Good afternoon, ladies and gentlemen, and welcome to Alkami Technology third quarter 2026 financial results conference call. At this time, all lines are in a listen-only mode. Following the presentation, we will conduct a question- and- answer session. If at any time during this call you require immediate assistance, please press star zero for the operator. I would now like to turn the conference over to Steve Calk. Steve, you may begin.
Thank you, operator. With me on today's call are Alex Shootman, Chief Executive Officer, and Cassandra Hudson, Chief Financial Officer. During today's call, we may make forward-looking statements about guidance and other matters regarding our future performance. These statements are based on management's current views and expectations and are subject to various risks and uncertainties. Our actual results may be materially different. For a summary of risk factors, please refer to today's press release in the sections in our latest 10-K entitled Risk Factors and Forward-Looking Statements. Statements made during the call are being made as of today, and we undertake no obligation to update or revise these statements. Also, unless otherwise stated, financial measures discussed on this call will be on a non-GAAP basis. We believe these measures are useful to investors in the understanding of our financial results.
A reconciliation of the comparable GAAP financial measures can be found in our earnings press release and in our filings with the SEC. I would now like to turn the call over to Alex.
Good afternoon, and thank you for joining us today. In the third quarter, Alkami again delivered revenue growth and profitability ahead of our expectations. Our end market continues to invest in modernization, which gives us a consistent pipeline of opportunity. The journey our market has taken from the ChatGPT moment at the end of 2022 to Opus 4.6 in February supports our DSSP strategy and creates new product possibilities. I will come back to that later in the call. In the third quarter, we signed five new digital banking relationships, three banks and two credit unions. Two of the five also contracted for DSSP, our digital banking, account opening, and data and marketing products together. This brings our year-to-date total to 16 digital banking new logos, compared with an average of 21 through the third quarter since becoming a public company.
Our digital banking pipeline is consistent with historical norms. However, we believe speculation about Alkami's future ownership created uncertainty for customers and prospects and has been an overhang on our new logo business. Providing clarity to our end market was an important reason we publicly communicated the conclusion of our strategic review in September. The results of that review was a unanimous decision by the Alkami board that executing our long-range plan as an independent public company was the best path to maximize shareholder value. With this behind us, we believe the business has the potential to return to its normal cadence. Along with the digital banking new logos, we signed 14 new MANTL clients. Five additional clients contracted for DSSP, bringing the number of clients contracted for all three DSSP products to 60.
We also now have 544 unique clients contracted for digital banking or MANTL, including 86 contracted for both platforms. Core conversions are an example of the high-impact work we deliver. A core conversion is when an existing client switches their core banking system. We completed seven this year, a record, including six of our largest clients representing more than 3.5 million digital users. Next year, we will nearly double that. In Q3, we brought 10 digital banking clients and 26 MANTL clients live, and over the last 12 months, we added 2.8 million users to our digital banking platform. Last week, we launched four digital banking clients in a single day, a record for Alkami. Five more follow, which will be nine launches in 12 business days.
From 2020 through the end of Q3 this year, we have brought 229 financial institutions onto our digital banking platform, along with their 8.4 million account holders. I am proud of our ability to execute critical client projects. When the real competitor is staying on a legacy platform, what changes an institution's mind is confidence that the migration will work. I am hearing much more clarity from our clients about where AI matters to them. People will still save, spend, borrow, move money, and open accounts. Those jobs have been the same for 100 years. What changes is who does them. Today, the account holder navigates menus and completes each task. In the AI era, they will describe a goal in plain language, and the software will work out the steps and ask for permission to act.
That shift shows up in two places, and we are building for both. The first is inside the institution's own app. Alongside the digital banking experience more than 24 million users rely on today, we are building an agentic conversational experience. It will spot a cash shortfall before a bill is due, suggest a safe amount to move into a CD, or line up a card paydown that will not put rent at risk. These experiences need all three parts of DSSP. The agent reasons over the data and acts inside digital banking. Opening the account or originating the loan depends upon the institution's policy and where they want it, a human approval. The second is outside the app because many account holders will also use general-purpose AI assistants. I think those assistants will be excellent at conversation and limited at banking.
