Welcome to the Verisure Group third quarter results 2026 presentation. Today, I am joined by CEO Austin Lally and CFO Colin Smith. For the first part of the conference call, all participants will be in listen-only mode. If you wish to participate in the questions and answers session after the prepared remarks, then please dial into the telephone conference and press pound key five on your telephone keypad to enter the queue. Please note that you will not be able to ask questions if you have joined on the audiocast link. Now, I will hand the conference over to the speakers. Please go ahead.
Thank you, operator. Good morning and welcome to our third quarter 2026 earnings call. To start, I want to remind everyone that you can find today's presentation and earnings release on our investor relations website, together with long-term trending schedules detailing our operating and financial performance. As usual, Austin will begin with opening remarks and operational highlights. I will then walk through financial performance in more detail before handing back to Austin to cover strategy, priorities, and outlook. With that, I will hand it over to Austin.
Well, thanks, Colin. Good morning, everyone. Thank you for joining us today. Q3 was another solid quarter of delivery, characterized by quality growth, driving strong cash generation and margin expansion ahead of our expectations. We grew our customer portfolio to around 6.5 million, prioritized quality customer intake, reached our highest-ever adjusted EBIT margin, and delivered positive free cash flow for the fourth consecutive quarter. With that, let's turn to slide two. I would summarize today around three themes. First, Q3 delivered quality growth, record margins, and strong cash generation. Annualized recurring revenue reached EUR 3.712 billion, up 11.8% year-over-year, which includes around 2 percentage points from ADT Mexico. Adjusted EBIT was EUR 285 million, up 13.9%, at a record 27.7% margin. And we were pleased with free cash flow of EUR 59 million in the quarter.
Second, turning to our outlook for 2026, Q3 trading again demonstrated the resilience of our recurring revenue model. Portfolio services economics remain strong, with ARPU, EBITDA per customer, and attrition all tracking in line with or ahead of our expectations. New installation volumes, however, remain softer than our original plan in a small number of markets where macroeconomic conditions and consumer confidence remain soft as the past few quarters. We again prioritize quality over pace, and we continue to exercise discipline in those markets. As a result, we are refining our 2026 ARR outlook to between 9%-9.5% growth. We expect our adjusted EBIT margin to be around 26.5% and confirm free cash flow of around EUR 200 million. Third, capital allocation. We are approaching our 2.5x leverage target. This will create capacity for shareholder returns beyond the ordinary dividend. With that, let's turn to slide three.
Our third quarter headlines tell a story of consistent delivery across the board. Our portfolio is now around 6.5 million customers, up 9% year-over-year, reinforcing our position as the global leader in professionally monitored security. New installations were 215,000, broadly stable year-over-year, reflecting our focus on quality intake against a more challenging consumer backdrop. Total revenue increased 9.2% to EUR 1.031 billion, and annualized recurring revenue reached EUR 3.712 billion, up 11.8%. Adjusted EBIT was EUR 285 million, up 13.9%, with margin expanding 116 basis points to 27.7%. This was a quarter of excellent financial delivery, our highest-ever quarterly revenue, adjusted EBITDA, adjusted EBIT, and adjusted EBIT margin, and our 16th consecutive quarter of double-digit adjusted EBIT growth. Turning to slide four. Here, we set out the continued compounding of our portfolio alongside sustained low attrition.
On the left, our portfolio grew by 534,000 customers year-over-year to almost 6.5 million, with 97,000 net additions in Q3. On the right, LTM attrition was 7.5% within our attrition corridor. Quarterly annualized attrition was 7.3%, 18 basis points above last year, including the higher attrition of the ADT Mexico portfolio acquired in Q4 2025. Our AI-powered customer insights engine, now live in Spain, France, with several other countries following soon, proactively identifies attrition risk. Attrition remains one of our most differentiated KPIs. With that, let me hand over to Colin.
Thanks, Austin. Let's next turn to slide six. As always, I'll talk to growth rates in constant currency where applicable. Q3 was a quarter characterized by strong financial delivery, with cash generation and margin expansion ahead of expectations. Annualized recurring revenue was EUR 3.712 billion, up 11.8% year-over-year, including around 2 percentage points from ADT Mexico, which, as a reminder, we lap in Q4. Total revenue was EUR 1.031 billion, up 9.2% year-over-year. Adjusted EBITDA was EUR 489 million, up 9.7% on a year-over-year basis. Adjusted EBIT was EUR 285 million, up 13.9%, with margin expanding 116 basis points to 27.7%, its highest-ever level, while absorbing EUR 7.2 million of Spain rebrand media investment. On a year-to-date basis, adjusted EBIT is EUR 831 million, up 15.6% year-over-year.
