Alfa Financial Software Holdings PLC (ALFA) Earnings Call Transcript & Summary

Hello, everyone, and welcome to Alfa Financial Software's 2026 Half Year Results Presentation. As always, I'm joined by Matthew White, Alfa's COO; and in his final results performance before retirement, Duncan Magrath, Alfa's CFO. And for the first time by Andrew Dickson, Alfa's incoming CFO. Duncan will take you through the financial review. Matt will cover operational delivery, and I'll return later to talk about the business, our sales progress and our outlook before we summarize and open for questions. So stepping back and looking at the first half as a whole, we are pleased with the progress we've made. Subscription revenue grew by 14%, subscription total contract value by 22%. ARR grew by 17%, and net revenue retention remained strong at 110%. Subscription revenues now account for 37% of total revenue, which reflects the continued evolution of Alfa towards a larger recurring revenue business. Sales activity has also remained encouraging. We secured 2 wins during the first half, and total contract value increased by 17% to GBP 247 million. And we continue to see good activity across both the late-stage and early-stage pipelines. We are already working with 3 of the 9 prospects in our late-stage pipeline, which gives us confidence in our future opportunities. We have continued to invest in our product and delivery capability. During the period, we invested GBP 19.6 million in software, particularly in originations fleet commercial finance and our AI capabilities. We also achieved 2 go-lives during the half, which is an important validation of both our product and delivery approach. Looking ahead, we remain confident in our full year expectations. We see artificial intelligence as an exciting opportunity to enhance our product, accelerate development and simplify delivery while continuing to create value for our customers. Turning to the key financial highlights. Revenue was GBP 65.1 million, representing growth of 5% at constant currency. Total contract value increased 17% to GBP 247 million, and ARR increased 17% to GBP 48.5 million. Subscription revenue grew 14%, while net revenue retention remained very strong at 110%. Operating profit was GBP 18.4 million, and operating margin was 28%. Excluding the impact of severance costs and FX hedges, operating profit was ahead of last year by 2%. And with operating margin only slightly down on last year. Duncan will cover this later. Cash conversion for the first half was 76%. That was influenced by the timing of customer receipts between the second half of last year and the first half of this year. Duncan will discuss that also in more detail shortly. Overall, these results reflect continued growth in our subscription business. ongoing investment in the product and good momentum across both the sales pipeline and delivery organization. Before handing over to Duncan for the financial review, as usual, I would like to welcome Andrew Dickson who recently joined us as CFO, and who will formally succeed Duncan on the Board on the 16th of September. Andrew, perhaps you'd like to introduce yourself.

Thank you very much. Since joining Alfa a couple of months ago, I've got to know people both in EMEA and the U.S. I've been hugely encouraged by what I've seen and I'm really looking forward to taking over from Duncan when he steps down the board in a couple of weeks' time.