To move money, open an account, or approve a loan, an agent needs governed access, identity, entitlements, limits, approvals, and the systems of record underneath all of it. Alkami runs that layer. We are building connectors into the leading AI assistants so that our clients' account holders can bank through whichever one they use. All of this comes back to control. Our clients want to decide which agents are allowed in, what those agents can do, and where a person has to sign off. Across more than 300 live clients, I hear the full range. Some want read-only access and nothing more. Others see agentic action with controls as the thing that lets them compete with a mega bank. Even the enthusiasts move in stages. Assist first, then recommend, then prepare an action for someone to approve. Acting inside rules they set will come last.
Here is why I think Alkami is positioned for this. We run a single codebase multi-tenant SaaS platform. Every institution on it configures roughly 4,000 policy settings of its own across approximately 25 systems of record that are integrated into Alkami. We built that so our clients could each run their own bank on shared infrastructure. It is also what an agent needs. One place to ask, one set of permissions to respect, and a record of every action taken. We are doing this now. There are three agents in beta inside our onboarding and account opening platform built for the people who run the institution. A KYC agent summarizes a manual review and recommends a disposition against the institution's own policy. A document review agent reads borrower documents, flags mismatches with the application, and coaches the applicant on what to upload.
An MCP server lets a financial institution ask a question from an AI assistant and get an answer grounded in their own application and account data. The main conversation we are having with all of our clients is that every generation's front door to the bank has changed. It was a branch, became a drive-through, then an ATM, the internet, and mobile devices. The next generation's front door will be an agent, and it will interact with digital banking the way the internet and mobile do today. Beyond deciding how agents interact with their customers, our clients share two other priorities. The first is their own transformation through the application of AI technology. As I mentioned on our last call, we built an internal platform in daily use by more than 100 people across our company.
It understands roles and skills, protects sensitive data, carries company context, works across teams, and optimizes across models. In our target market, financial institutions average fewer than 200 employees. We are already running our internal platform inside a company several times their size, and some are asking if they can use ours as their AI-powered enterprise workspace. The second is new agentic applications. With SaaS, we build apps on rented infrastructure. With AI, we build apps on models. Our clients don't just want a helper agent stitched to our SaaS application. They want an AI app built on a foundation model that solves a problem SaaS applications can't solve. An example is the combination of data analysis and workflow optimization necessary to fight fraud. Roughly 10% of our revenue today comes from AI products already in our portfolio, including fraud management. This is an extension of something that already works.
Both client priorities will shape where we invest in 2027, which we will discuss on our Q4 call early next year. In closing, since our founding 17 years ago, Alkami has proven itself as an industry innovator while driving sustainable growth and expanding profitability. Our next chapter builds on that foundation as we become the technology partner of choice for regional and community financial institutions. I will now hand the call to Cassandra to discuss our financial results.
Thank you, Alex. Third quarter, we outperformed our guidance across revenue and Adjusted EBITDA, enabling us to raise full-year targets for both metrics. Our recurring revenue model continues to drive compounding expansion throughout our customer portfolio, translating into strong margin progression while continuing to invest strategically to support durable shareholder value creation. I will start with our outlook. For the fourth quarter of 2026, we expect revenue of $139.7 - $141.2 million, representing growth of 15.6% - 16.9%, and Adjusted EBITDA of $30 - $31 million or a 21.7% margin at the midpoint. Our fourth quarter guidance reflects the timing of one-time revenue that benefited the third quarter and will not recur in the fourth. It also reflects slightly higher third-party costs based on the mix we saw in Q3. Those items are why our Q4 outlook is slightly below where some of you may have modeled it.
They do not change our view of the year, which is why we are raising our full-year outlook. For the full year, we now expect revenue of $530.5 - $532 million, representing growth of 19.6% - 19.9%. Adjusted EBITDA of $98 - $99 million, or an 18.5% margin at the midpoint, and stock-based compensation of less than 14% of revenue. Our revenue outlook reflects a heavier fourth quarter launch schedule, continued cross-sell momentum, and mid-single-digit ARPU growth. For the year, we continue to expect over 500- basis points of margin expansion driven by operating leverage and cost discipline, while we continue to invest in product innovation and AI. Alex's remarks on our future reinforce our path to Rule of 45 by 2030, which assumes gross margin approaching 70%, roughly 300- basis points of annual Adjusted EBITDA margin expansion, and stock-based compensation falling below 10% of revenue.
