The Sage Group plc (SGE) Earnings Call Transcript & Summary

Hello, and welcome to Sage's Innovation and Product Webinar. I'm James Sandford, VP of Investor Relations here at Sage. And on behalf of the executive leadership team, thank you for joining us. We've got a great agenda planned for you today, but please note that we will not be giving a trading update and no new material financial information will be provided. [Operator Instructions] And finally, the event is being recorded and will be available for replay on our Investor Relations website. We have a lot of great content for you today. So let's get started. I'm pleased to hand over to our Chief Executive, Steve Hare.

Thanks very much, James, and welcome to everyone. We're coming to you today from the U.S., and we're here in our offices in Atlanta. This is the third in a series of these types of webinars. And the aim today is to give you insights and visibility both from -- into the team, but also how we're approaching innovation here in Sage. So we're going to do a bit of a deep dive on what we're doing around innovation and particularly what we're doing around AI. And what we'll also do is connect this so that you can see how we're driving value and how we're doing things differently as we look forward. And so there's a few takeaways that I just want to set out upfront. First of all, the network -- the Sage network is our platform. So that platform and our ecosystem is what helps us to drive innovation. That network platform is also how we drive our global cloud solutions. And those solutions are in turn driven by our AI-powered network. So the AI is embedded and is an integral part of that network platform. So as James said, we're not just -- it's not just James and I, we have a full team here ready to talk to you. So why don't we start by meeting the team. So first of all, we have Walid Abu-Hadba, who is our Chief Product Officer. And also today, we have Dan Miller, who is also the EVP of the mid segment, and that obviously is powered very much by Intacct. And we have Aaron Harris, our Chief Technology Officer. So we have a big lineup who can really take us into the -- deep into what we're doing. So we'll give you the key kind of overview in terms of the strategy, but then we'll take you down into the detail of what we're doing. And in particular, this linkage to how we're turning that into outcome and values. So before I hand over, I just want to talk, again, come back to this kind of anchor slide, our strategic framework for growth. Now what innovation allows us to do is to achieve our purpose of knocking down barriers. It underpins and enables our ambition. And what you'll see during the course of the next kind of 1 hour, 1.5 hours is how much we come back to that network. So to be the trusted network for SMBs is such a crucial part of how we're going to deliver value. And that's also how we deliver on our priorities, which remain unchanged from what you've seen before. It remains as critical as ever to scale Intacct. That's central to our ambition. And increasingly, you're seeing how we're doing that globally, not just here in North America. Expanding that proposition, so solving more of the CFO pain points. Continuing to build that small business engine, essential to the scale of the Sage network, scaling that network, increasing participation, increasing the connectivity, allowing our customers to connect with who they need to, whether that be their customers, suppliers, government, that digital interchange. And continuing to learn and disrupt, innovation being at the heart of what allows us to create these amazing outcomes. So the trusted network is the platform. This is how we deliver amazing solutions to all of our customers, wherever they are in the world and whatever products or outcomes -- whatever products they're using, this is how we bring it together and create that connectivity. Effectively, this is how our customers are able to use -- are able to deliver and consume their back-office solutions from Sage. And what I'm going to do now is hand over to Walid, who's really going to take you and show you how we turn that into reality. Walid?

Thank you, Steve. As Steve said, we are right now introducing our Sage network platform. But let me step back for a moment here and tell you what's important here. In the industry right now, we're living actually in the most exciting time since probably the era of the PC. We are at the convergence right now of 2 massive technology shifts. The number one is AI and machine learning, which will disrupt everything we do and will disrupt this market in better ways, in a good way. But also, we are at the intersect of what we call connected accounting. For a long time, the accounting profession has been automated. There's a lot of [ tests ] being automated, but it's still what was living in islands. Suppliers and customers and accountants live in different islands. They have to share things through maybe some files, emails and so on and so forth. What we are -- what we have built over the last 4 years, and I'm going to take you through it right now, what build a digital network that allows suppliers, customers, vendors and all types of interactions to come together in a single environment. And also think of that applying AI and machine learning to that environment will give us a tremendous amount of benefit. The benefits are massive, and it's for our customers, for our vendors, for the suppliers, for ISVs, but also internally for Sage. As the Chief Product Officer, I always say, I've never been in a role or a job where I never had to ask for resources for the past couple of years, incremental, because of the amount of AI, the amount of innovation that we bring in made our developers are so much more productive. Now our -- in just short, our Sage network platform is a collection of services and products that sit at the heart of everything we do. We've centralized all these services, from banking services, for example, AR, AP, accounts payable, accounts receivable, workflow automation services, all these type of services are sitting right now in a single platform, we call this Sage network platform. That allows us to actually interact with it through a standard way of mechanism of interaction. It allows ISVs to actually go ahead and build value-add applications on top of that. But also importantly, it allows us to introduce AI and machine learning at the network level -- at the platform level, where every application, therefore, will be able to take advantage of that. And Aaron will spend time with you today taking you through that example. And Dan will spend time with you to take you through the product examples behind that. Now keep in mind, it's very important to keep all these things connected together. That is the key ingredient here. This is where Sage excels. We are the #1 platform for small- and medium-sized businesses. There is nobody broader or deeper than us. We cover medium and small. A lot of our competitors either cover one side or the other, but we cover all of them, and we operate in more geographies than all of them combined. Now I'm just going to give you through some data points to show the depth and breadth of our platform. Just take a look at these numbers out there. I just want to take a couple of these numbers. We have 50 million interactions today with vendors. That means we have known about the graph, what we call it. We understand how suppliers and customers interact with each other. We know when they send the invoices. We know what the invoice looks like. We understand that. We have 300,000 bank accounts lit up on our network. That means 300,000 customers who initiated these bank accounts in there across 11 countries. Another important one. We actually have 10 million individuals who get their payroll, and they get their payroll stubs and their payroll information through our Sage network. This is very powerful. We already have -- our platform exists, and it's very deep and broad. Now this is really kind of very exciting to me because as a Chief Product Officer, it's really fantastic to talk about introducing new products. In the past, for us to introduce a product or a service, it would have taken -- let's stand up a whole development team. Let's go ahead and start working together on this, and it was discrete. With the invention of the Sage digital network or the Sage network platform, we're allowed -- we can bring in products and services at a record time. This project here, what we call Sage Inbox. Sage Inbox is a tool that will be available to all our products and actually competitive -- our products and other customers who do not use the Sage general ledger. This will be available to everybody that allow you to fully automate, for example, accounts payable, accounts receivable. And it's -- by the way, it is an open platform. What that means simply is ISVs and developers will be able to actually put their own automation services inside this platform, and they will be able to use the digital network platform, services, like banking services, authentication services, workflow engines, compliance engines that will be available to them, and that will create a tremendous amount of ecosystem around our products and services that will be a force multiplier for Sage and for our customers and vendors. So we're very excited about this. And I would like to actually show you what this Inbox is right now. I'll take you through a double. But just remember, this happened in a record time. In the old days, this would have taken us 18 months probably to come up to market. Right now, we bring this up, and we're going to release the first version of this coming up soon for Sage 50 in the U.K. in the next few weeks. But this took us less than 6 months from a -- from 0 to bring it to this time, and we're going to accelerate this, introduction of this kind of product lines. So let's go ahead and... [Presentation]

I don't know about you, but I get very excited about product demos and just my nature. But what you've seen here is an application that sits on top of the digital network platform that has AI implemented in it that transforms the workflow. That's what we're all about. And now, Dan will take you through more product information. Dan?