Thanks, Andrew. The first half of 2026 was a solid financial performance against last year, which was always going to be a tough comparator. Revenue was up 4% at actual rates or 5% at constant currency with growth in subscription and delivery revenues, partially offset by a lower level of software engineering revenue than the very strong first half of 2025. The gross margin percentage was down 400 basis points. Two things to note here. Firstly, last year, we had a very high level of chargeable software engineering revenues. And secondly, the margin in this period is weighed down by severance costs. Operating profit was down 15%, delivering an operating margin of 28.3%, although very much impacted by severance costs and FX hedges. Excluding these, operating profit was actually ahead by 2% with the operating margin only slightly down on last year. The effective tax rate of 26.1% was in line with 2025. Turning now to the cost lines in a little more detail. Cost sales was up 16% and SG&A up 12%, but both lines are distorted by the severance costs and FX hedges. In the first half, we incurred GBP 1.6 million of severance costs, including the associated legal fees, and this was split GBP 1.2 million into cost of sales and GBP 0.4 million into SG&A. In relation to FX, we had a GBP 1.7 million gain on the U.S. dollar hedges in 2025 compared with GBP 0.3 million loss on the hedges in 2026. So the year-on-year swing on that line alone is GBP 2.0 million. Strip both of these out and the picture is much more measured. Cost of sales increased 10% and SG&A was up just 2%. Within SG&A, profit share is down GBP 0.5 million or 19% simply reflecting lower profits with share-based payments also being down. Depreciation and amortization was as expected, up GBP 0.6 million or 39% and I'll come back to the trajectory of amortization under modeling guidance. Other operating costs were up 6%, driven by the growth in head count and the cost of expanding into new markets and territories. Turning now to TCV. Total TCV grew 17% to GBP 247 million, up from GBP 211 million this time last year. Looking at the revenue streams, subscription TCV and delivery TCV were both up 22%, with software engineering down 29%. Software engineering reduction reflects a lower level of chargeable development work needed for new customers as the profile of new customers changes, so will TCV. Next 12 months TCV was up 12% to GBP 101 million from GBP 90 million last year, again, with growth in subscription and delivery, partially offset by the reduction in software engineering. As I flagged up last year, we now supplement the TCV disclosure with ARR and NRR and I will turn to these next. Annual recurring revenue was up 17% at 30th of June 2026 to GBP 48.5 million from GBP 41.6 million. As a reminder, we calculate this by taking the average subscription revenues over the last 6 months and annualizing them. And we exclude any revenues, which we do not expect to last 12 months at the point of origin. That 17% growth is very much in line with the underlying growth in subscription revenues. Net revenue retention was 110%. To help understand this, I've included a breakdown in the table. We are, in effect, a 0 churn business for Modern Alfa systems, and so I've started the table at base of 100%. We then have a 3% drag from 1 specific terminating V4 customer who gave notice back in 2018 and not yet fully transitioned away from Alfa. We grew 6% from net upsell across the existing customer base and grew 7% from customers not yet live. It is the impact of new customers not yet live that I will return to in a minute because I'm not sure that this future growth is sufficiently understood by everyone. So looking now at overall subscription revenues. Subscription revenues were up 14% on last year to GBP 24.1 million. Looking at the breakdown, 73% came from customers already live on V5 or Alfa System 6, 20% from new customers currently implementing ASV, 4% from customers upgrading from V4 and the remaining 3% from V4 customers who've not yet upgraded. Subscription TCV was up 22% on last year driven by growth from both new customer wins and the existing customer base which underpins our confidence that this revenue stream will continue to grow strongly. We now have 44 customers contributing to subscription revenues, up from 41. Within that, Alfa Cloud customers on V5 or AS6 have increased from 23 to 24, with a further 3 in the late-stage pipeline, up from 1 last year. We have 15 customers on private cloud and 2 remaining V4 customers. As I said before, we do not expect to convert every private cloud customer, but we would expect to convert many of these over time. and that remains a potential incremental source of growth for the next few years. Staying with subscription, I want to dive a bit more into the growth from customers not yet live on Alfa. I've shown this slide before, it shows the history of our V5 and AS6 customer base, and I've previously used it to make the point that there has been 0 competitive churn. I'm showing it again but wanted you to focus on the customers and implementation at the top. We currently have 15 customers in implementation, of which 2 are V4 upgrades. So there are 13 new customers who will drive future subscription growth once they go live and reach their full run rate revenues. All the customers in implementation started from 2023 onwards. I'll show on the next slide how the customers in implementation, who we started work with in 2023, 2024 and 2025 will strongly contribute to the future growth in subscription revenues. Many of you will be familiar with the illustrative graph I produced a few years ago, showing the typical life cycle of a contract. I have shown a small copy of this at the top of this slide, as a reminder, and repeated the full slide in the appendix. This time, rather than using illustrative figures, I will be demonstrating with actual and internal forecast data. The previous slide showed all V5 customers and was organized by when we started working with them. I have taken the 2023, 2024 and 2025 cohorts and excluded any of those, which are customers upgrading from V4, so only including new customers. The 2023 cohort is deep orange with 5 customers in it. The 2024 cohort is medium orange with 6 customers in it, and the 2025 cohort is the light orange with 2 customers in it. It shows revenue by half years with actual data from H1 2023 to and including H1 2026. It then shows internal forecast data for the period after that. You can see how the revenues build as customers move through implementation to go