We will provide our full 2027 outlook with our fourth quarter results. We are sharpening our focus on profitability, and we expect that focus to drive margin expansion in 2027. AI is also changing how we build products, support our platform, and serve our customers. We expect those efficiencies to show up across our cost structure in 2027. Our 2027 plan will be built on the momentum we see today. 19% live ARR growth this quarter, a backlog of approximately $56 million of ARR that we expect to mostly go live over the next 12 months, and continued ARPU growth driven by strong cross-sell and sustained interest in our DSSP offering. Now turning to third quarter results. Revenue was $134.9 million, up 19% year-over-year, and above the high end of our guidance. Subscription revenue grew 20% and represented 96% of total revenue.
Our year-over-year growth was primarily driven by new customers going live and expansion within existing customers. Growth also benefited from an easier prior year comparison and from higher third-party and core conversion revenue than expected. We increased live ARR by 19% to $533 million. Our $56 million ARR backlog represents 31 new clients and roughly one million digital users, and the majority is expected to go live over the next 12 months. As Alex noted, the strategic review created some uncertainty for customers and prospects this year. With the review concluded, our teams are fully focused on winning new business and executing for our clients. Our strategy is increasingly centered on expanding value per client, and our results reflect continued progress.
In the third quarter, average ARR per client reached approximately $1.7 million, and we now have 56 clients at or above $2 million in ARR, up from 18 at the end of 2021. This illustrates the central premise of our long-term model. As mega banks and fintechs continue to innovate, our clients adopt more of the platform. We continue to see strong momentum with DSSP, which makes our clients more competitive and drives higher quality long-term revenue for Alkami. We exited the quarter with 323 live digital banking clients and 24.4 million registered users, an increase of 2.8 million users or 13% year-over-year. Over the past 12 months, we implemented 36 clients supporting 1.4 million digital users and existing clients increased their digital adoption by 1.5 million users.
Digital user growth has been higher than we expected this year, and our clients continue to grow faster than the overall market. This is a testament to how our product helps our clients grow. Our digital banking contracts provide strong visibility into attrition, typically several quarters in advance. Over the past three years, we have churned less than 1% of our digital banking ARR annually, usually resulting from a client merger. ARPU increased to $21.88, up 5% year-over-year. Growth was driven primarily by cross-sell execution, increased user adoption among existing clients, and the increase in live banks, which tend to onboard at higher ARPUs, given the commercial to retail mix. Remaining performance obligations were approximately $1.7 billion or 3.2x live ARR, providing strong visibility into long-term revenue. Third quarter gross margin was 63.1%, below our expectations, and driven primarily by higher third-party product costs.
Our cost of revenue includes third-party product costs, which fluctuate with product mix, hosting costs to run our platform, and the cost to implement and serve our customers. Roughly a third of our revenue is third-party products that are fully integrated into the Alkami platform, so gross margin can move with mix. We remain committed to expanding gross margin as we scale and improve the efficiency of hosting, implementation, and client support. We expect AI to contribute to those improvements over time, with the benefits building gradually. Third quarter operating expenses were $59.4 million or 44% of revenue, representing approximately 600- basis points of year-over-year improvement. That improvement was driven by personnel cost efficiencies as we achieve new levels of scale, and it was realized across all areas of operating expense.
Adjusted EBITDA was $26.3 million above the high end of our expectations, with an Adjusted EBITDA margin of 19.5% and an expansion of over 530- basis points year-over-year. Revenue grew faster than cost of revenue and operating expenses, which drove the expansion. In the third quarter, we generated operating cash flow of nearly $22 million, driven by improved profitability and disciplined working capital management. We ended the quarter with $85 million in cash and marketable securities. We also repurchased $10 million of stock for a total of $20 million through September 30, 2026. We remain committed to a disciplined and balanced approach to capital allocation across acquisitions, debt reduction, and opportunistic share repurchases funded by free cash flow. In closing, our third quarter reflects the strengths of our client base, disciplined execution, and a clear focus on profitability.
Our clients' growth is driving ours, and we are building on that momentum as we convert growth into margin and create long-term value for our shareholders. We are excited about what lies ahead. With that, operator, please open the line for questions.