Great. Thank you, Walid. So as we jump into building on what Walid was just talking about, we have a broad range of solutions that are designed to serve customers from small businesses up through midsized businesses and bespoke development solutions. Their anchor ground finance and accounting, that's where Sage's DNA has been. And so we've built a line of solutions to solve for customers across all of these sizes of businesses. We've also built out a set of solutions that expand the jobs to be done for our customers. We solve for finance, payroll, for AP and AR automation, for financial planning, BI and analytics and industry-specific operations capabilities that you might need if you're a retailer or if you're in construction, where you have that operational need as well. As we talk about specific verticals, we know that as our -- as businesses get larger, they have more specific needs where they want to be able to see that the solutions are designed around them. And I'm going to talk a little bit more about that as we get into some of the detail here. From this broad range of solutions, we want to bring together how we package that as we go to market. There's a number of offerings that we can provide to customers. We want to be able to tell a simple story to them so we can onboard them quickly into a set of solutions that are going to solve for their business need, whether they're a midsized business, a small business or it's an accounting firm that's helping their clients, we can bring this together a set of solutions that have an integrated experience that is designed around the users and they're going to cross workflows. We can solve for that broader range of solutions, all within one solution or one set of solutions that's easy to adopt, has one price point associated with it. And we can create greater value and create a longer-term relationship with those customers. So let's take a couple of examples of that. As we look at the midsize business, that office of the CFO becomes paramount, being able to solve for the breadth of jobs that they have in that office, being able to solve for planning, their operational needs, financial needs, HR and payroll capabilities and then bringing together industry-leading analytics that are built directly into that ecosystem. Being able to provide AI-embedded capabilities, as Walid was talking about, being able to build that into that ecosystem so that it's designed for that kind of -- the solution that they need as a midsized business. It's really designed around their workflow, so they have that full integrated set of solutions. For our smaller customers, it's really about ease and ease of adoption, being able to bring together an integrated experience where they can start with some simple jobs to be done, start with the accounting workflow. And then being able to add as that business grows, being able to add payroll or HR capabilities, maybe time management or expense management and be able to choose the right set of solutions that's going to work with them as they grow and expand but also within one kind of package solutions designed around that particular kind of business. And even before they're really thinking about accounting needs, they have a -- there's -- maybe they're a sole trader in the U.K. or a sole proprietor here in the U.S. They want to be able to bring to -- they have the accounting need, but they don't -- they're not quite thinking about it yet. Being able to work with a bank where they already have a relationship and see those products embedded, those capabilities embedded into their workflow where they can issue invoices, manage their procurement workflow, being able to manage basic accounting or work with their accountant where those are actually handled behind the scenes by Sage with third-party partners like our partner Tide, who's doing this for us as one of the thought-leading banks in the U.K. And then lastly, building a set of solutions that are designed around that accountant workflow, helping the accountants build relationships with their clients, from client management, at the very beginning through creating proposals for their clients, being able to manage the accounting workflow and capture transactions, being able to then -- be able to forecast and look to the future for where their clients are going to go so they can accomplish the best possible business results. The accountants can have a more fulfilling and profitable practice as they use the set of tools that are designed for them. But it's not just for the Sage solutions, it's also for the -- we're also able to embed whatever the financial system is that the client is using with that accounting firm. So if they are a QuickBooks or a Zero customer, they can embed that into our client management, giving us a real advantage in being able to serve those particular kind of firms. The second thing I wanted to dive into a little more deeply, Walid talked about what we're doing as we accelerate internationally, we have a number of global solutions. What we're doing in -- with Sage HR, what we're doing with our accountant solutions, what we're doing with our distribution and manufacturing solutions, but I want to concentrate on 2 specifics. I want to concentrate on Sage Active and Sage Intacct. For Sage Active, Sage Active is a relatively new product that's designed by our EU teams for EU clients. It's now available in Germany, Spain and France and it's designed for the small business. The second product I want to talk about is Intacct. Intacct is now sold out of 6 regions: the U.S., the U.K., Canada, South Africa, Australia and France. But it's actually broader than that. Intacct is -- has expanded out of South Africa into the rest of Southern Africa. It's expanded from Australia into New Zealand and this expanding into other regions -- other countries within the EU. It's already used by customers in 78 countries for tax compliance. And we have companies that we serve that have business entities that are managed in 130 different global currencies. Basically, it's a global solution designed to solve for global customers. So a little bit more about these 2 products. First about Sage Active. As I said, this is a product that is -- it's very DNA is EU-based. We brought together our EU product teams to build a solution designed for EU customers, where we have local expertise that's built into the solution. It's designed for small businesses that are growing and have the need for flexibility in managing their business process where they can adapt how they're handling purchasing or their invoicing process and grow over time, being able to use the open and API to be able to embed third-party solutions. From the very beginning, it was built to be multi-legislative for compliance and handle multiple languages. It's built on the Sage network that Walid was talking about earlier. So that has enabled us to move significantly faster where we can leverage solutions that are already in place to be able to get to market faster, and the pace is impressive, with new releases basically every single month. For Intacct. Intacct has been able to leverage the global expertise of Sage, being able to collaborate with our colleagues that are in every part of the world. I remember some really fun meetings where we're working together with our French colleagues about how we make sure that the product is designed with that local expertise. Really, how does the accounting workflow work in France? How is it different? How do we make sure that it reflects the experience that we want our French colleagues or our U.K. -- or our French customers or our U.K. customers to be able to have? We've built that natively into the product. We've also taken advantage of Intacct's global capabilities, things like multi-entity or multi-book accounting, which are native to the product and are available to be able to help customers be able to manage their global businesses. All of these products -- these 2 products -- all these global products are really helping us accelerate our international presence. The third thing I want to talk about is how we're able to focus on the vertical needs of our customers. It's often not just vertical. It's also micro vertical. So we -- the customers in the mid-market space want to know that we can come alongside them as a partner and provide them the expertise and the product that's right for them. So the differences might be, you have a nonprofit that's a community development nonprofit. As a community development nonprofit, they're issuing grants. And so they need to manage those grants. You might have a different kind of nonprofit, maybe a faith-based nonprofit that is managing restricted funds. And that's a key part of their accounting workflow. Those differences are things that we know well, and we can make sure that the product is designed around them with the platform that we have available to us today. Where Intacct has had historic strength in service-based verticals. Sage's has historic strength in construction and also in our product-based industries. And we've been able to bring that expertise together. And here's a little more detail on the specifics of how we brought that together. So first, for service-centric organizations. There have been a number of recent announcements that we've made to really accelerate the workflow and the automation that we can provide to customers. There's just a couple of the examples of things that we've done for service-centric businesses. So for -- an example might be what we've done with our advanced ownership consolidations. If you're a hotel chain or you're a family office, you have -- potentially have many different owners, a complex ownership structure for that company. And you want to be able to provide clear financial reports with partial ownership on a timely basis. Well, historically, the way that's been done is hours in Excel, potentially days in Excel to create those financial reports. Now we can do that in a matter of minutes with just a few clicks because of the kind of capabilities that we've built in this new capability that's been released. Another example is things that we've done to bring together the front office for education, Catholic diocese and healthcare, bringing together our -- for our for profit -- for our public and just about to be public companies, the control processes that they need with our partnership that we have with PwC are just a handful of examples of the kind of automation that we can provide for our customers. As we look at construction, there's been a number of new things that we've been able to create for our customers, building on our history of bringing great solutions like 300 construction and real estate and 100 contractor to market. There's a line of new cloud solutions that have been -- that bring together preconstruction and estimating, enhancing our construction financials and project analytics, as well as a new solution that's become part of the ecosystem for construction management that has become part of that portfolio of solutions designed for the broader range of needs of the construction business. We're building this out over the next quarters to be available. So there's an easy way to adopt it for those smaller construction firms so they can get started quickly and then build with that solution over time. And then lastly, I want to talk about what we're doing in distribution and manufacturing, built on Sage's deep expertise with X3 and our 200 line. In less than 18 months, we've been able to launch this brand-new solution into 7 countries. This new solution is designed around the Intacct experience and being able to build out a full set of solutions for a cloud-based distributor or manufacturer, but we've also integrated it with our 200 line of solutions. So those customers can move their operations capabilities into the cloud while retaining their current general ledger. There's new monthly deliveries of these capabilities. And so these are just 3 vertical areas. There are many more that we're investing in to accelerate growth from a vertical perspective. We want to -- as we're doing these things, as we're innovating across these spaces, one of the things we want to be able to do is come alongside our customers as they evolve. One of the things that happens for a business as they are succeeding, as they're thriving as we're wanting to make them do, we're wanting to make sure that they can flex, they can change over time. So just an example would be for a small business, being able to come alongside a small business and start with some simple capabilities, maybe the accounting. But then within the product through digital delivery, be able to turn on new additional capabilities. Turn on -- maybe they added an employee and they want to turn on payroll. Maybe they have a handful of employees that they now need to manage time for. Being able to turn on basic HR capabilities when they need them. And also coming alongside them with content. Content that through our membership capabilities that allow them to be able to learn what does it mean to be a thriving business. What does it look like to change their business process? How can they improve their -- the way they do collections. This content is available to them to help them be able to scale and grow. And for our midsized customers, we're looking to be able to come alongside them with scalability capabilities. With the Intacct product, we've introduced our next generation of platform scalability. So that's based on some very easy to calculate API tiers. And because we have simplified this, it's actually pretty thought-leading because a lot of the models that are out there require some complex calculations to know what you're using. We want to make this very simple for customers to know what are they going to need to have to be able to scale with their business and that they know where they stand with the platform. And then I want to talk also about what we're doing to come alongside our customers as they evolve. We have been able to achieve world-class retention rates because we don't make our customers change. We want to help them evolve as their business changes. A great example of this is the loyalty program that we've created for our construction customers. The loyalty program brings Intacct to our Sage 300 CRE customers as part of their subscription so that they can have that as a part of their program. And then for other customers and maybe in our 200 line, being able to have access to Sage Intacct planning so they have the collaborative cloud planning and budgeting capabilities available to them. With all of this, we are simplifying the life for our finance professionals, increasing the value that we're providing to our customers with a broader range of solutions while we simplify how they consume them. And deepening that customer relationship. We're accelerating internationally with global solutions that have local expertise built into them. And we're expanding our vertical solutions solving for more problems with customers, helping them with their pain points, which increase -- while we're doing the -- it increases the value of the contract and it allows us to address more of the market. With that, I'm going to hand it over to Aaron.