live. It is important to note that this only shows revenue from customers we started working with in those years, and so it is not total subscription revenues. It is intended to show how the new customers and implementation are a significant contributor to overall growth in subscription revenues. The overall percentage rate growth rates will be lower than this once you add existing live customers into the figures. You can see that the growth in revenues from the 2023 cohort starts to flatten off in 2027 and 2028 as those customers reach go live and hit their full run rate. If you compare the revenue for these 3 cohorts in 2026 to 2028, you will see that the revenue more than doubled over that period. These are revenues from customers we are working with today and there is no contribution here from customers we are not yet working with. We can be pretty certain and unless the project stop this growth will come through into our subscription renews. There is some risk over timing, so it takes longer to get projects to go live than we expect. The ramp-up in revenues will be later. But at the moment, this is our best view of the likely outcome for these cohorts. It is this growth in subscription revenues as customers we go live, which gives us confidence in the strength of subscription revenue growth over the medium term. Turning now to software engineering revenues. Software Engineering revenues were down 17% from a very strong first half last year, although 43% higher than the year before. Within this, chargeable development work for new subscription customers decreased by GBP 2.7 million, with development work for existing customers down GBP 0.5 million. Offsetting this, customized license revenue was up GBP 1.7 million to GBP 2.8 million, reflecting the completion of the accounting on our last significant perpetual license sale. We're now into the perpetual license tail at GBP 4.8 million on the balance sheet at the half year, which will be GBP 3.9 million by year-end. It will steadily unwind with GBP 1.7 million recognized in 2027, GBP 1.1 million in 2028, with the balance declining through to and finishing in 2031. There was no one-off license revenue recognized in H1 2026. TCV is down 29% on June 25, reflect that lower level of future charge or development work for new customers. I would note though that while TCV is down from December, our visibility of future work has improved since then. That work is not yet in TCV, and this is an area where we could do better than current expectations if client approvals come through more quickly. Turning to our final revenue stream delivery. Delivery revenues were up 5% year-on-year. 54% of delivery revenue relates to new customers in definition or implementation, up from 40% last year. which again speaks to the volume of new customers moving through the pipeline towards go-live. Partner days were 8% of our total delivery days, slightly higher than the 7% last year. TCV is up 22% on June 25, primarily driven by 2 new customer wins, and you can see the improvement in coverage across both the next 12 months but particularly in the period after 12 months. We are looking to recruit more people into delivery for 2027 as looking at TCV on our late-stage pipeline, we need to increase the people to deliver the growth we expect. We have 44 customers contributing to delivery revenues, up from 43. Within that, live V5 and AS6 customers have increased from 26 to 28, and we have 3 new customers not yet live, up from 11. Turning now to cash flow. Cash conversion was 76% for the half year. This is lower than our normal level, but as a result of the very high conversion of 108% in the second half of 2025, which included GBP 2.8 million of accelerated receipts in December. So this is a timing effect between the halves rather than any change in the underlying quality of our cash generation. I've included the cash flow performance by half in the appendix to demonstrate this. Our modeling guidance for cash flow remains unchanged. Capital expenditure remains in line with the prior year. Net tax payments increased to GBP 4.7 million. Last year benefited from cash received on R&D claims, and we've had none of that so far in 2026. We paid GBP 13.7 million dividends in the period made up of the GBP 4.5 million ordinary dividend and the GBP 9.2 million special declared with the full year results. Now some words on capital allocation. Alfa remains a strongly cash-generative business, and our approach to capital allocation remains disciplined. We continue to generate excess cash even after allowing for the investment we are making in the business. For the last 5 years, we've paid an ordinary dividend and then returned excess cash to shareholders through special dividends and buybacks. Having reviewed current market conditions, we've decided to retain the excess cash for the time being to provide us with optionality over how we use it. If in due course, we conclude that we have no use for it, we will return it to shareholders. Our policy of paying an ordinary progressive dividend remains unchanged, and the amount to be paid for 2026 will be announced with the full year results. Next, a brief update on modeling guidance. Starting with the outlook for 2026. We expect continued growth in subscription revenues. Delivery revenues will grow more slowly than we originally expected, but we expect that to be offset by improved Software Engineering revenues so that overall growth is broadly as we anticipated. Capitalized development costs are expected to continue at similar levels to 2025. As the capitalization of internally generated intangibles has grown and those assets come into use, amortization will increase to a similar level as capitalization. Cash conversion is expected to be 80% to 90% for 2026 as a whole, which is unchanged from what I said in March, notwithstanding the 76% in the first half. The effective tax rate is expected to be around 26%, slightly higher than the U.K. corporate tax rate, reflecting the overseas territories we now operate in. On currency, the sensitivity shown up for a full 12 months and unchanged from previously disclosed and ignore the impact of hedges. For profit, we are fully hedged on U.S. dollars. So any movement in the U.S. dollar exchange rate will have no impact on profit. We do not hedge account, and so revenue is unaffected by our hedges. And so a $0.01 movement in the average exchange rate for the second half would have half the full year effect i.e., GBP 250,000 impact on revenue. I will now hand over to Matt for an operational update.