Thank you. Ladies and gentlemen, we will now begin the question-and-answer session. To ask a question, you may press star followed by the number one on your telephone keypad. If you are using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star followed by the number two. With that, your first question comes from the lineup. Chris Kennedy with William Blair, please go ahead.
Good afternoon. Thanks for taking the question. New sales activity was very stable throughout the year, but Alex, you did mention some uncertainty kind of pressured new logo wins. Can you just provide a little bit more color on that and the pipeline?
One moment please, as we reconnect the speaker line. Ladies and gentlemen, please continue to stand by as we reconnect the speaker line.
Telling people that we are reconnecting.
Hi, Chris. Would you mind if you could repeat your question?
Sure. Hey guys, thanks for taking the question. New sales activity was pretty stable throughout the year, but you did mention kind of strategic uncertainty did pressure new logo wins. Can you just give a little bit more clarity as to that comment and talk about the pipeline?
Sure. This is Alex. Thanks for the question. As I mentioned in my opening comments, our pipeline remains consistent with other years. If you think about a customer decision when they are looking to switch off of a legacy digital banking platform onto a new digital banking platform, they are signing a five to seven-year contract for what you could argue is the most strategic system that they have.
If they are looking at multiple options and one of the options their company within the market there has been a for sale sign put around its neck and a lack of clarity in terms of the company's future, it is a very reasonable thing for a bank or credit union CEO to make another selection that does not have as much risk for them as the perceived risk that they would have making a selection in a company that they do not know the future ownership of that company. I know this, Chris, because this is conversations that I had with CEOs while they were making decisions, and it was unfortunate.
Understood. Okay. Thank you for that. Cassandra, you did mention ARR growth, the backlog, continued ARPU growth. Is there any way to kind of bracket potential revenue growth as we think about 2027? Thanks for taking the questions.
Yeah. No problem. Thanks for the question, Chris. I am not ready to guide for 2027 just yet. We do feel that our plan will be based on the trends that we are talking about on this call today. We have 19% growth in ARR this quarter, which was really strong. We have about $56 million in backlog that will likely go live over the next 12 months. We are still driving really good ARPU growth. So kind of the underlying and fundamental drivers of our business will hold in 2027, and that is our expectation today, and we look forward to updating you further on next quarter's call with our 2027 guidance.
Great. Thank you.
Thank you. Your next question comes from the line of Aaron Kimson with Citizens. Please go ahead.
Great. Thank you guys. It seems like there was some management turnover in Q3. Your CTO, Deep, appears to have left and had his role rolled into the Chief Product Officer's role. Chief Accounting Officer left, the Chief Customer Officer for MANTL segment and ACH Alert, who came over in the MANTL acquisition, left in the first week of October. The question I am getting from investors is whether this is having any effect on the day-to-day of the business. I would love any perspective you can offer there.
Yeah, all three of those are very unrelated. With the situation, Deep was a great hire for Alkami, and made a massive contribution to the company. If you are the CTO of a digital banking platform with 24 million users, you are an emergency room doctor 24 by 7. That was just a situation with Deep where he wanted to make a personal decision. There are two different organization models. One where you have a CPO and a CTO and one where you combine them. What I decided to do is combine those two organizations so that we could be more fully aligned in terms of customer delivery. That was an organizational design decision. But we have a bunch of talented executives, and we have a good bench when somebody like Deep leaves. I am excited about Ben taking over the combination of the two.
That's helpful. Then as a follow-up, can you talk about progress you've made with the features and functionality around your treasury management solutions and how close you are to getting that where you need to be to compete more effectively with larger banks?
Yeah. Thank you for that question. Just to recap, in the middle of 2024, we benchmarked our treasury management software against what we consider to be the two or three leading contenders if a community bank in our ICP was making a decision to buy digital banking. At that time, we determined that there were somewhere between 24 and 28 features that we wanted to build. Once again, those were not necessarily a new feature. It was a level of sophistication on a particular feature that would be necessary. For example, householding and how a community bank did their householding. We built those features with our customers. So we had a group of customers that actually sat through and reviewed every single sprint cycle to make sure that we were building them well.
The majority of those, I think with the exception of maybe 6, went into customer usage through the third quarter. We've got another 6, I think, that are in beta right now. But by the end of the year, we should be largely complete with what we consider to be the feature set that allowed us to be able to be selected and have a happy customer within our ICP, which is community banks somewhere between $500 million and $10 billion in assets.