Thanks, Dan. I'm going to spend my time with you talking about our AI strategy. I'm going to give you a number of examples of how we've embedded AI experiences within our products. And I'm going to take you through where we're going with AI in the future. But to start with, I want to give you an insight into the North Star that drives all of our innovation with our AI investments. And it starts with the age-old problem with the accounting industry that essentially, it's bogged down in things like the monthly close, annual inventory counts, quarterly tax filings, annual audits. And the problem with all of these cycles is that they only provide point-in-time visibility, point-in-time confidence, and it's often weeks or months after the fact when the information is already stale. Our vision is really straightforward. We want to transform the industry into an industry that provides continuous, strategic value to businesses. And the way we do that is really straightforward. We've broken it down to 3 areas of innovation. The first is continuous accounting, where we've got this really provocative ambition to eliminate the monthly close. We want to pair that with continuous assurance, so that not only is your data real time, but it's also trusted. Then lastly, as we freed up accounting teams from the repetitive, low-value work of doing the mundane accounting tasks, getting through audits and now giving them real-time trusted data, we want to use technology to help them to make better decisions and to be better strategic advisers to businesses. And of course, as we've been telling you consistently today, the way we do that is through the Sage Network. So the Sage Network is our platform. It provides the map of our customers across all of their business ecosystems to form these connections. Those connections become data pipelines that we use to train AI to develop new ML capabilities. The platform also provides the developer toolings to make this an efficient experience for us. Our core AI team builds services that are commonly used across products to meet common customer needs, like automating data entry, doing classification of accounting transactions, looking for anomalous transactions. And while that team is building the core services, the product teams then take those services, embed them within their products and design an experience that's appropriate for their market, for their customer, including pricing and packaging. So a really good example of how this works across our portfolio of products is accounts payable automation. Every company in the world has accounts payable. They've got invoices to process. They've got vendors to collaborate with. And it is often reviewed by customers as the #1 area where they need to provide some automation and gain efficiency within the accounting team. So we've built a collection of AI services that enable that automation. We have AI for reading data off of the invoices. We have AI for classifying into the correct accounting treatment for those invoices. AI for doing the matching of the invoices to purchase orders and other documents where necessary. We've made it very easy to get invoices into the system, whether it's by email or a mobile app or APIs. And one of our objectives throughout consistently has been make it very, very easy for customers to adopt and enable and to make it global from the beginning so that we can support customers across the network. So if you just take a look at invoice processing within Sage Intacct, fully embedded. We've already brought it to market in the major markets where we sell Intacct today, and it's a single click to turn this on. We've got lots and lots of quotes from customers talking about the efficiencies that they've gained. I think this one very succinctly sums it up. "Our customers are doubling roughly, or in some cases, tripling their productivity in the accounts payable team." But to then show you how this work, we've brought the same capability to Sage 50 in France, to Sage Accounting in the U.K. and Canada, to Sage Individual in the U.K., to Sage AutoEntry. And in each of these markets and each of these types of customers, we've had to create the right experience and meet local market requirements. So for example, in Sage 50 France, we've had to prepare our customers for the coming e-invoicing legislation. For Sage AutoEntry, there's a total focus on the workflows of accounting firms and the way they interact with their clients. Sage Individual recognizes that individuals want to just use a mobile app to get their invoices into the system and that they've got a lot of anxiety about making tax digital. And so therefore, they need technology to help them to automate the classification. And we have to think about the entire end-to-end workflow when we build this automation. There's many steps across the workflow, but the last step in the workflow is always bank reconciliation. So here again, we've used AI to do the matching of payments and receipts from vendors and customers. We've got AI that for smaller businesses can automatically classify transactions. And of course, we use AI to do complete statement reconciliation. And what's really, really key about this, and again, a great proof for our strategy here is we now have 34 products across 11 countries that have integrated the bank reconciliation service into the products. So I just want to summarize a few points here on what makes the AI better because it's built on the Sage Network. So the first is that every customer benefits from that global reach. Businesses are increasingly operating internationally. An Intacct customer in the U.S. can process a French-Canadian invoice now because of the reach of the network, as an example. The volume of transactions that flow through the network, together with our relationship with our customers, enables us to amass incredible training data sets. So for processing invoices, we now have more than 10 million invoices that we are training the models on a daily basis. And as the models continuously get trained by more activity, every customer benefits with increased accuracy. One of the proprietary inventions from Sage is building models that are tuned across customers that are using the same vendor. So very commonly used vendors, these models are leading to nearly 97% accuracy and reading the data off them. And we have another unique advantage, which is that we're building these capabilities into our accounting products, which means that not only do we have access to the invoice data, but we also have access to the accounting data, the general ledger data, which enables us to build a broader reach of capabilities than if we were restricted to just accounts payable. So our ambition is really, really clear. This is something Walid says almost every day, every meeting, we will have AI in every product. And there's a couple of reasons why we are very confident that we can do that in a way that customers really benefit from. The first is we've built a world-class AI team. If you look at our team of machine learning engineers and data scientists, we now have 18 PhDs across those 2 teams. Collectively, they filed 17 patents in the area of artificial intelligence, 6 of those in generative AI alone. And while they've also -- they've built incredible services, they've also built an incredible platform for doing the machine learning work itself. I'm going to talk more about that here in a minute. So we've got a world-class team. We're also very targeted and precise in the way we invest in AI research. So I've already told you about our North Star, continuous accounting, assurance and insights, everything we do is focused on furthering that ambition. And rather than invest in big, risky foundation models, we instead focus on narrow fit-for-purpose models that are targeted at the specific problems we need to solve for our customers. So when we give our engineers flexibility to choose the appropriate foundation model, we pair that with our purpose-built models. The results are really, really impressive. So going back to the invoice processing example, we start with a common off-the-shelf foundation model, which has pretty good accuracy. But when you combine that with our models that are fit-for-purpose for the accounting industry, we're able to eliminate more than 50% of the errors that arise from the foundation models. And lastly, we've got a track record. We've been doing this now for -- since -- we've had AI in products since 2018. We've had the team now for 5 years. We now have more than 7,000 models deployed