Thank you, Duncan, and hello, everyone. I'm going to start, as I always do, with a reminder of our strategy. And as a reminder, it's not an update. Our strategy is stable and it's consistent, and we see that as a very good thing. But I think that the reminder is important because the most important thing to understand about Alfa is the context in which we operate, and that is, the market that we serve is extremely complex, highly regulated, infinitely demanding and ever-changing. Our opportunity is huge. We're the leading player in a massive market, and we currently have only a small market share. So our strategy for creating long-term sustainable business value is designed to maximize and enable us to grasp that opportunity. And that strategy is to strengthen to grow our differentiation by investing in our 3 key differentiators: our smart, diverse team, our product and our delivery methodology in tooling. Secondly, to sell to enable profitable growth by focusing on building our community of single-tenant SaaS customers, increasing our subscription revenue and enabling incremental sales. To scale, to increase our capacity for developing and delivering out the systems and to extend our reach. And finally, to simplify to enable more concurrent Alfa systems implementations more efficiently. And I'm going to focus in a little more on this final aspect of our strategy, simplification this morning. Reducing friction in the implementation of Alfa systems has been a key element of our strategy for a long time. Doing so will lower the cost of delivery, shorten the time before customers go live and allow subscription revenues to begin flowing sooner. Just as importantly, it improves the economics of projects for customers. We are and we have always been the premium provider in our industry. by reducing delivery costs, we will enable prospects that may previously have been too small or too cost sensitive to justify an Alfa implementation to afford an Alfa the premium offering. To repeat our market share is actually very small. So the amount to go after is huge. So we expect reduced cost per implementation to result in us being able to reach significantly more customers. AI provides fantastic tooling for increasing the efficiency of the implementation process. The technology is improving quickly and we're seeing real benefits. Data migration provides a great example use case. In 1 example, we've seen the effort required for development of data transformation code reduced by 75%. and AI provides a fantastic tool for reconciling the migration output as well. Our AI-enabled Alfa Recon tool is one of our biggest incremental sales opportunities. But importantly, most of the work carried out as part of an implementation of Alfa systems is done not by the Alfa team, but by customers or by implementation partners. The role of the Alfa team is generally to provide expert assistance in implementation tasks and our implementation partners are also investing in AI as a simplification technology and they're competing with each other to find the most impressive efficiency gains. So again, the reduce -- the result of this is reduced implementation costs and increased addressable market. Moving on to our product and our market. As I said when presenting our FY '25 results, we see AI amplifying the value of Alfa's product. Alfa systems value has never simply been the code base. It's the combination of decades of domain experience embedded within a SaaS platform built for one of the most complex vertical markets in enterprise software. We provide a governing control plane for the world's largest and most complex finance organizations. At its core, sits a robust ledger and system of record for auto, equipment and commercial finance, providing a vast, well-structured data framework. Around that are deterministic transaction processing configurable workflows, embedded authority models, security, resilience, integration capabilities, scalability, performance and extensibility, and the embedded and configurable workflows are standardized, they're auditable, they're repeatable, reversible and integrated. Now these aren't just technical features. They form a trusted operational platform for highly regulated businesses, and customers can rely on us to evolve with the pace of technological change, allowing them to focus on enabling investment in the economies that they serve rather than being distracted by fast-moving technology. And alongside all of that, there's Alfa clouds, SaaS, delivery and implementation track record. Those provide huge competitive advantage. And while we expect AI to enable efficiencies, including head count reductions for our customers, Alfa systems is priced based on the number of asset finance contracts managed on Alfa rather than per user. So our revenue model is not impacted by increasing customer efficiency. Next, a few words on Alfa systems AI functionality. Our Alfa systems AI products now live under one umbrella brand, Thea. Thea core is the layer within Alfa systems that allows Alfa innovations to communicate safely and efficiently with AI, regardless of the underlying service. Importantly, this is only available for Alfa Cloud customers, providing a compelling case for upgrade for customers not yet using our SaaS offering. An example of a quality of life feature built on Thea Core is via notes. This provides a summary of notepad entries for an agreement, which is hugely powerful for many of our customers. Thea Lens provides intelligent document processing functionality. Alfa can already work with third-party IDP solutions are now working on our own functionality in this area, powered by Thea Core. And Thea Connect provides a model context protocol or MCP server. This has not yet been launched as part of our marketing agenda, but it is available, for example, use cases. And MCP is an emerging standard for connecting AI assistance to business systems, and it's now firmly part of conversations with our customers and with prospects. Thea Connect is likely to be key in allowing customers to plug AI tooling directly into helper processes. Moving on from AI, and we've progressed with our key market expansion exercises in fleet, commercial finance and U.S. auto originations. All 3 are progressing with customers, which is our preferred methodology for investment. U.S. Auto ignitions is exciting because of the scale of the opportunity. Every U.S. auto finance provider requires originations functionality, and this is a new addition to our offering, the value is substantial. Fleet is exciting because it opens up the European auto finance market, where auto fleet management often sits alongside retail finance. Our first implementation of our fleet functionality is progressing well. And commercial finance is an adjacent market, which will, in time, increase our TAM and we're stepping up our marketing efforts within the commercial finance well. We're also investing in our portal for customer and dealer access, again, with customer partnerships. And we've completed a pilot accelerating software development using AI tooling, we've had some excellent results. We found many compelling use cases, and we're now moving into a BAU phase. We expect to have usage of this exciting tooling to continue to increase the pace at which we can deliver new features for customers. We continually assess the shape of the team required in order to deliver efficiently for customers. We've seen reduced demand for customer-led enhancement of our software, and we've reshaped the team in response. In the first half, this has resulted in 31 nonvoluntary departures from Alfa, mainly from product engineering. We've continued to recruit where demand is stronger, including into cloud hosting operations and into delivery with both graduate and experienced hire recruitment. We've refreshed our new high induction material, and we're now rolling this out in all regions. The new approach enables new hires to be onboarded more efficiently and more effectively. And our cross-company program of AI literacy ensures that all areas of the business have the resources that they need to maximize opportunities for efficiency. For example, every business area has AI champions as points of contact to roll out and adapt learning and development materials for their group. We have a fantastic team and a culture of delivery and of growth all focused on building this special company together. 2026 people initiatives include our culture playbook, ensuring that we maintain and grow our culture as we scale our team internationally. Our unrivaled track record of delivery continues, and it's this ongoing delivery, the layers new subscription revenues onto our model. In the first half, we achieved 2 go-lives of new Alfa System 6 customers. The first was for an existing Alfa V4 customer, which upgraded on to AS6. The go line involved migrating portfolios in 2 different countries onto a single segregated instance of Alfa Cloud. The upgrade allowed our customers to simplify their internal systems infrastructure as well as to access the product benefits of the latest version of Alfa. The second new customer go live was for a limited new business payment, but with a ramp-up in new business volumes expected over the coming months and with migrations of the existing finance book expected to follow. In due course, the result will be our largest Alfa cloud implementation. So we expect this customer to be an important part of our growth in the coming years. We've also sold a new subscription upgrade product to our first customer. This is a win for our customer as well as for Alfa for our customer access to upgrades on a subscription basis makes costs predictable and upgrades easier to access. For Alfa, the new model increases subscription revenue and assigns to us the benefits of increasing efficiency. And we're confident of further sales of this product in the future. So we have the leading product, an outstanding team and a clear track record of delivery in a complex vertical where competitors frequently fail. Our market opportunity is huge. Our simplification agenda accelerated by advances in AI tooling will enable us to reach more customers and layer high-value recurring subscription revenue onto our model more efficiently and investment in our product is expanding our opportunity. So we're really excited about the future. And I'll hand over to Andrew Denton for an update on the prospects for future customers.