Year-to-date margin increased to 27%, up 143 basis points, supporting today's upgrade of our full-year EBIT margin outlook to around 26.5%. Adjusted EPS increased to EUR 0.16 from EUR 0.09 per share prior year, reflecting strong profit growth, lower interest costs, and good progress on tax. And free cash flow was positive EUR 59 million, our fourth consecutive positive quarter, taking the year to date to EUR 154 million. Let's now turn to slide seven. ARPU delivered strong growth in the quarter at EUR 48.10, up 3% year-over-year. Our ARPU growth reflects a lasting impact from the January price increase and underscores the stability and strong engagement from our customer portfolio. We also delivered increasing upsell and low discounting. As I've often noted, we have no front book/back book dynamic in our portfolio. New customers join at the same level of ARPU as our portfolio.
Recurring monthly costs were EUR 12.30, up 3.5% year-over-year, with around 50% of the increase from ADT Mexico's higher inherited cost base, which will reduce over time as we integrate and introduce Verisure processes and systems. Bringing this together, EBITDA per customer reached EUR 35.80, up 2.8% year-over-year, increasing the profit contribution from each customer every month. Portfolio services EBITDA margin is at 74.4% in the third quarter. Turning to slide eight and customer acquisition. Q3 new installations were 215,000, broadly stable year-over-year. We saw country-specific macro-led pressures in France and Latin America, specifically in Argentina, while our major markets elsewhere delivered very good year-over-year growth. The complex consumer macro made it challenging to grow at the pace of our 2026 plans. We chose, as always, to prioritize quality over volume, with no compromise on credit scoring, entry pricing, or discounting.
With an average customer lifetime of around 15 years, maintaining acquisition quality is imperative to long-term value creation. Cost per acquisition was EUR 1,658, up 9.2% at constant currency, including around EUR 33 per installation of Spain rebrand media. Excluding rebrand, CPA was EUR 1,625, up 7% year-over-year, reflecting the lower upfront revenue per installation, as previously guided, like-for-like media inflation, and underabsorption of the fixed cost element of CPA against lower installation volumes. Our acquisition multiple was 3.9 times, up 0.3 times year-over-year, or 3.8 times excluding rebrand, with higher CPA partly offset by increased profitability per customer. New customer cohorts continue to generate an unlevered IRR of around 20% over 15 years, well in excess of our cost of capital. Moving now to cash flow on slide nine.
As noted, the company generated EUR 59 million of positive free cash flow in Q3, an improvement of EUR 86 million year-over-year, and our fourth consecutive positive quarter. We have moved firmly through our cash inflection point. Year-to-date free cash flow is EUR 154 million, which is EUR 278 million better than the first nine months of 2025. Stepping through the quarter three cash flow bridge. From adjusted EBIT of EUR 285 million, we add back D&A and customer acquisition EBITDA, and deduct portfolio and other CapEx leases, and a EUR 64 million working capital outflow. The working capital investment reflects inventory build and lower trade payables. That takes adjusted portfolio operating cash flow before customer acquisition to EUR 515 million, up from EUR 488 million a year ago.
We invested EUR 195 million in attrition replacement, acquiring around 118,000 customers, and EUR 160 million in organic portfolio growth, growing our book by 97,000 customers in the quarter. Adjusted operating cash flow was EUR 159 million. Below the line, we paid EUR 18 million of tax, with net interest paid of EUR 72 million, down from EUR 132 million a year ago, reflecting the excellent progress made over the past 12 months on refinancing. Last, a note on the right of the chart. Adding back the EUR 160 million of organic growth investment, pre-growth cash generation was EUR 219 million, up 62% year-over-year, underlining the growing cash generative strength of our portfolio. Let's turn now to slide 10. On the left, our leverage ratio was at 2.7x at the end of Q3, including the EUR 103 million interim dividend paid in the quarter.
This remains within our year-end 2026 guidance range of 2.5x-2.75x . Net debt closed the quarter at EUR 4.95 billion. On the right, looking ahead. First, our cash inflection is firmly validated, with EUR 154 million of free cash year-to-date, driven by strong progress in operating cash flow, our working capital improvement program, and lower financing costs. The group's weighted average cost of debt was approximately 4.3% in Q3, within our target range, with 2026 interest savings versus 2024 at the upper end of our EUR 200 million-EUR 220 million target, as guided at the time of the IPO. After the quarter closed, we repaid EUR 350 million of our Term Loan B, funded by a EUR 250 million Term Loan A upsize and EUR 100 million from our RCF, cutting our weighted average cost of debt by around 5 basis points.