in production that are continuously trained, continuously redeployed as they improve. And collectively, they're making more than 20 million ML predictions every single day. And so it's that world-class machine learning infrastructure that really powers all of this. It's got the ability to fully automate the training and the deployment of these models. It's designed to support both global models. An example of a global model is reading data off of an invoice where everybody benefits from the shared experience, but it also supports customer-specific models where the models are tuned for an individual customer and their individual needs. So the example there would be the actual coding of that invoice according to how that business likes to evaluate their performance. But it's not just the technology itself. It also depends on the relationships that we have with our customers, where they're willing to sign agreements that allow us to do these developments, where they're incentivized to participate in our development programs. And just a couple of examples of how this turns into real customer benefit for our customers who are using invoice processing, once they start to use a vendor that is used commonly across the network, we reach nearly 97% accuracy. And an example of a customer-specific model with classifying those invoices, after just 1 month of use, that customer-specific model is 15% more accurately -- more accurate from when it started. So benefits across the global spectrum as well as individually for customers. So getting back to that North Star, the first was automation. The second is trust. Real-time data is great, but we want you to be able to trust that real-time data to make decisions. And one of the investments that we made very early on as we started the team was our outlier detection capability. It's very, very straightforward. This technology learns the typical patterns of business within individual businesses. And it's then able to identify when a transaction is anomalous, when it falls outside the typical patterns of the business. So it could be an accounting error. It could be a clerical mistake. It could be fraud. It could be a disruptive change to the business operations. And so our AI is able to call out for a human exactly what within a transaction might need additional review to take a look, maybe to change it. And if there is a change made, of course, that change gets fed back into the machine learning that retrains the model to understand this. This is a great example of the power of that platform, where we're now doing 15 million transactions reviews per week. We're catching about 2,000 problems for our customers, which might result in overstating revenue or filing a compliance report with bad information. And because of that partnership with our customers, we've been able to develop a training data set of more than 300 million journal entries. And this is the quote that I -- I use this quote a lot. It's my favorite quote in all my years working in accounting. "Actually having the -- giving the confidence to the accounting team that the rest of the business trusts them to give them good information is priceless." A couple of more examples. Sage Intelligent Time, where we're using AI to automate the capture and the categorization of billable hours for services organizations. The results are incredibly impressive. We've seen some businesses increase their billable hours by nearly 17%. We've got incredible accuracy. And you can see in this quote, "Sometimes automation drives confidence." We're also using AI to power Sage Earth, our carbon accounting solution. We're using AI here to take all of the expenses and the purchases within a business and classify them according to environmental categories so that we can accurately predict their carbon impact. And this is a really staggering statistic here about how we're able to do that. We've been able to pull together a training data set of 870 million transactions over 7 years with a total economic value of $5 trillion. It is a massive amount of data that we have available to train these models. And of course, everything depends on the trust that we have with our customers. Without their willingness to sign these agreements that allow us to do the development, without their willingness to actually use these capabilities, all of this is for not. So we've had to be very, very clear over the years about this focus on trust. We have a dedicated trust team. We've got clear principles that say, without perfect confidence or nearly perfect confidence, we don't release products. I mentioned we filed 6 patents in the area of generative AI, 5 of those patents have been solely around the safe use of generative AI. And we're also very transparent with our customers about the principles that drive our behavior and that define the processes within our business. So moving on to generative AI. Obviously, this is an incredibly exciting time. I took you through the North Star of transforming the industry into an industry of strategic value. With generative AI, our opportunity now is to wrap all of that in a more human experience. So despite the tongue-in-cheek illustration here, what we're really trying to convey is that generative AI both accelerates the strategy, but it also gives us a way to create experiences where humans can do better work than they were doing before. So I'll give you an example that ties into our continuous insights. If you're a small business owner using Sage Accounting, you don't have a choice but to file compliance-related reports. You have to get an income statement out of the product to a bank, for example. But small business owners don't often have the training or the education to really understand the insights that are found in these reports. So we've been able to use large language models to analyze the information in these reports and then to present it back to the business owner in a way where the insights are called out, where a summary of what you might learn from the report is written in plain English. And what I'd like to do now is take you through the vision for where we really think this can go, and this is ongoing investment in Sage right now, and that's the Sage Assistant. So we think it's great to build AI into our products. But we've got a principle of meeting our customers where they work. And if you're a business owner or an accountant, you work in Excel, right? It is the most commonly used tool in the world. So we've embedded Sage Assistant within Excel so that a user can ask questions related to their accounting data or they can ask for business advice or ask for some tasks to be completed that are related to an accounting workflow. So for example, a small business owner might ask Sage Assistant, "What's my forecasted cash balance at the end of the month?" We can then use the Assistant to interact with the accounting data, use our ML models that forecast out time series financial data. And because it's in Excel, we can actually bring in the evidence to support the forecast, and we can visualize it in a way that can be used with other decision-makers. And so for example, they can integrate the results into Microsoft project. Well, what if that cash balance gives that small business owners some discomfort and they want to increase the cash balance? Well, here, we can give advice. In this case, they can actually accelerate their collections of receipts by about $50,000 by reminding a few customers of the early payment discount. We can then use our AI that predicts which of these customers are likely to pay early based on those discounts. And because we're talking about accounting workflows here, we can go another step and ask the AI to write the email. They will actually be sent to those customers, reminding them to pay early or suggesting they pay early and via our integration with Sage Inbox, deliver that email and track the correspondence with the customer as they make those payments. The most difficult question for many businesses is when do I invest more in sales and marketing? And this is a really incredibly valuable answer to give to a customer. We take them through how do you measure efficiency? How is efficiency looking for sales and marketing in your business? And how is it trending over time? And we can look at the Sage Network to see, you know what, if you need to hire a sales rep, it's taking about 60 days on the network. So you should probably open your position in 30 days. So that's it for me. Thanks for joining, and I'm going to hand it back to Steve.