Thanks, Matt. I'll continue with the business and sales update. We're pleased with the progress we've made in the pipeline since our full year results. During the first half, we converted 2 prospects into wins, demonstrating the continued demand for Alfa systems, and the effectiveness of our sales strategy. At the same time, we maintained a strong late-stage pipeline of 9 prospects spanning multiple geographies, customer types and industry segments. Importantly, several opportunities have continued to advance through workshops and contracting activities. And we are already undertaking paid work with some of these prospects. This remains a key indicator of commitment and provides a strong foundation for future conversions. The pipeline is geographically diverse across the Americas, Europe, the U.K. and broader international markets. We continue to see particular interest in our investments in originations, fleet and commercial finance, reinforcing our belief that expanding the product capabilities increases our addressable and serviceable markets and strengthens Alfa's competitive position. And we've been particularly pleased with the level of incremental sales driven by our new commercial finance modules. So looking forward, demand for assets and automotive finance remain -- software remains strong. What continues to differentiate Alfa is the combination of our people our product and our delivery track record. These advantages have underpinned our success to date and remain central to our long-term strategy. Artificial intelligence is creating exciting opportunities across our business. As Matthew explained, we are using AI to accelerate software development, simplify implementations, improve internal efficiency and create practical functionality for our customers with entrusted Alfa workflows. We are particularly pleased with the progress of our Thea AI product group. We continue to invest in market expansion through originations, fleet and commercial finance, and we believe these investments will continue to support future growth in both delivery and subscription revenues. While foreign exchange remains a headwind given the success of our North American business, our expectations for the full year remain unchanged, and we continue to see a significant opportunity ahead of us. So to summarize, the first half of 2026 has seen continued progress across the business. Subscription revenues grew 14% and subscription TCV increased 22%. ARR grew 17% and net revenue retention remained strong at 110%. We Subscription revenues now represent 37% of total revenue, demonstrating the ongoing transition of Alfa towards a business with a larger recurring revenue base. Sales performance was encouraging, with 2 new wins, growth in total contract value to GBP 247 million and a healthy late-stage pipeline of 9 prospects. And we continue to see strong interest across the markets we serve and good activity in the earlier stages of our pipeline. We continue to invest in our product, in our people and in delivery capability. Investment in originations, fleet, commercial finance and AI is expanding our addressable market and strengthening our competitive differentiation while AI is helping us simplify delivery and accelerate element. Most importantly, we remain confident in our future prospects. The combination of a growing subscription base, a strong pipeline, expanding market opportunity and continued product innovation positions Alfa well for the remainder of 2026, far beyond. Thank you for listening.