We have no material debt maturities until May 2030, and at Q3 close, we had available liquidity of EUR 911 million. We are making excellent progress towards our 2.5x leverage target, which will give us increased financial flexibility and create capacity for shareholder returns beyond the ordinary dividend. We reflect on Q3 as a solid quarter of delivery and progression against a complex macro. Our financial framework is working optimally, with quality growth generating strong margin progression and free cash delivery. With that, Austin, back to you.
Thank you, Colin. Over the next few slides, I will cover the playbook that has made us the global category leader, how quality customer intake enhances long-term portfolio value, and the growth opportunities and cost transformation that will support our next phase. Our global leadership position has been built over more than 35 years through a consistent playbook to unlock the market, resting on four pillars. Continuous innovation is core at Verisure, delivering proprietary, differentiated products and services. Our human-led monitored security focuses on protection in the home, at the perimeter, and beyond the home with Guardian. Innovation is centered around modules that open new segments and drive upselling. More on this shortly. Secondly, category-creating marketing. We invest over EUR 300 million a year, around 8% of sales, in category-creating marketing to drive demand and stay top of mind.
Our brand leadership is now reinforced as Securitas Direct in Spain and Portugal transition, uniting our markets under one brand, Verisure. We began preparing to retain this leadership in an LLM world around 18 months ago, with a dedicated lead in every market. AI-led search rewards authority signals, deep digital presence, and real-world footprint, including our local branches able to reach customers within a day for a security audit or urgent installation. As number one in almost all of our 18 markets, Verisure is well-referenced across LLMs. 25% of consumers now start their category journey on LLMs, but typically revert to our website. Our share of pool remains stable across our multiple channels, and since then, Q3, we have added a new channel, our pilot advertising on ChatGPT, now live in four markets. Go-to-market muscle.
Our specialist field force of 12,000 security experts remains a strategic asset, increasingly complemented by an increasing number of strategic alliances that accelerate growth and customer experience. Our first-class customer service, supported by AI, data, and analytics, is one of the factors that keeps our attrition low. Increasingly, on-device AI diagnostics enable more faults to be resolved remotely and at first contact, reducing technician call-outs and simultaneously improving customer satisfaction. In the third quarter, the total number of customer calls analyzed by AI more than doubled quarter-on-quarter to almost one million. All four pillars are underpinned by our culture and talent. With over 30,000 colleagues across 18 countries, we are proud to now also be recognized as one of the Fortune 100 best companies to work for in Europe.
Verisure has always operated with a clear philosophy: acquire the right customers at the right returns in the right markets. That discipline has made our portfolio of around 6.5 million customers one of the highest quality recurring revenue assets in European consumer services. On the left is our self-reinforcing growth model. Quality at the point of sale builds a growing long-duration portfolio. That portfolio generates strong economics, which we reinvest in growth, and increasingly, return to shareholders. Every decision is designed around one objective: maximizing long-term value per customer. On the right, you can see what that focus on quality delivers. Annualized quarterly attrition of around 7.3%. Industry-leading retention. ARPU of around EUR 48, up 3% in Q3. A portfolio EBITDA margin of 74.4%, underpinned by expanding EBITDA per customer, and an average customer lifetime of around 15 years.
The customers we add today determine our attrition and ARPU trajectory for years to come. So by maintaining our high standards, we protect the long-term recurring earnings power of the portfolio. Slide 14 highlights three areas that will support our next phase of growth and margin expansion: technology and innovation, alliances and partnerships, and our Fit For Growth cost program. First, technology and innovation. We have signed an exclusive partnership with U.S. company Ting Labs to bring its electrical fire prevention technology to our customers in Europe, fully integrated into our alarm ecosystem, with launch expected in early 2027. Kivala, our digital building access solution acquired in Q2, is now integrated and launched in France. Our AI-powered GuardVision outdoor detector is live in 12 markets, with outdoor solutions helping to prevent intrusion attempts.
With over 200,000 GuardVision detectors already installed today, we are seeing strong attach rates in Italy, France, and Germany. Second, alliances and partnerships. We are accelerating partnership sales, with 23 alliances now active across the group. 2026 was a record year for new partnerships, most recently with the insurer Frende in Norway and Danske Bank in Denmark. Alliances contribute over 10% of new installations, typically at a lower-than-average CPA. Third, Fit For Growth. This is an ambitious, broad-based cost program spanning material costs, digital first, organizational design, and sales and marketing efficiency. Increasingly, we're using AI to reimagine our processes, from AI-supported customer care and remote diagnostics to our advanced monitoring model in our alarm receiving centers, and a new AI-powered recruitment platform. Taken together, these initiatives support our future growth and margin expansion. With that, let me move to the key takeaways.