Thanks very much, Aaron, and thanks also to Walid and Dan. So a lot of content there, which hopefully you all found really helpful. But I just want to say a few words before we go to Q&A, just to emphasize what we've said a number of times over the last hour, which is, the network is the platform. And therefore, our ambition to be the trusted network for small and midsized businesses is crucial in terms of understanding why it is that Sage is differentiated. This platform is also the platform that enables us to connect all of our customers and deliver services, cloud services to all of our customers, but connect them to the people they need to be connected to, whether that be customers, whether that be suppliers or whether it be government. The future is a digital economy. And the Sage Network enables small, midsized businesses to use this platform to participate and connect into that economy. It is also the intersection of where connected accounting and artificial intelligence comes together. So thanks very much for bearing with us over the last kind of hour or so. And what we're going to do now is go to Q&A. And I think we're going to, James, start with the phones.

Many thanks, Steve. Yes, it's now time for the Q&A session. [Operator Instructions] It looks like the questions are coming in thick and fast, particularly on the chat. I think we've got a couple on the phone, so let's turn to the phones first. Operator, can we have our first question, please?

And now we're going to take our first question. And the question comes from the line of Toby Ogg from JPMorgan.

Yes. A couple from me. So first one, just for Aaron, just on the Sage Assistant you've referenced there at the end of your presentation, could you just give us a sense for where we are currently with respect to the development of this? What's a realistic time frame for when this could become generally available? And any kind of early thoughts on what potential commercialization of that could potentially look like? And then just second one for Walid. You made a comment earlier around leveraging AI internally to make some of your developers more effective. Could you just elaborate on that again in a bit more detail and give us some sort of indication around how much of an improvement in productivity you've been able to see? And also whether there are further opportunities to increase productivity within the engineering teams.

So the question about the Sage Assistant, we're actually quite far along in the development. Much of what I showed you in that demo is already developed and working. The next stage for us is to bring in some customers to do some -- to give us some feedback. What's going to be the biggest determinant in when we get this to customers will be our confidence that this is safe to use. And by safe to use, I mean you can't respond to an accounting question with an incorrect answer. We also have to think about how much do we bound the context within which the Assistant will operate. So we don't, for example, wanted to go asking questions that are too far afield from accounting. So it all comes down to, at this point, our level of confidence that we can deliver this to our customers safely. In terms of commercialization, it's very early days, not just for us, but for others. It's not entirely clear what model will work the best. But one thing that I think that's critically important is that this artificial intelligence is using the same APIs that developers have been using on our products for years. And so as long as we have pricing for the APIs, that's going to drive revenue for us when those APIs are used by artificial intelligence.

For the productivity for using AI internally, there's 2 dimensions to the question. Number 1 is actually using AI or generative AI for the rest of our colleagues. This is marketing sales and support. And we are in the process of deploying some of that right now to the organization. We're getting a lot of excitement. But in development, particularly, we've done a few pilots in the organization. We're seeing at a minimum right now, 10% productivity. And the way we measure productivity is the throughput of lines of code, the testing, the Q&A functionality. We're seeing also productivity at the early stage of careers. So younger people in career who are coming to the development organization, we're seeing them come up to speed significantly with that kind of AI, with the AI products that we have right now deployed. And we're seeing a lot of Q&A and debugging capabilities going in there. Overall, so far, at this stage, around 10% productivity is what we're seeing, and we think that this actually will increase in the future, the more these things mature.

Speakers, there are no further questions over the phone. I would like to hand back the call to James for any written questions.

Okay. Thanks very much. So we've had a number of questions come through on the web chat. They're quite interesting questions, actually. The first one, I think, has probably been addressed to an extent in the presentation, but perhaps you could elaborate on it a little bit more. What progress have you made in using customer data to create new artificial intelligence functionality to improve your products?

So I mean, quite a little -- quite a lot of progress. The video that Walid showed is with customers now. So that is deployed to production. And we're gathering feedback. Again, our -- the biggest determinant for us is will our customers trust the AI enough to use it. The demo that I showed you, again, we've used AI successfully to arbitrarily interact with customer data in a variety of our accounting products. It's able to actually understand our APIs and to invoke the APIs dynamically as needed, the embedding of that capability within productivity products is straightforward. I showed you Excel, but we've also embedded capabilities within Teams. Our expectation is that in the next couple of months, we will start customer testing. But I want to be very careful about setting any time lines for release because this is such a new area. And if there's one industry where trust is paramount, it's the accounting industry. So we can't do anything that's going to erode our customer trust and this technology.