Thank you for the presentation. We have had a number of questions presubmitted and submitted live. [Operator Instructions] We're now moving on to our first question. Revenue only up 4%. That's a big step down from where we used to see Alfa. Is this the new norm? Or does H2 pick back up?

Okay. It's Duncan Magrath, I'll take this one. I think the best way of thinking about this is to think of the 3 different components of our business or the 3 revenue streams. If you think about delivery business being roughly 50% of the business, if you think of subscription being 35% to 40% of the business and Software Engineering being 10% to 15% of the business, if you've got those blocks in your mind. And we've got very different growth characteristics of each of those blocks. So if you look at the first half, if you look at those blocks, as we've talked about, subscription was up delivery was up 5% and Software Engineering down 17%. So the 4% is very much a combination of those 3 different parts of the business. So if I deal with it in 2 bits, what is -- does the second half pick back up. If you -- in terms of the profile of each of those pieces of business, I'm expecting subscription revenue to be slightly stronger growth in the second half and delivery in Software Engineering to be similar. So we should see growth being slightly higher than the first half, but not dramatically higher. So overall, for the year as a whole, the analysts have us on about 6% growth for the second half, that gives us to sort of nearly 5% growth for the year as a whole. The question also says is this a new norm? And I think if you, again, look at those blocks of the business, again, it's not the biggest one, delivery, half the business at the moment. We still expect delivery typically to be somewhere between mid-single-digit to high single-digit growth. So let's call it 5% to 8% type growth per annum is not a bad sort of medium-term indicator for that business. And if that's driving half the business, then you've got sort of 2 to 4 percentage points of growth for the whole company coming from that part of the business. Subscription, 35% to 40% of the business, mid- to high-teens growth is where we would be expecting that business. And so you can see that we would be getting 5 to 6 percentage points growth from subscription for the whole of the business. So before Software Engineering, we're about anywhere between sort of 6% to 8% -- sorry, an 8% to 10% growth business. And then Software Engineering is the one that's much more difficult to predict because it really much depends on what's coming through the pipeline, the nature of the customers. et cetera. So I think no, it's not the new normal. I think if you were modeling something I've said it before, modeling Software Engineering as a sort of flat part of the business wouldn't be a bad place to start. It will be sometimes much better than that, and sometimes it will be below that, which we've seen this year. So not a new normal. And I think the important part is strategically, we're trying to grow the subscription part of the business as a proportion that will grow as a whole and the growth in that part of the business is the fastest-growing part of the business.