Awesome. Thank you.
Thank you. Your next question comes from the line of Jacob Stephan with Lake Street Capital Markets. Please go ahead.
Hey, guys. Appreciate you taking the questions. Cassandra, just wanted to touch on the Q4 Adjusted EBITDA guide a little bit. I know you talked about some third-party cost pressuring gross margins a little bit, but maybe help us understand a little bit more in detail what those third-party costs are, and do you expect them to continue over the next couple quarters?
Sure. Those costs are related to about 35 different products that we fully integrate into the Alkami platform and sell to our customers. From time to time, we can see mix shifts that happen that move the margin a little bit. That did happen in Q3, and I've forecasted that to continue in Q4. In addition to that, I would just call out that we did have some one-time revenue that I had expected in Q4 that actually ended up coming in Q3. So there was a little bit of shift that way. So just a little bit of timing, but on the whole, feel really good about the year, which is why we raised our guidance for revenue and Adjusted EBITDA. On the timing of revenue, some of that is related to core conversions.
We've done a high number of those this year, and they've gone very well, which has been great. But the timing of which can be somewhat difficult to predict. So that's what drove the shift between the quarters, and that also had an impact on Adjusted EBITDA.
Okay. Got it. Maybe just on capital allocation, as we eclipse the review period here. I guess, how are you thinking about buyback pace versus delevering or if M&A is even part of that equation right now?
I would describe our posture on M&A as opportunistic. Our net leverage ratios are coming down, but we're still focused on fully integrating the MANTL business and making that a very successful part of our DSSP, so we don't want to distract from that strategy. In terms of share buybacks, it really is opportunistic when the share price is attractive, and it will be funded by excess cash flow. So, we do intend to use our authorization, and we'll do that when it's prudent.
Great. Thank you.
No problem.
Your next question comes from the line of Jeff Van Rhee with Craig-Hallum. Please go ahead.
Great, thanks. Thanks for taking the questions. A few for me. I realize you are not going to give a 2027 outlook, but you printed a 5% ARPU growth this quarter. I am curious, as you look into 2027, how do you think about factors that might accelerate that, and how do you think about factors that might decelerate it? At least give us some things to think about when we are building a plug for 2027.
Sure. I guess in terms of things that would accelerate it would really be around our new customers going live. Right now we see those customers coming on at a really high ARPU. One thing I would just point out is there can be variability quarter and quarter on new customers going live in terms of that ARPU. Just for an example, when we onboarded new customers in Q2, their ARPU was nearly double the new customers that went live in Q3. So there can be some meaningful swings on the new customer side. Growth from existing customers in ARPU is much more steady. We just see that steadily tick up quarter- on- quarter. That is really driven by our very successful cross-sell motion, and expansion of existing customers.
That is another lever that we have that would continue to drive growth and potentially accelerate it would be the pace at which our customers adopt new product, and our ability to deliver that to them.
Mm-hmm. On the inverse, just things that might be a pressure to ARPU?
A pressure to ARPU would be our success in the bank market, which has been going very well. Banks tend to onboard at a much higher ARPU. If we saw more of a shift to credit unions next year, you could see ARPU growth slow down. I think overall, we'd net be fine because they tend to have very similar ARR overall, but it could drive a change in the ARPU growth as a result. Conversely, on existing customer growth, if we saw the pace of cross-sell and product adoption slow, that would impact our ARPU growth.
Got it. Alex, on the prospects taking a pause during the strategic review, can you just revisit that? There have been on-again, off-again, at a minimum, rumors about possible sales for quite a while. When did you really see the prospects start to delay in a meaningful way? On the deals that were affected by it, were those deals pushed or have in fact most of them gone ahead, closed, and just gone with another provider?
If you think about a customer's buying cycle, it could be 9 - 12 months. The public commentary about Alkami started in December of last year. I've had many conversations with existing customers and with prospects, and I respect the questions that they're asking. The questions that they're asking are, "I'm making a long-term investment. Can you categorically tell me exactly what's going to happen?" You know better than I, as a public company, there's not a lot of commentary that you make about interest in the company or lack of interest in the company. As follow-ups with some of those customers that made a different decision, I talked to them and they said this weighed heavily on them because they have-- Look, there's three or four really good companies in this space. Alkami's got great software. We've got great people.