First, Q3 delivered record profitability, margins, and cash generation. Our portfolio grew 9% to around 6.5 million customers. We reached our highest-ever margins, and free cash flow was positive for the fourth consecutive quarter. Second, our formula for value creation has not changed. Quality customer intake remains our core operating principle. In markets where consumer confidence is below its long-term average, we choose quality over pace. By maintaining our standards, we protect the long-term compounding earnings power of the portfolio. Finally, we are creating capacity for additional shareholder returns. We have moved firmly through our cash inflection point and are delivering free cash flow consistently each quarter. We are progressing well towards our 2.5x leverage target, after which we create the capacity for additional shareholder returns beyond the ordinary dividend.
Verisure is defined by disciplined quality growth coupled with consistent execution, and Q3 is another clear demonstration of that. Before we move to the Q&A, I'd like to note upfront that we will not be commenting on any questions regarding recent press speculation. With that, I'll hand back to the operator, and Colin and I look forward to your questions.
Thank you. As a reminder, if you wish to participate in the question-and-answer session, please dial into the telephone conference and press pound key five on your telephone keypad to enter the queue. If you wish to withdraw your question, please dial pound key six on your telephone keypad. We please ask all analysts to limit their questions to two per person. We will take a short pause as we wait for the first question. As a reminder, it is pound key five to join the queue. Thank you. The next question comes from Annelies Vermeulen from Morgan Stanley. Please go ahead.
Hi, good morning, Austin and Colin, and I have two questions, please. Firstly, on the guidance, could you unpack the moving parts within the 9.5% ARR guidance for 2026 and what that assumes for Q4 in terms of installs and attrition? As part of that, are you still confident on hitting your medium-term guidance of 10% ARR growth per year, given these moving parts?
Then secondly, just on installations and new customer acquisitions, you've touched on this already, but could you talk a bit more about how you're navigating the increase in customers using search and how Verisure is positioned for that? Austin, I think you mentioned you're advertising within ChatGPT, so is that increasing your marketing costs? I appreciate a large number of your customers coming through non-digital channels like telesales and alliances and so on, but I'm just wondering how that digital customer acquisition piece is evolving at Verisure with the rise of LLMs. Thank you.
Thanks, Annelies. Good morning. Why don't I go first on outlook, and then Austin can pick up LLMs. Look, I think to start, we'd characterize the refreshed 2026 outlook as balanced and broadly in line with that we previously provided earlier this year in February. For the company, the outlook underscores that our financial model is working well as designed, and as talked a lot with investors over the past year. If I start with ARR, as you know, we've got an outlook projected now at 9%-9.5% in 2026. I mean, that reflects the complex macro that we've worked with this year, particularly low consumer confidence, and our focus on disciplined growth, always prioritizing quality intake with no compromises, as I said earlier on credit scoring or on entry ARPU.
We are seeing some country-specific pressures in France and Argentina, with heightened consumer volatility in both markets. If I just talk to how we get to the 9%-9.5%, as we've said, we lapped the ADT Mexico acquisition in the fourth quarter, and we've said often that that's worth around 2 percentage points of ARR growth. For the fourth quarter, look, I'm not going to guide on installations, but the one point that I would like you to note is that we are lapping a very strong growth quarter in Q4 2025, where we grew installations at around 5.9%. At the same time, we're upping our EBIT margin to around 26.5%. That delivers over 100 basis points of margin progression this year, and that's driven by excellent ARPU growth and really good cost control. That manifests, of course, in increasing earnings per customer on a month-to-month basis.
We are well on track towards our 30% long-term EBIT margin target. Of course, free cash flow, we're guiding today at around EUR 200 million for the year, and cash conversion continues to improve. We paid our first interim dividend in September, and that really did symbolize a start of a new phase of shareholder returns. Look, in short, we're pleased with how financial delivery is shaping up for 2026 against what's been a complex year macro-wise.
If I just look forward into 2027, I mean, as you would expect, we intend to update on our 2027 outlook alongside full-year results in February. Today, we're focused on refining our 2026 outlook. On the medium term, we're not changing medium-term guidance today. We're working through our budget and operating plans, including growth, capital allocation, and pricing changes for 2027. To confirm, we intend to update the market in February, providing an update on our outlook for 2027, as is customary.
I'll say, good, Colin. Can I maybe pick up the LLM thing? Annelies, good morning, and nice to hear from you. I would say it's still early days for AI's impact on consumer journeys. So far, we haven't actually seen a material change in channel mix for incoming bookings. Now, we're estimating that about 25% of consumers are now using LLMs to search the category. I mean, that started being almost all ChatGPT. We now see Gemini becoming more significant. But we're also seeing that behavior being a bit complementary. They typically still revert to our website because they don't fully trust AI-generated answers. So, reliance on traditional channels like Google, it's still there. Now, that's going to change over time. We started preparing for this shift, as I mentioned, like around 18 months ago. We put a Generative Engine Optimization leader in every market.