Great. Thanks very much, Aaron. Another question on AI. Can you comment on the recent pricing commentary that's been given in the market generally on AI? How will you charge for AI features? Will you charge a premium?

Yes. So I think, look, as Aaron has already said, it's early days. We're seeing how things are going to evolve in terms of what we charge customers. And I think it's worth just differentiating between where AI is embedded within the products. And as both all of the team have said, we have AI kind of embedded everywhere throughout the products and throughout the network. And as opposed to, for example, an assistant that is giving you advice and giving you input. And I think we'll see how it evolves. There will undoubtedly be different tiers of offerings that we give to our customers. And particularly the generative AI, the Assistant is going to add a lot of value, but we're not going to sort of be too rigid in these early stages about exactly how we charge for that, whether it's completely separate or whether it's embedded into an overall solution that we then charge for.

Great. Thanks very much, Steve. They're still coming in on AI, specifically. So another one, how does AI and generative AI change the midterm road map, both in terms of how you're able to leverage it internally and also how it's able to drive growth?

So yes, I mean, I'll take that. It's really 2 things. It's acceleration. So accounts payable automation is a really good example. We're using more traditional AI to read data, to classify data. We're now seeing how generative AI can actually help us to automate the entire workflow with more autonomy than today and also to help customers to ease off on some of their controls. So in the one hand, it accelerates that North Star vision of driving automation. And the second, where you're going to most visually see this in the midterm, is these new interactive experiences where we expect there will be employees within a business that today don't log into an accounting system that will now actually be able to participate in an accounting workflow like approving a purchase order, submitting a time sheet using a digital assistant. And of course, we're very bullish on the idea that our users want to pull accounting information and do accounting analysis within Excel and other Office productivity products. So that's going to be the visual area of capabilities that you see in the midterm. And it's a key focus of investment right now.

I think this kind of ease of use, the way that business users will be able interact, super important. And I think maybe Walid, comment a little bit as well on why this is so important in terms of how we're able to deliver functionality to all of our customers regardless of what ledgers...

It's an incredible, actually -- this is actually a fantastic question in 1 sense, too. By the way, I'm going to make an announcement. In the next 18 months, almost -- not almost, every developer in Sage will be working on some sort of an AI capability. That's a transformation that we have begun on. But let me just address what Steve just commented on that. This is now the platform, and I used the word open platform in the essence of this platform is not only locked for Sage general ledger or Sage data only. We're seeing right now competitors and other vendors who will be able to interact and feed into our platform. That's very, very important. For example, we know there is a lot of small ISVs out there who could add a lot of value, but they don't have the resources or the resources or the money or the investment profile to be able to use AI. They will come into Sage Network, and they will be able to use our AI services as part of delivery for their end customers. And that's a really important thing. We will love all the customers, for example, Intuit customers and ISVs will be able to participate in our network. We think other competitors like they are Zero, they will be able to participate in our network. And that's how we are really, really creating this platform. It's open. It has a bunch of services that will be able to allow you to interact with them through an API layer. So it's a matter of calling. This is a standard layer of how to call that application. It will be available with no constraint. That means we're not going to limit it only to Sage data. We will be able to allow you to interact between our models and your data and take that information. And that's very, very force multiplier for us.

Great. Thanks very much. The web questions continue to come in thick and fast. So I'm staying on the topic of AI. You spoke about trust in the presentation. What permissions do you have to use the existing customer data on the platform? And does it matter where the data is stored?

Yes, it's a great question. And again, it gets to the point of why we're able to work with such large training data sets. And it's important to differentiate between data that's used in the process of developing a model versus data that's used in the training of a model. So those are 2 different use cases. In the first use case, we have customer programs where our customers agree to participate with us. Often, they're very interested in the solution that we're building. And via these programs, they then allow us to access their data for development purposes. Now, of course, we're still adhering to all of the laws and regulations around data privacy, rights of the data is anonymized, for example. But that agreement is what allows us to do the model development. The model training is essentially no different than how we host data for reporting or analytics. We have special purpose platforms through which this data flows to be developed. But fundamentally, it sits within the same constraints that we use for data that we throw into dashboards and analytics. But we're very explicit with our customers in the agreements about how we use data and our data principles also make it very clear how we make decisions regarding their data in the process of developing capabilities.

As for their sovereignty, we will follow local laws on data sovereignty. Full stop.

Great. Thanks, Aaron and Walid. Another 1, have you considered how artificial intelligence can assist with bank feeds?

Absolutely. I'm not going to announce any products today, stay tuned over the next few weeks and months. This is absolutely the richest area right now, that an AP and AR. This is a massively rich area to apply artificial intelligence and machine learning tool.

Yes. So for small and micro businesses, often the bank feed is the source of bookkeeping because they don't often do record invoices ahead of time. And so where we already have AI today in bank reconciliation is that classification of these transactions as they're pulled into the product. And we also have AI that does the matching of the invoices -- or the transactions and the bank feed to the underlying accounting record. But as Walid said, there's tremendous opportunity here to do more.

One thing is, I don't want to announce it that way. Invoicing in Europe and bank feeds are an incredible opportunity. So imagine that we are going to marry invoicing services with bank feeds and give the end customer and the accountant and the regulators and the government entities, one truly trusted ledger that is just -- it's 100% proof.

That's great. Thank you very much. Back to sources of data for training. One question that's come in is, is the data from cloud-connected products included in the retraining of AI models? Or is it limited to cloud-native data only?

Absolutely. I mean this is core to the strategy, right? One of the reasons why the Sage network is so important to us is that it enables us to bring this capability to both our native and our connected customers. And in fact, the prize for driving engagement with our connected customers is massive scale, right? We have a huge customer base on 50, on BMS. And to sort of drive the point home, the first product we embedded AI into for AP automation was Sage 50 France, which is a connected product. And so it is 1 network. It is one set of data pipelines. So now everybody benefits across the product line from the data flowing for both connected and native customers.

Let me just address the connected product discussion here. It's actually we are basically revitalizing all the connected products. They're really now becoming, in a lot of ways of their network products, they are all of them network products. So when we create banking feeds -- when we created the banking feed environment, we just created one set of services, and all these connected products kind of use that. This is the same structure we are going to use for all the services we provide. When I showed you the Inbox, the Inbox first release is going to be Sage 50 U.K., which is a connected product. And now we're bringing cloud technologies, AI technologies through the platform into all our franchise products.