Thank you. Well, we have 2 now for TCV. So the first question, TCV up 17%, but revenue barely moving. What's the lag time on all the contracted work actually converting? And then the second question is how confident are you that the current TCV growth will translate into sustainable growth earnings.

Yes, I'll take that one as well. And again, I think if you think of those individual building blocks, it's important to think about it that way again. So TCV was up 17% year-on-year. A lot of that growth came from subscription. So GBP 31 million of the overall GBP 36 million of growth in TCV came from subscription, as I said, the fastest-growing part of our business. GBP 11 million of growth came from delivery and actually Software Engineering was a reduction of GBP 6 million. So it's the same picture as revenue growth that I just walked through. So in terms of the lag, it's very much where we'll see the growth percentages that I just talked about coming through. And it's more the makeup of the TCV and how that plays out is the important thing. So the big growth driver will be the movement -- the biggest growth driver is obviously subscription revenues, that's the fastest growth. In terms of confidence, the second part of the question, which is translated into sustainable earnings growth. I covered on one of my slides why we're confident in that subscription mid- to high-teens growth going forward because as customers go through implementation, you get to go live a number of contracts on the system increase and our subscription revenues increased. So we are confident in the subscription growth number. We're confident also in the delivery because of the late-stage pipeline. And again, I'll come back to the fact that Software Engineering is a little more difficult to predict because it very much depends precisely on what needs new customers might need for changes to the software. But we've always looked ourselves a Rule of 40 type of business over and meet any sort of medium-term time frame and that's always been around the sort of 30% type margin business with a 10% to 12% overall growth rate. And we still believe that's the shape of the business that we are today.