We want to be a pioneer here, so we started at the end of Q3 already advertising on ChatGPT in four markets. We actually started in the U.K. in June. In terms of your question on cost, it's not big money. I mean, at the moment, it's a fairly limited investment. We're trying to figure out actually how well it works. We'll then look at the ROI on things like ChatGPT and Gemini compared to other alternative ways to spend. I've said in previous calls, we actually think that this will be good for our business because, I mean, AI search, I mentioned in my remarks, it rewards brands with authority signals. So, awards, customer reviews, having a large real-world footprint, having a deep digital presence, having a lot of superior content. I mean, we're the number one brand in almost all of our markets.
If you look at the LLMs today, I mean, Verisure is very well referenced across those models. If you look on ChatGPT, we're number one for best alarm on most markets, and we're actually tracking that on a weekly basis. There's also some discussion. I mean, I've seen discussion about whether LLMs actually favors local players. I think for us, this is actually very good because we're both a local and a global company. We've got the leading brand, but we've also got the leading local presence.
We've got more than 600 local branches. I mean, basically, a Verisure salesperson is closer to any potential customer than any competitor in terms of driving time, in terms of speed of response, and appointment. We're like the most local player in the industry, but we're also the most scaled global player. Now, we keep monitoring it, particularly, what's the right spending profile and what do we get back? But we actually think over time, this is actually going to build advantage for us rather than being a drag on us.
That's super helpful. Thank you both for the detail.
Thank you.
Thanks, Annelies.
The next question comes from Joachim Gunell from DNB Carnegie. Please go ahead.
Perfect. Thank you. You report very strong underlying profitability here, and then raised just a little bit margin guidance here despite inflationary backdrop. Just how much of this underlying strength is structural, versus, say, some timing or one-off effects? Can we expect this strength to carry into 2027?
Morning, Joachim. It's Colin. Thank you. I mean, I think, let's start by talking about the EBIT margin in Q3. I mean, the one thing that I would remind everyone of is that that EBIT margin was delivered inclusive of the fact that we are investing in our Spain rebrand. We invested over EUR 7 million in the quarter on that rebrand. So, actually, I think the margin underlying that and normalized for that is actually even a little bit stronger. Let me unpack kind of what drives our EBIT growth as a business, just as a reminder. I mean, I think the primary driver is volume. We grew the portfolio 9% on a year-over-year basis, and that's the single biggest driver of EBIT progression.
The second one is the fact that our portfolio is becoming more valuable, and our EBITDA per customer is actually increasing, as I noted earlier, by 2.8%. So, you've got a volume and you've got a rate driver here, where together, those two things drove our EBIT year-over-year up by around EUR 80 million. We see those as being absolutely part of our model and something that we would expect to continue going forward. Now, the investment comes in terms of our CPA, as you touched on.
We have an increased CPA year-over-year. That does include those rebrand costs, and that basically led to us investing around EUR 30 million- EUR 40 million more in customer acquisition. Then lastly, you have higher D&A on a year-over-year basis. Net-net, those are the four kind of driving components of how we get our EBIT nominal and our EBIT margin up year over year. I think there's nothing in there that's in any way different to the decades of performance that we've delivered as a company and as a team.
Lovely. Just a follow-up, just say high level, if this call it inflationary backdrop and then softer consumer world continues to be a pressure into 2027, in what way do you think that the playbook here is different from you in light of how Verisure performed in, say, 2023 when attrition crept up, ARPU growth slowed a bit, and then new installations volume declined year over year?
Let me say a few words on this. I'm actually delighted with the attrition performance, if you go back and contrast it to 2020 to 2023, because it's been rock solid, right? We did see a modest increase back in 2022, 2023 that you haven't seen in 2026. I think the reason for it is we've got capabilities now on things like sentiment analysis and AI predictive models that we didn't have then. By the way, we're not fixing it with discounting. That's the point that Colin made. I mean, we're doing it with service, with speech, with analysis, with innovation. I think the other thing, though, that's very consistent with 2022 and 2023 is we understand when it's a bit more difficult in a market like France, we just don't chase the volume, right? We go for quality.
Maybe, Joachim I can just build a little bit on CPA in particular, just to add to what Austin said. I mean, I think the one thing that I would point to would be cost pressure, and we've noted this before, on hardware cost. I think the one thing that we should bear in mind is that hardware cost as a relative proportion of our customer acquisition investment is reasonably low. So only around 25% of our CPA is hardware. I think the company's clearly been focused on ensuring that we can secure supply in this supply chain world that we live in. We've done that, and I'm very happy with where we're at. We actually invested in working capital in the quarter to make sure that we could do that, as you may have seen.