We've got customers that stay with us. Other companies have the same thing. When faced with making a strategic multi-year decision, like I said, a five to seven-year decision, I can empathize with the customer decision-maker, and I can understand the decision that they make when there's a set of uncertainty around the company. That's why it was so important that we issue the press release that we issued. We issued that really for our end market to understand that the board of directors made a unanimous decision that the company should pursue shareholder value as a standalone company. So those have gone away for good, Jeff. Like I said, I spent a lot of time with those decision-makers. I understand why they made the decision, and I respect it.
Mm-hmm. Yeah, that's helpful. Should we, as I'm thinking ahead in terms of new bookings then in terms of pipeline, should we expect a quarter or two of continued light new client bookings, just as you weren't maybe putting people into the funnel and working them through? I'm trying to think about how that air gap might present.
Yeah, I think it'd be irresponsible for me to forecast number of new logos. I think our job is to continue to try to sell and give you a solid update after Q4, and that will show whether it's been continued headwinds for us or we're able to compress that. When you think about it, your question has a lot of logic to it because it's basically where are customers in their buying cycle and are they early enough in their buying cycle that this can change their mind, or are they so late in their buying cycle that they've already decided? Your question has a lot of logic to it. I think it would be irresponsible for me to forecast something right now.
We just need to be heads down, close out Q4, and then we can spend time reviewing Q4 and give you a view of the 2027 opportunity set.
Yep. Okay. Fair enough. Maybe one last quick one for you, Cassandra. On the one-time items in the quarter, can you size the one-time revenue that shifted from one quarter to another? And just what was that related to?
It was about $800,000 or so. Most of it was tied to core conversion revenue. These are projects that we do on behalf of our customer when they are switching to a new core. We have to do work in our digital banking application to make that work. We have had a high number of those this year. There can be variability into when they actually get completed.
Yeah. Fair.
Jeff, when we are doing those core conversion projects, we may go through four or five different mocks because the core convert—essentially, over the weekend, they are switching out their back-end core, and we have to come up live on Sunday and Monday morning. Based upon how that testing goes, we will either extend out the work for the conversion to happen two weeks later, or we will be doing really well on the conversion, and we will decide, together with the customer, to convert two weeks earlier. We have a schedule, but because it has to go so right, just like a new logo has to go, there is some variability in terms of the date that we actually do the conversion versus the original forecast of when it was going to happen.
Yeah. Well, judging from the go-lives that you outlined in the very recent past year, it sounds like you have really found your rhythm there. I will leave it there. Thanks for answering the questions. Appreciate it.
All right. Thank you. Your next question comes from the line of Ella Smith with JPMorgan. Please go ahead.
Good evening. Thank you so much for taking my questions. First for Alex, can you please remind us how you were thinking about the database conversion that moved from this year into 2027? Are you expecting any duplicative costs into next year, and does it change your margin expansion targets in any way?
That is still something that we investigate, and we will make a decision to do that when we feel like it is the right time to do that. This year, we also pushed through. There is a new PCI compliance, PCI 4.0, and we made some significant architectural enhancements to the way that we handle PII data on behalf of our customers. That was a thing that we decided that was more important for our customers than doing the database conversion. Right now, I would not forecast you that we are or are not going to do that conversion in 2027.
Very clear. Thank you so much. For Cassandra, I was hoping to hear your philosophy surrounding buybacks. Are your buybacks meant to offset dilution, or is it purely opportunistic in the market?
I guess it's both. Yes, we're opportunistic. We're very focused on managing cash flow. We obviously have had a good year for cash flow generation, especially as our profitability has really ramped up which has been great. But we still have less than $100 million of cash, so we have to be mindful of that. In addition to it, the nice benefit of buybacks is that it does offset dilution. So it really is with those two goals in mind. I think that's the answer to your question.
Very clear. Thanks so much.
All right. Thank you. I'm showing no further questions. Oh, one moment. We do have another question coming from the line of Andrew Schmidt with KeyBanc Capital Markets. Please go ahead.