Brilliant. So the next question is about Software Engineering. Software Engineering revenue is down 17% year-on-year. Is that just normalizing after last year's bumper year? Or is client work dry up?

Let me have a go answering this, Andrew Dickson here. So as Duncan said, it's actually quite difficult to accurately forecast Software Engineering revenue. This is partly because it's very closely linked to the timing of new customer wins and also customer requirements. But if we actually look at the figures, the GBP 8.6 million that we reported in the first half of this year, whilst it was down against H1 '25, it was still up significantly against the first half of 2024, where we reported GBP 6 billion worth of Software Engineering revenue. So to some degree, it is really a function of timing. However, I think at the same time, it is worth noting that the first half of this year did include GBP 2.8 million of customized license revenue, which is expected to be slightly lower in future periods. I think that's the best explanation I can give for that.

Brilliant. So moving on to our next question. Do clients ever build this build this stuff in-house instead of buying Alfa systems. Is this still a competitive threat? Or has that basically gone away?

It's Andrew Denton here. I'll have a go at that one. It definitely used to be more of an option. We -- if we look at our total addressable market of just over $3 billion per year annual spend, about $1 billion of that is spent on self-build systems. But things certainly seem to have changed. We're in a position now, I think, where large institutions are more inclined to do what we call stick to the knitting and concentrate on being large institutions. And why is that? Well, there are a number of problems with self-build. One of them is what does good look like. People are investing in these large technology-led transformations in order to make their businesses better. And there is a real risk that what you end up doing is pouring concrete around our existing business processes, whether they're good or bad. There's the risk. We've had 36 years of refining our system. If you're building your own right now, you have to get it right first time. There's the fact that it might fit you like a glove right now, but nobody knows what the future will bring and you've got a better chance of retaining the flexibility you need to be future competitive by buying a package. And then finally, all of the regulatory change is on you going forward. whereas you buy something like Alfa, and we will take care of it. We'll also take care of what the regulatory change looks like. So pretty compelling reasons not to do it, I think, which is what these organizations are seeing as why they're not on the whole doing it. There might be a question within the question around AI. We're often asked about the threat of AI creating a situation where people build these systems. We're very certain that nobody is going to live code or cloud code and Alfa. On AI, as we're finding internally with our users, the likes of card definitely speeds up development. But speeding up, development doesn't actually solve any of the risks I just outlined.

Brilliant. So on to the next question. Share price was over 200p earlier in the year, and now it sits just at 167p. A reasonable performance today, but why do you think the market keeps marking this down when the underlying numbers still look decent.

I'm going to take this one again. I'm not as far as to say today has been a good day. I'm looking at 173p, but you're right. I think that the market doesn't quite get it. Why is that? I think some of it is within the material that Duncan outlined around the general market, not quite understanding that I use rather more [ focus ] language and say that a lot of that revenue is in the other it's just not fully cooked, which Duncan did a thermal scientific job of explaining. But either way, we are very certain that the ramp-up in subscription revenue from the projects that we're working on is around the corner for us. And I don't think that's appreciated. And also I think sometimes the financial markets do paint with a very broad brush. So when people worry about AI, I think they worry about all technology. And I think that enterprise technology, such as the market that we're in is somewhat more impervious to the threat of AI. And then finally, it also feels somewhat like we've been caught up in what some people call SaaS Pocalypse. And that was the concern, I think that AI making businesses more efficient would that will be deleterious of revenue for those businesses that licensed on a per seat basis. We've never done that. We've always licensed on volume because even before AI Alfa made these businesses more efficient. So a combination, I think, still is not quite understanding the business and where we're going and what our future looks like. And also us being swept up in larger market tracts movement. So I agree with the question or the implicit question there that we are somewhat undervalued and under understood.

To the next question. The big question seems to be whether Alfa can convert the strong ARR and TCV growth into earnings. What should investors expect from operating margins over the medium term?