We now have around six months of stock on hand, and I think that is absolutely the right strategy in this environment. The one thing that we are seeing is some cost inflation on that hardware. As I say, it's nothing particularly material in the grand scheme of things. As I look to the consensus for CPA in 2027, I think that's something that I would acknowledge and feel generally comfortable with. I think on the consumer side, you've rightly called out the fact that we do need to be mindful of impacts in previous years. We don't see the CPA environment growing anything like what we saw in 2022 and 2023.
I would just bear in mind that a slightly elevated CPA environment is actually positive for the business, because it gives us a bit more pricing power on the front end and drives that EBITDA per customer that I was talking about earlier, which ultimately drops straight to profitability. So, hope that helps give you a bit more color.
Okay, perfect. Thank you.
Thanks.
The next question comes from Suhasini Varanasi from Goldman Sachs. Please go ahead.
Hi, good morning.
Morning.
Just to pick up from me as well, please. Morning. Just to go back on the new installs question, please. Given the softer macro environment and considering that Q4 does have pretty tough comps, just want to get some sense on, do we expect flattish trends in new installs similar to Q3 in the macro? Can it even be down year-over-year? Just cognizant that you've also invested in inventory in working capital. Just wanted to get some color there, please. The second one is on your portfolio services EBITDA margin, which is broadly flat year-over-year, maybe down 10 basis points. How should we think about the scope for improvement over the coming quarters, please? Thank you.
Morning, Suhasini. Nice to speak. I mean, I think, look, first on installs, as you'd expect, we won't be giving any guidance on Q4 installation volume expectation. I think what we focused on this morning is providing our latest view on 2026 outlook that includes ARR at 9%- 9.5%. So that's kind of what I'd want to say on that one. Can you just repeat your second question, please?
Portfolio EBITDA margin.
Yeah, on portfolio EBITDA margin, my apologies. We are at 74.4%. I think first things first, that is a very strong monthly contribution on average across our customer portfolio. I think if we look at the RMC side of things, you will note that we were up by 3.5% on a year-over-year basis. I mean, there is an opportunity in here in that the inherited cost base from Mexico is significantly above the group average. Now, we acquired the Mexican business a year ago, and it is going to take us a little bit of time to implement the processes and the systems and the know-how, frankly, that we have deployed over the rest of the business over the past 10 years. We do see that as an opportunity, and that is a very specific country item.
I think the more broad opportunity on this, as we have talked many times before, is what we can do with AI. Particularly in certain pockets of our RMC cost pool. We are focused on customer care, and we are focused on field maintenance, so kind of sending engineers out to fix a problem. What we are doing with AI is we are entirely reimagining customer process.
We are not just using AI to automate certain tasks within that process that used to be carried out by humans. Instead, we want to rethink the entire workflow. I think, look, this is not going to happen overnight, of course, but we are excited about the opportunity that that can bring. As I have said before, out of our total RMC, those two areas, customer care and field maintenance, make up around 1/3 of our RMC. There is plenty to go for there in terms of opportunity.
Thank you.
Pleasure.
The next question comes from James Rowland Clark from Barclays. Please go ahead.
Hi, good morning. Two questions, please. My first is on the quality customer intake you have spoken a lot about today. I just wonder how you can explain the upfront fee still coming down further than I think the last few quarters, I think down just over 10%. How do you measure that customers are high quality at this installation stage? So what are the KPIs that you are looking at when you have not actually seen whether they are going to churn as a cohort at this point?
Secondly, on attrition, that is nudging up year-on-year by nearly 20 basis points on your Q3 annualized figure. I think last quarter you were very pleased with how you were managing that to flat year-on-year despite ADT Mexico. What has changed in the attrition, please? Is that Mexico seasonality or is there a bit more attrition coming through in those weaker markets at the moment, like France and Argentina? Is there something else to think of? Thank you.
I mean, we would regard the attrition change as being a little bit of a blip, right? It is a relatively small change. I mean, I think the absolute level we still think is best in class in the industry, right, and well within the corridor. I mean, there is always a bit of country mix in this. There is definitely a Mexico effect in this. Remember, I mean, that was an attrition profile that was in line with the rest of our Latin American business, but it was a bit higher than the group average. Now, in terms of, let us say, the management systems to keep this under control, I mean, I think the first thing I would say, the level of upfront investment that the customer makes is always sufficient to qualify the customer. Okay? I mean, we do not give away alarms for free.
The customer still has to make a decent contribution to the cost of the system. I think one of the misunderstandings people sometimes have is they wonder if there is a linear correlation between the upfront revenue and the attrition profile. I say, if that were the case, there would be no attrition in Belgium, right? Because the upfront revenue in Belgium is the highest in the company. Actually, the attrition profile in Belgium is bang in line with the group average, right? Bang in line with what we would do in a market like Spain. So, what we control is making sure that there is always a qualifying level of investment for the customer, number one. Second topic is around not discounting the ARPU. A salesperson, for example, who is sitting in front of a customer does not have the authority to do deals on the recurring revenue.