Yes. I think one of the ways to think about this is in the very early stages, where cloud connected customers participate in the network. It started with a bit of -- it was like a one-way thing because what cloud-connected customers did was they connected to the network, and they pulled something down. So they pulled a bank feed down or they pulled down pay slips or whatever it was, they pulled information down. But actually, it was very -- there was -- it was quite limited in terms of how much 2-way flow there was, because very few cloud-connected customers would issue an invoice or do something which created a flow across the network. And that's one of the big changes here. What this is allowing us to do is create functionality, which we can then deliver to all of our customers. And using AI, as the guys have said, it means what you can do is you can sit in whatever your collaboration tool is, whether it be Sage Inbox or Teams or whatever it is. And you can sit there and you can make -- you can do things like approve invoices, you can approve holiday requests from the HR system, whatever. And when that needs to go back into the application, the machine will do it for you. So you as the user do not need to go and sign into multiple applications; HR, payroll, accounting because you can do your simple tasks, and then the machine will do the rest for you. And that then is accompanied by the Assistant and the AI-enabled functionality coming back to you with prompts and suggesting that you might want to do something else. Do you want a bit more information, when you're approving a holiday request, you want to see the person's calendar, and then the machine can go get it for you and show it to you. So there's a lot of applications which apply to all customers, whether they're native or connected.

Terrific. Thanks very much, Steve. A question about accuracy. So Sage is in an industry that requires 100% accuracy, really. And Aaron, you talked about 97% accuracy in the context of AI as being good. So I suppose the question is, is that 97% high enough?

It's all about the experience that you deliver to the customer. The only way to achieve 100% accuracy is to have humans as backup. That 100% accuracy in reading data off of forms doesn't exist. So what we have to do is create an experience where we don't project a 100% accurate view on that. So we show on the invoice when they're reviewing it, where we're confident that we've gotten the value correct and where we're not confident that we got the value correct. And that's where they can come in and fill in the missing bits that AI wasn't able to predict. And when the user makes that change, when they fill in the missing value, that actually gets fed back into the model, so it increasingly gets more and more accurate over time. 97% is very, very high. It's enough that our customers are very confident in it. We'll never achieve 100%, right? That remaining 3% investment needs to go towards the experience.

Let me give you -- just to elaborate on what Aaron said and the power of what he said in a second. A normal business in a medium-sized business, let's assume they get 100 invoices a day. Today, they have to process that 100 invoices, everyone in the invoice, individually. They have to process it by hand. With the AI models we created, that 100 invoices, 95%, 96%, 97% -- 97% of these invoices will just go automatically in there. And only they have to deal with the exception, which is the 3%, 4%, 5%. The amount of productivity that we gave back to the business, it translates into real money. Now you don't necessarily have to hire an extra AP clerk to actually process that. So that's the power of the 95%, 96%. However, our models in the user experience that Aaron showed you, it's not like we take all the 100% and just say, okay, you go figure out all the other 3% who did not work or 4% did not work. We have a really simple user interface that allows you to go and see the ones that did not actually make the bar and make your own connection on those.

I think it's worth just emphasizing that as we look forward over the coming years, today in most countries when people talk about invoicing and digital invoicing, they're talking about PDFs. So 80% of invoice flow today is emailed with PDF attachments. This is why e-invoicing is so exciting. If -- when e-voicing comes and it's already going to be mandated across the whole of Europe, it's just a question of timing as to when it's going to be implemented, and it's coming very soon in France. When we get through invoicing, then we'll start to have true digital interchange. And when we have true digital interchange, automatically, that's a lot more accurate than trying to read a PDF, which is effectively a paper document in a digital form. And so as the economy becomes truly digital and we start to get much more digital interchange with governments between customers, suppliers, et cetera, automatically, the power of AI and the power of machine learning will mean that as we automate those flows, they will become more and more accurate. And then as we always said, you elevate the role of humans. And what humans can do is look for those exceptions and make sure that those outliers are dealt with appropriately, but the machines help you identify the outliers in the first place. So you're not having to review every single transaction.

Thanks very much. So we're getting quite a few questions now coming in on the Sage Assistant. So I'll turn to those now. The first one , can you explain what is powering the Sage AI Assistant? Are you partnering with any vendors to deliver the underlying large language model?

So the answer is yes. We announced last year, a strategic partnership with Microsoft. We worked with them on the development of this capability where they made the back end large language model available to us via an API. In this case, we're using GPT-4. Obviously, as we go through and build more and more skills into this Assistant, there's actually quite a few models that come into play that are more purpose-built for specific scenarios. So for example, with what I showed you with Sage Accounting, where we're summarizing reports, we're actually using an anthropic model called Claude for that because that actually turns out to be -- for our needs, more useful and more efficient. So we will always use the right model for the requirements. But absolutely, our strategic partnerships have paid off in this regard in getting us early access and escalated support in building these capabilities faster and making them more reliable.

I think it's worth, I'm just emphasizing a point actually, that Aaron, you brought out earlier, which is I think sometimes people think that the Assistant, it's all about going and finding as much as possible, whereas actually, as you pointed out, and maybe just really emphasize this point. Actually, in some ways, our challenge is making sure that it's focused on looking at the right data set. And in some ways, maybe this isn't the right word, you tell me if it's the right word, we're almost slightly restricting our Assistant because we don't want it to kind of go too wide because we're trying to focus it on this, giving our customers quite precise advice as opposed to higher-level advice, which may or may not be correct advice.

I mean in the case of the Assistant, the large language model is understanding and interacting in the conversation, and it's making decisions about what task it needs to perform to answer the question. But when it actually comes to performing that task, we're using traditional technology. We're using the APIs that already exist that enable developers to query data from their accounting system. We're using APIs that exist to automate workflows. And so there's always this moment where there's a handoff between the nondeterministic technology, right, where it can be a bit unpredictable. And the very solid traditional technology that we've had for years and years that exists to be very accurate.

Fantastic. Thank you very much. We've got another one coming on the Assistant. Will the Sage Assistant be available on cloud native, or also on a broader range of products, including connected?

Okay. Yes. Look, as I said, every single product we have is connected to our Sage network platform. The Assistant will be part -- just like the Inbox, the Assistant will be part of all our -- it's part of the platform. Therefore, it's part of every single product that interacts with the platform. And we will make it available to ISVs third party through the platform. They have to come through the platform, and we're going to make the Assistant available to them to use as an API.

Thanks, Walid. Moving on to the Sage Network. While Sage products like Inbox are targeted to be vendor-agnostic, how easy or difficult is the integration with some of your competitor products?

Let me take this one. It's really easy. If you know how to call an API, which is almost every developer on the planet does, then you, by definition, know how to call Sage Network. Think of Sage Network as no different than, let's say, Azure or AWS, in that sense. It's a platform with a bunch of predefined services that we have in there, and we can -- we publish those services. And it allows you through a call. Now we have to have a commercial interaction with you before you make that call, you have to be -- you have to sign up an agreement with us. But other than that, from a technology point of view, it is going to be, Aaron, also brought something a comment that I want to emphasize. We are also as part of the Sage Digital -- Sage Network platform, we're actually providing also a developer experience. So we're working on a set of developer tooling that we will give to third-party developers and internal developers to make that experience even that easier, but you could use it today without having to have through that set of tools. You can use it through your set of tools.