Hey, Alex. Hey, Cassandra. Thanks for sneaking me in here. Apologies, I hopped on a little bit late, but maybe if you talk through just the gross margin visibility here. I think there was a question on this, but also about the exit rate. Are we still thinking about the 65% exit rate and I guess just stability on a go-forward basis? I know some of this is database costs moving around, some other things, but just thoughts on gross margin progression going forward. Thank you.
No problem. We did have a small impact to our gross margin in Q3 with some of the third-party product mix. We have factored that into our Q4 guide, so expecting that to continue to have a small impact. Over the long term, we do expect to continue to drive gross margin expansion. For us, we really have three main categories of spend that impact our gross margin. One is our third-party product cost. We know that about a third of our revenue is driven by that line, so if there is a shift in mix, that can definitely have an impact. We largely expect that to be stable. We are not banking on huge improvement from third-party gross margin in order to hit our gross margin targets. The next largest category is our hosting cost that we use to run our platform.
We do expect to get more efficient with that over time. That is an area that we have had a lot of success in becoming efficient over the past three or four years. Lastly is around how we support and implement our customers. This is another area that we expect to get more efficient, especially with being able to leverage AI more meaningfully in how we operate those two functions. We do expect that those improvements and efficiencies will show up gradually over time.
Hey, Andrew, this is Alex. Let me give you some insight into our thinking. Let us take hosting cost. We have done a really good job over a number of years of improving our posture from a hosting cost perspective, and we have a good sense of other things that we can do to improve our hosting cost. Those are all engineering activities, right? We can say, here is a series of things that we can do to improve our hosting cost on the way to the target that Cassandra laid out, which is a long-term model of about 70% gross margin. But in any particular quarter, we are going to make decisions based upon the priority of the business.
As I said earlier on the call, we believed it was more important to make an architectural change in the way that we handle PII data on behalf of our customers than do an engineering project that might further optimize gross margin. The thinking process of the management team is we've got a view of how we get to the long-term gross margin target. We've got projects that we can do to get there, but we might make a decision in any one quarter to prioritize something that's customer-facing versus something that gets us the next point of gross margin.
Got it. Very helpful. Thank you both. Then, maybe I could just ask about win rates, and I think this was discussed earlier to some extent, but obviously I think as you know, the strategic review does tend to affect those. I'm curious now that that's concluded, this probably takes time to sort out, but have you seen any changes in terms of pipeline conversion, win rates, things like that? I'm just curious how more recent conversations progressing as we move past the strategic review. Understanding it's still a little bit early. Thank you.
Yeah, I would tell you a bright spot for us is our win rate improvement in the bank market. We set some targets for the first three quarters of the year in terms of our win rate. And we've hit those targets in the bank market. So we're very excited about the progress that we've made in the bank market, and as I said on earlier comments, I understand the decision that customers made. And I think that starts to change with some of the uncertainty getting behind us. But certainly, win rates in the bank market were a bright spot for us.
Got it. Thank you for that. If I could just ask one last product-related question. Obviously fraud, cyber, hot topic in everyone's minds. You have the ACH Alert product, which is helpful. But if you talk about maybe just any ambitions to build beyond that, you are in a good position to bring those products, whether it's through your own products, through partners. But just curious how you're thinking about the fraud opportunity. Thank you.
Yeah, thanks. I was at a customer's site yesterday. I will just share the conversation that we had with them, and these were in some of my prepared remarks. What our market is looking for is an AI product that solves a problem that because it is built on a foundation model that a SaaS product couldn't solve. One of the big challenges for our customers when they are managing fraud is the fact that the combination of the analytics to figure out what the threat vector is, and then there is usually about four to five different functions in their company that are trying to sort out what their role is in addressing this particular fraud vector.
If you think about the combination of a high degree of analytics over a pretty big surface area of data, along with dynamic workflow, that is the kind of thing that building an AI application on top of a model solves very differently than how we would build a SaaS application in the past. So that is an area that is of high degree of interest to us. It is a high degree of interest to customers. Andrew, that would be the kind of thing that we would be doing research with six build customers to understand if we can truly solve the problem, which is the combination of data analytics and dynamic workflow management.
Understood. Thank you, Alex. Appreciate all the comments.
Thank you. We have no further questions at this time. Ladies and gentlemen, this concludes today's conference call. Thank you all for joining. You may now disconnect.