'm about to show you in terms of apps: it's just one example of how the apps could connect if you have this tech. If you are using other apps, insert them in the relevant places. I'm just using this as one example. There are so many different ways you can skin this cat in relation to which apps, but ideally what you want is for the apps to talk with each other, which is exhibit A. So in the morning I am going to jump onto my Practice Protect portal through my single sign-on solution, and once I open that up, I'm going to have all of my little apps there. I'm probably going to check my emails through Microsoft and check whether we've sent a request to a client saying, hey, it's time to do your work, we'd like to call your work in. Or a client has said, hey, can you get my tax work done? Or whatever work it is.

Amy Holdsworth: I'm then going to go into Content Snare and do a content request, so request information. This is for a current client. There are other nuances when it's a new client, just so we're clear, but for current clients, I'm going to send a request from Content Snare for all the relevant documentation and information I want. Once that's done, I'm then going to create an engagement in Ignition, which is going to be a proposal with all the relevant services I'm offering. I'm going to be very transparent about the services and the price I'm charging, and I'm hopefully also going to take payment upfront, because I am a very big fan of that one, where possible. Off the back of an accepted proposal, it should flow through to Xero Partner Hub, Xero Practice Manager for those who, again, aren't quite sure of it. And yes, it has a different colour, for those who are used to it being green. It's changed. It's now dark navy blue.

Amy Holdsworth: The job should automatically deploy into XPM, which means you can then go off and do the actual work. Because a job's sitting there, you can connect up a tax form. You can then make sure that's connected to your ledger, your financial reporting and, if you need to, workpapers as well. If, on the other side, you need to do anything CorpSec-related, you've got CAS 360 there, and if it was a self-managed super fund, you've got Simple Fund 360. That depends on what work you're actually doing. Once you've done the work, because all of this information links together, you could then do your time and billing through Xero Practice Manager or back through Ignition, whichever way, because the information syncs. The invoicing and payments, if you haven't already triggered them through Xero, could be triggered off the back of Ignition, or through XPM, or through Xero directly, and that should all link through to your document management solution, which in this instance I have chosen as FYI.

Amy Holdsworth: Once you've done your document, you've packaged it up and you've sent it out, we're going to send that out via FuseSign for the client to sign. Then once it's all back, lodged and saved, I'm going to use ATOmate to send out any ATO correspondence I need to for the client, which will also autosave back into FYI. So as I said, this is just one example. I'm not saying this is the be-all and end-all, because there are lots of variations, and lots of firms have different opinions on the app stack they use, and reasons, good, bad, indifferent, doesn't matter, that suit their firm. If you want further information on any of this, I'm happy to chat further with you, but that's just one example, so you can see that these apps all actually talk to each other. In an ideal scenario, this is what you want. By doing this and then adding AI over the top, this is where you'll start getting efficiency gains.

Amy Holdsworth: Hope that's helped a little bit. But legacy apps aren't always the villain, to the point I was just talking about. The idea that a new platform solves everything on its own, ripping out workflow systems, etc.: I think you really need to be conscious of what you're doing. And again, I say legacy apps when in actual fact it's probably the apps you're currently using, realistically. So it's something to be mindful of: it may not be the app. Dare I say it, it could be the operator of the app, or the team using the app, that is potentially the challenge there. So keep what works. Don't throw the baby out with the bathwater, so to speak. If it does the core job well, it can stay. For example, I'm just going to pick on it, because I was using it already: Xero Practice Manager. There are varying degrees of love for that little product, and there have been for quite some time. I've been using it for nearly 20 years, since it first came onto the market as WorkflowMax, way back when.

Amy Holdsworth: Anyway, there is a very big love-hate relationship with that product in the industry, but it does everything it needs to do for most firms using it. Are there better apps out there? Potentially, yes, but not always. And the one you've got might actually be completely and utterly fine. So just remember that. If it does the core job well, it can probably stay. Bridge the gaps: connect the edges and bridge the gaps with integration before you replace it. Really look at the integration of the apps before you decide to go, I don't like that, that's not right, it's not fabulous. Make sure it's integrated, and then also be honest about its limits. Only replace an app when it genuinely can't stretch to what you want it to stretch to. So just keep those little tidbits in mind.

Amy Holdsworth: All right, step five: some practical first steps for directors and owners. This will help you build a more connected, resilient practice, which will also set the scene for you to add an AI layer over the top of things, and I think to really trust that it's working for you as well. So, pick one workflow and map it. Pick one real workflow. Pick a client journey and document how it actually runs. I'd suggest the one that's annoying the crap out of you the most is probably the one you should pick and try to map, because that's probably the problem child that actually needs some love and attention. If it's the biggest bugbear in your practice, whether for you or for a team member, pick that. Also consider that as a business owner, it might be different to what the team is screaming at you about as a pain in the butt. But pick one and actually map it, step by step. Map the full process.