Great. Another one on the network around trust. Part of the philosophy of the network is to enable trusted interactions between verified participants on the network. Have you considered embedding a capability to add digital signatures into your accounting systems to act as legal agreement between 2 parties?

Without -- the issue to that is we are in active research to go ahead and do that. That will be decided on a product by product. And then if we find a service that will satisfy the digital network requirements, then we can explore that.

Okay. Very good. Thank you, Walid. So moving a bit more generally now on to products. Could you provide a sense of the pricing for some of the new products you've talked about today, such as Sage distribution and manufacturing operations and also Sage Active?

So I'll take that. So the line of solutions that we're talking about today, Sage Intacct or positioning Sage Active for our smaller businesses is going to be in a, say, 1,000, 3,000, kind of Pound or Euro product. And then our manufacturing solution is a purpose-built solution. We haven't announced pricing. That's -- but it's designed as a module that's added on, and it will be competitively priced within for that particular ecosystem of solutions.

Great. Thanks, Dan. Quite a general question now. With regard to the general ledger. Can you explain how controlling the general ledger gives Sage a privileged position in SMB workflows?

Maybe I'll start, and then I'll hand over to Walid. I think, look, we talk about having an open network, which is very important because we want to create scale. So we want as many people to participate in our network, whether they be our direct customers with ledgers, with HR capability, payroll, et cetera, or whether they be third parties, as the team have explained, connecting through APIs. But it's also important to understand that the data and the transactions which are flowing through that network is not -- that's private, right? That's a protected set of data, which is important that our customers trust that it's a protected set of data. And that goes to the heart of the general ledger, because the general ledger is where a lot of that data sits. And so again, for our customers, it's important to them that they understand and can trust that, that data sits in a secure place. It is being used, but it is being used in an anonymized way for the machines and for artificial intelligence to learn to be able to produce better insights and indeed, ultimately, better products and better functionality for them. So it's important that customers know that there aren't anyone in the world with an artificial assistant can't just come and start looking at all the different ledgers. So the ledgers remain important as a place where the kind of core prime data sits and whether that be one of our ledgers or whether it be a competitor ledger, it still remains an important place where data sits.

So that's a really important point to make that we talked about when we talk about trust. I'm going to make a very general statement. Every customer controls their general ledger and their information. This is not something that we could reach in and just get that information without your permission. So nothing here will allow us to actually get your data without your permission or access to the information. That's very important. But the importance, Sage is the broadest company when it comes to general ledger across the world. We are the #1 in terms -- as it comes between medium and small. Nobody covers actually that information and have the general ledger. That gives us a competitive advantage because that allows us to understand better what people are doing with these general ledgers and how actually we can automate and how we actually use AI to make these general ledgers more effective. So that's a very important key differentiator for us to have the general ledger that we have, the general ledgers that we have and the customers using of those general ledgers. So that is really important to distinguish from other competitors who either don't have general ledger access or they have a very small portion of the general ledger information. As it relates to other customer -- other general ledgers' interaction, this is when we want to change the game. Up to now, these are all islands, including us, we had an island of our general ledgers, everything. What we did with the digital network is we're opening those pipelines with allowing a general ledger from other competitors that we have to interact with us in a controlled manner, okay? And again, in a controlled manner, in a safe manner, in a trusted manner. But we will allow you to exchange invoices and exchange transactions between our general ledger and their general ledger without having to go through a significant human interaction. And that's the experience that we only could give in the marketplace.

Brilliant. Thank you very much. Another question on Sage Active. Does Sage Active provide an upgrade path to any of your existing products? And also, how is it positioned versus Sage Intacct as you roll both products out across Continental Europe?

Yes, that's a really good question, especially since we recently launched both Sage Active and Sage Intacct in France. So one of the -- the first question around upgrade path, yes, that is part of the design principle, as Walid was just talking about, having a set of general ledger capabilities for micro businesses up through the large end of the mid-market. Having us have that capability means we've thought through. So what does a customer need to do when they're ready to grow to that next level of capability. And so we've built program into place to be able to provide that migration path with as -- and we're going to continue to invest in how seamless that transition can be where they unlock new capabilities as they move up. That's the first part of the question. The second is, as we're launching product into new regions, we're making sure that we're solving for -- the particular customer that we're solving for. So one aspect of that is we've launched Active into France, it's really targeted at smaller businesses where -- there's a large number of business in that region also in Spain and in Germany that Active is really purpose-built for that type of smaller business customer where Intacct unlocks capabilities around multi-entity and some of the more complex use cases that they need to be able to have as those businesses continue to grow and thrive. And so we've built those capabilities with clear differentiation so we can guide the customer to the right product for them to start with and then they can grow with us over time.

Great. Thanks very much, Dan. More of a financial-related question now. How should we think of R&D as a percentage of sales in view of these new product developments and pipeline at Sage?

Well, I'll kick off, and then Walid, you can comment as appropriate. But I think we've talked a lot about the shape of the P&L. And obviously over recent years, we've increased the percentage of revenue that we spend on R&D. And we've been very clear that we don't see that we need to increase it any further. So the way that Aaron and Walid and Dan and the team are thinking about this is this -- what we're talking about is fully funded. So I don't want anybody to think we're going to suddenly start saying that we need to spend more as a percentage of revenue. And actually, if we weren't here live on camera, I'd be saying to Walid, we obviously need to see efficiencies. So maybe we need to speak a little bit less. But as financial investors, you don't need to worry that we're going to suddenly say we need a load more funding for R&D. We're very committed to doing it within the envelope that we've communicated.

3 things here. Number one, I am actually very happy with the shape of the P&L. That's actually really from a product person, usually, they would never say that, but especially in front of my boss. I am very happy with the shape. The second thing is, we're doing 2 things that are converging for us. Number one, the productivity that actually copilot and the AI has given our development organization, ability to actually be more productive. So we're using that productivity as the slack for us to actually produce more services and products. Second thing is, Sage has embarked on partnership deals and partnership strategy that will allow us to partner with Microsoft, with Amazon, with potentially other big partners to actually help us co-develop these things instead of us developing, for example, as Aaron described it, instead of us developing all these language models on our own, we're actually partnered with Microsoft, and we expect us to partner with other great suppliers like Amazon, like Google and so on and so forth, to actually partner with them to -- it's not the purpose of reducing the R&D spend. It's actually optimizing our R&D spend and getting products to market faster.

Thank you very much, Steve and Walid. This brings us just about to time. So I think with that we'll close. I'd like to say a big thanks to Steve, Walid, Aaron and Dan for this session. I hope you found it informative. Thank you very much for joining us. We look forward to seeing or speaking with you again soon. Goodbye.