I mean, the recurring revenue is sort of sacrosanct in the model. We may offer, for example, a sharper price on the upfront if we are facing a competitive bid for the customer, but never at a level, right, where we would put the quality of the customer at play. The other thing we are also doing, pretty much on an ongoing basis, is we are looking at the early behavior of the cohorts who are joining the portfolio, and we are comparing those early behaviors to previous cohorts in previous years. So the 15-year customer lifetime is not some theoretical extrapolation.
It is actually something that we have observed over time by studying different vintages of customers. The reason we do it is, it is obviously super important to us. I mean, in a portfolio of nearly 6.5 million customers, if you were blending poorer quality customers into the portfolio, it would take you a while to see it if you were not vigilant and really checking those early behaviors and those signals all the time.
I think, James, just to add, I mean, on the upfront, and we have talked about this before. I mean, firstly, look, the upfront in the quarter was EUR 362 on average, which is well within the comfort corridor that we have around this metric. Austin talked earlier about the fact that this is an important qualifier for quality. Again, something else that I have mentioned before is the apartment mix. So, as at today, we have got a portfolio that is made up of approximately 25% apartments.
Our acquisition mix in the quarter was around 35% apartment. I mean, clearly, an apartment typically has fewer components than a villa, and consequently has a lower upfront. Now, what we do, and we have talked about this, is we try and preserve as much of the ARPU as we possibly can. The delta in ARPU between villas and apartments is actually quite tight, but we do have a lower upfront. I just wanted to add that additional color.
Yeah, very good.
Thank you. Sorry, just a small follow-up there. What are the early behaviors you're looking at on these new cohorts to compare to old cohorts?
Well, the first basic one would be early cancellation. Somebody who contacts you in the first, like, two, three, six months and basically says, you know what? I think I've made a mistake. That's the first one. Second one would be things like usage behavior, right? Someone who's basically bought the system and doesn't really get into a good usage pattern. That's one we feel we get quite angry about that, because that basically means we've not done the right job. I mean, part of our process is the early contact with the customer to get them into the right usage habit.
I mean, it's worth contrasting that to cameras, for example. We get asked about camera competitors, because obviously there are a lot of lower-cost cameras out there. I mean, what you discover is actually within one or two months of someone installing a camera app, only about 30% of these cameras are still active. I mean, it is a solution where there is a big, big drop-off, but we just do not have that.
Great. Thank you very much for the color.
Thank you.
Thanks, James.
The next question comes from Andy Grobler from BNPP. Please go ahead.
Hi, good morning. Just one from me, if I may. Just in terms of some of the regional differences in installations, you called out France and Argentina as being particularly weak within the group. Can you talk through in fairly broad terms about how far down those countries are, and also kind of the offsetting growth rates in other countries that are doing a bit better at this stage? And I guess, just as a bit of a follow-up, you've talked to some of this through the call, is there anything that makes you think that anything on these metrics is going to change over the next three to six months? Thanks very much.
Look, I don't want to be Nostradamus on the economy. I'm not an economic forecaster. We're obviously watching this in real time, like you are, right? We're looking at all the consumer confidence pictures. I mean, if you stand back, it all adds up still to 9% portfolio growth. I think in a more difficult environment, our portfolio is up strongly year-on-year. That's what I'm pleased about. We noted France, obviously, because it's a big country and there's quite a bit of discussion about the consumer position, the debt, the upcoming election. I mean, what I would say in France is, we sort of see customers sitting on their hands, right? France tends to be, if I look through history, one of the markets where it's a bit like Canadian and coma. The French consumer, if there's some doubt, sort of delays. That business will come back.
We know that. We're already seeing some positive news on bookings there. I think, again, without getting into too much color, I've been very pleased, actually, with momentum in markets like Iberia, markets like Italy, markets like the U.K., right? And they've been, as part of this other Europe, they've been a strong contributor to our growth rates in recent years. Part of the message I want to get across is that the macro effect is in a few markets. I mean, it's not something that's consistent right across the footprint. We've actually got plenty of bright spots.
Okay. Thank you very much.
The next question comes from Virginia Montorsi from Bank of America. Please go ahead.
Good morning. A quick follow-up to James' initial question on lower install, if I may. Can you just help us? I think we're all appreciating the focus on customer quality, the discipline on discounting, and the quality of the cohort that we've observed. But if I look at the upfront revenue per install, which has been down year-on-year in Q4, I think I struggle a little bit to reconcile it with the focus on the quality of the cohort. Can you help us maybe try to better understand how these two elements can coexist? Is decline mostly than deliberate pricing? Is it more mix? How do we think about payback? Because when we think about the focus on quality and yet the low installs, I'm just struggling a little bit to go through moving pieces, if I may.