Amy Holdsworth: Then, once you've done that, ask: what do we have? Do we have a people, process or tech problem? And answer that honestly. You might be surprised, might be very surprised. Sometimes the fix isn't easy for any one of those, but it can potentially be easier than you first think.

Amy Holdsworth: Audit your legacy apps, or the apps you're currently using. Look at each system. Coming into Christmas, I appreciate our brains are all possibly a little bit fried at this time of year, and very soon we're probably going to want to start winding down. But I'll also tell you that you've got about a month before you're allowed to wind down. Just giving you the heads up. You've still got loads of work to do beforehand. But in that lull, realistically, clients don't want to know you after about the second week of December. They've already switched their brains off. And unless you've called in a heap of work for your team, they're not really going to pick things back up again until after Australia Day. So bear that in mind as well.

Amy Holdsworth: Which means you've got about a six-week period where, unless you've called in a lot of compliance work to process over that period, it's going to be a little bit challenging for you to... I mean, you've got to have some free time. So this is a good time to actually review the apps and go, what's working, what's not working? Where are we getting touchpoints that are breaking down? So audit it then. Connect your apps before you replace them. If you're having any challenges around that, either speak directly with your app vendor, or jump on any of the forums, the Facebook groups, that kind of thing. You're welcome to reach out to me as well to understand whether the apps are connected correctly or not. Just make sure they're connected, to close some of the gaps in your connected apps. Once you've done steps one to four, then and only then may you add AI, and apply it where the process is clean and the value is going to be clear.

Amy Holdsworth: Because as I say, after you've done one, two, three and four, you may go, hmm, I don't necessarily need the AI layer for this process. Very big fans of AI, just saying you may not need it, depending on what you've solved by going through steps one to four. The final one, and I cannot stress this enough, is bring your team with you. If you're in any way similar to how I felt about AI up until about six months ago, which was that it scared the crap out of me, genuinely, and you should ask Trent about that, because I would have conversations with him frequently about it. How I got around that was, while my team were getting really excited about it and I was a bit, I don't know, and sometimes it's the opposite, you might be excited and your team's not, how we solved that was I basically forced myself to be involved and said, show me.

Amy Holdsworth: So we have a 30-minute lunch and learn every Monday. We share screens and go: this is the cool thing I've learned, this is what I've done, this is what I'd like to do, and hey, can I connect this up? And I go, whoa, hang on a second, I don't know whether you should do that yet. But that's how you actually get them involved. So bring your team with you, because it will make for a much better journey. Get the foundation right, then let AI amplify it. It's typically an integration problem first. AI adds value on top of connected data and defined processes, not instead of them. Diagnose before you buy. Whether it's people, process or technology, the fix for each is different, as we went through. And then finally, bring your people along. They're not an obstacle to automation. They are what makes it stick, so bring them along for the journey.

Amy Holdsworth: I think I was actually on time, a little bit, so thank you, everybody. I believe there are a few comments or questions in the Q&A that I should get to. All right: Amy, how much is the monthly cost per user to have all of these apps? Mmm, good question, Claudia. How long is a piece of string? It somewhat depends, if I'm being honest, because not all apps charge equally per head. I have a cute little, forgive me, Excel spreadsheet calculator that goes through how much all the apps cost. So if you want to hit me up after this, I'd be happy to speak to you about that one. Xero costs are dependent on the number of ledgers you have. That's how they charge, just so you know.

Amy Holdsworth: What else was there? I think the biggest issue some places are facing is trying to connect the services they've previously used and are established with, only to run into issues where there is no compatibility or current option. It means the question is whether to fill in the gaps with human work, wait for the options, or start all over. I feel like, yes, I understand that part, but I think not everything is changing so rapidly that there isn't going to be an evolution and a step-by-step process you can apply to that sentiment, Malachi. Yes, there are some services that are going to get changed and tweaked, but not everything. So baby-step it. The accounting industry is fast-paced, especially when it comes to technology, and especially the Australian market. We are so incredibly lucky. We are so ahead of the curve when it comes to technology in the accounting landscape, genuinely. Sometimes we forget that.

Amy Holdsworth: I think we do have the opportunity to go, all right, what's actually working and what's not working, that kind of thing as well. Aden or Trent, are there any other questions I've missed? Did I miss any other comments in the chat I need to go back and look at?

Host: Not that I have seen. No, I think you're good.

Amy Holdsworth: Yeah, cool. Amazing.

Host: Excellent. Well timed.

Amy Holdsworth: Let me stop sharing my screen.

Host: That was awesome.

Amy Holdsworth: Appreciate that. Appreciate it.

Aden: Well, thanks, Amy. Really appreciate it.