It's also just worth pointing out that when you assess the payback on a customer investment, a small change in upfront revenue doesn't make much difference to the calculation. It's basically the amount of earnings per customer per month times 12 months times 15 years. I mean, that's the power of the portfolio.
Virginia, it's Colin. I mean, I would, as I noted earlier on, I think the one point I'd focus on on the upfront is the fact it's still EUR 362 in Q3, which is strong, but really driven the movement by the fact that we've got a higher proportion of apartments in our acquisition mix, and apartments typically have a lower upfront. However, we do, and to Austin's point, we preserve the earnings per customer from an apartment customer is to almost the same level as we do a villa customer. I think the returns that we make on customer acquisition are still very strong, and clearly we're updating this on a month-to-month basis. We've got a highly disciplined focus on returns, and that talks to the fact that we're not going to do things that are uneconomic.
To get an extra couple of thousand of installations because we've got a high bar. We see all of these as individual investment decisions. We're investing EUR 1,600 in a new customer, and we want to make sure that that customer has the same characteristics as the other 6.5 million, and they're going to be with us for an average of around 15 years. So, returns are solid. We're still at IRRs of around 20% measured over 15 years. And frankly, when we look at it on an apartment basis and a villa basis, it's actually quite similar. Not quite as high as a villa, but very close.
It's still the best of the industry. And I, just repeating something I've said in previous calls, we don't guide on volumes. One of the reasons we don't guide on volumes is I don't want to create inside Verisure pressure for people to go out and just chase installations. I mean, we could sell more alarms. We know that. Instead of going for EUR 48 of ARPU, if I were selling alarms at EUR 30, I'd sell more. My conversion from booked meetings would be higher. But I'd be giving up effectively EUR 18 of EBIT per customer per month for 15 years. And that would be, I think from an economic standpoint, a bit silly. So, priority number one, it's making sure the customers that are coming in are basically going to have the same economics as the company and the portfolio so that we don't regret it later.
Understood [crosstalk]
Thanks, Virginia.
Thank you.
Thank you.
The next question comes from Jane Sparrow from JP Morgan. Please go ahead.
Morning. Just one question, please. Just on the dynamics of alliances and partnerships, which you called out as having accelerated in growth. I know, obviously, in the presentation, you highlight they deliver lower than average CPA. Can you just whether there are any other dynamics of growth from that particular channel that we should be aware of? Do they affect other key metrics, working capital dynamics, et cetera? Anything to highlight there?
No, not really. I mean, I think that what you typically find with high-quality alliance partners is it's actually, on average, a higher-quality customer, because maybe they've got a well-established banking relationship. They've got a strong credit history with the institution. It's often bundled with a mortgage, like homeowner, or with an important insurance policy. Again, one of the reasons why the CPA is often better is it's not being driven by the same level of high marketing investment that our traditional field model would involve. There's usually a commission to the alliance partner, but that's effectively replacing what we might have spent, for example, on marketing expense or maybe even on higher levels of sales commission. So, it's a kind of win-win.
I think for the alliance partners, I mean, it's not only revenue for them, but it can also help them with the stickiness, right, of the core offering like they have, particularly on things like mortgages and insurances. Now, it's over 10% of our sales today, but we actually think that it's got the potential to grow significantly over time. One example, we partnered with MasOrange, so number one telecom provider in Spain. That's opened access to a customer base of more than 30 million subscribers in our largest market. We're only a few months in, but we're already pleased with the annualized run rate that we're seeing. If I look at the 23 alliances that we've signed across the group, many of them are actually quite immature, quite young, quite small, with the opportunity to scale up.
I gave an interview this morning to Bloomberg, and we were chatting about alliances. I was asked where did I think there were big opportunities. One market I highlighted was Italy. We're now at 500,000 customers in Italy, and we've built every single one of them ourselves organically with no help. Yeah, but now we've got a national footprint. I mean, a precondition of doing an alliance is you've got to have national coverage, and you've got to be the number one brand, because alliance partners, when they work with you, they're putting their reputation on the line. So, I'm glad you asked about this, because it's one of the things I talk a lot about in the company as a growth vector.
Thanks very much for the color.
Thanks, Jane.
Thank you for your questions. I will now hand over to CEO Austin Lally for closing remarks.
Well, thank you all for joining us today, for the questions, for your continued interest in Verisure. We think Q3 was another proof point of how we run Verisure. Quality growth, driving strong cash generation and margin expansion ahead of our expectations. We remain convinced by the scale of our structural opportunity that lies ahead and our conviction in. So thanks.