Good morning, everyone, and welcome to the Angi Inc. Third Quarter 2026 Earnings Call. After today's prepared remarks, we will host a Q&A session. To ask a question, you may press star then one on your telephone keypad. To withdraw your question, please press star then two. All participants will be in listen-only mode. Should you need assistance, please signal a conference specialist by pressing the star key followed by zero. As a reminder, this conference is being recorded. Before we begin, the company would like to remind you that during this presentation, the company may make certain statements that are considered forward-looking under the federal securities law. These statements may relate to the company's outlook, strategy, and future performance, and are based on current expectations and information currently available to the company.
Actual outcomes and results may differ materially from the future results expressed or implied in these statements due to a number of risks and uncertainties, including those contained in the company's most recent quarterly report on Form 10-Q, its most recent annual report on Form 10-K, and any subsequent SEC filings. The information provided on this conference call should be considered in light of such risks. The company will also discuss certain non-GAAP measures, including Adjusted EBITDA, which they will refer to today as EBITDA for simplicity during the call. The company would also like to refer you to its earnings release, public filings with the SEC, and the investor relations section of its website for all comparable GAAP measures and full reconciliations for all non-GAAP measures. Now, it is my pleasure to turn the floor to Angi CEO, Mike Steib, and its CFO, Julie Hoarau.
Mr. Steib will begin with some introductory remarks. Please go ahead.
Thanks, Alan, and good morning, everybody. I am excited to be here. Before I begin, I want to thank Jeff for the work he has done at Angi and for his support during this transition. I also want to thank our board and our leadership team for embracing a lot of change in a very short period of time. I thought to get started, I would share a little bit about my background, why I came to Angi, and what I have learned about the business in my first few weeks, and what we see as the path to victory. I have spent much of my career building marketplaces and transforming businesses, especially in my last three roles as CEO. At The Knot, we built America's most used and highest-rated wedding planning app, and we delivered annualized shareholder returns of more than 30%. At Artsy, we built a global transactional marketplace for art.
We grew revenue aggressively, and we made an unprofitable business sustainably profitable. At TEGNA, we used technology and AI to improve productivity, to build better digital products, and deliver approximately 40% annualized shareholder returns and a $6.2 billion exit. We did that by building strong teams, great products, and operating companies that move fast and invested shareholders' money wisely. That's the playbook I'm excited to bring to Angi. Angi is one of the most recognized brands in home services, a network of almost 200,000 pros around the world, and a leading position in a nearly $2 trillion global home services market. These are advantages that would be difficult to replicate, but today, we're not getting enough out of them.
Too many homeowners come to Angi looking for help and don't end up hiring a pro, and too many pros miss opportunities to turn these leads we deliver into jobs. We can make Angi work better for both, and we can run this company far more profitably while we do so. Let me start with how I think about the business, because we've reported a lot of different metrics in the past, but we haven't always made it clear how they connect to revenue and profit. I want to change that, and I want to walk you through how this business makes money in simple terms. That will also let us tie what we're working on directly to revenue and profit in the future. Homeowners come to Angi looking for help with a project, and they tell us what they need. That is a service request.
We match that request with one or more qualified pros, and each match is a lead. Pros pay us for those leads, and we set the price of each lead based on its expected value to the pro. Pros can pay per lead within a monthly budget they set, or they can pay a subscription for an estimated number of leads each month. We call that budget or that subscription capacity. As we deliver leads, pros spend against that capacity, and thus we generate revenue. The business comes down to three things. One, generating service requests. Two, turning those service requests into valuable leads through matching and pricing that we sell to the pros. Then three, building the right amount of pro capacity to purchase those leads. We generate service requests in four ways. About a third come to us directly.
These are homeowners who search for Angi by name, visit our website or app, or respond to our emails, our calls, and texts. Others come from unbranded search, where Google sends us free traffic from people looking for, say, a plumber or a roofer. We also buy clicks through paid marketing, and we buy service requests from third-party affiliates. Finally, we win homeowners through LLM integrations and LLM advertising, a nascent but an exciting growth channel. We drive the number of leads and the value of leads through matching, connecting each service request to the right pros, and by setting the price of each lead fairly in order to balance pro retention and spend. We build capacity in three ways. Pros sign up online, our sales team signs up new pros, and we work with existing pros to increase their budgets. Here's what's been happening.
Over the last few years, Angi has seen declines in service requests and leads. Some of that has come from changes in organic search traffic, as unbranded search has fallen by about two-thirds since 2023 because of changes to the Google search experience, which have reduced traffic across the whole industry. Unbranded search now contributes only about 5% of our service requests, so that pain is largely behind us. Some of the decline has been self-imposed, as Angi has reduced affiliate marketing channels by about 80% since 2023, narrowing the number of partners we work with and focusing on channels that deliver good outcomes for pros. Affiliates now represent about 6% of service requests, so we are much less reliant on that channel than we used to be. Finally, Angi has reduced paid marketing over time, where spending wasn't earning an acceptable return.
Paid marketing remains an important and sustainable source of homeowner demand. These changes have put significant pressure on revenue, but they mean that as we roll into 2027, we do so on a much steadier base of service requests and leads with myriad opportunities to improve. On capacity, fewer service requests and leads have reduced the amount of pro capacity that we can serve. Most of our capacity now comes from more tenured pros, a much stickier client base. We retain around 80% of that capacity annually, and that means we need less new sales activity to replace capacity churn and grow the marketplace. This allows us to shift some resources from our small and medium-sized pro efforts to our large pro sales efforts.
While smaller pros are an essential foundation of the marketplace, these large pros generally have much bigger marketing budgets, and they have more resources to respond to homeowner inquiries, which is good for both capacity and lead value. We've reduced our smaller pro new business sales team by about 70% since 2023. Our sales team is now sized appropriately for the business that we have today, and once leads return to growth, we can easily dial up our investment in sales and online enroll to grow capacity. In short, the team has been managing declining search traffic, has been reducing affiliate and paid marketing activity where the economics weren't attractive, and adjusting the size of our sales organization. That has hurt revenue, but it was the right work to do to sustainably grow cash flow.
We have a healthier business to build on now, and I see clear opportunities to improve our products, run the company more profitably, and return to growth, which brings me to our near-term priorities. First, we are going to maximize cash flow. We can run this company at much better margins and at higher velocity. I've seen companies like Angi become too layered and expensive to operate. Unnecessary bureaucracy erodes focus, it adds complexity, and it frustrates talented people who are trying to build great things. Angi had become too layered, and we've already acted to address it. Yesterday, we made the sad decision to reduce our headcount by approximately 300 roles, primarily in middle management and support functions. These changes are expected to generate approximately $40 million-$50 million in annualized operating expense savings and reduce CapEx by approximately $10 million.
Some of these savings will be reinvested, but the majority will go right to the bottom line. These actions improve efficiency, but more importantly, they make Angi a more nimble and fast-moving place where talented people can do their best work. Second, we will bring greater focus and speed to how we operate. Simplifying how we look at the business, as we have done on today's call, is part of that. Removing unnecessary layers and support functions also puts our leaders closer to the work, gives talented people more responsibility, and it minimizes wasted time. We will also be putting in place a high-velocity operating system with clear objectives and key results and fast data-driven decision-making. My teams at prior companies have implemented this successfully several times, and it can really speed our return to growth. Third, we are going to turn a good product into a great product.
We are already making progress against two clear opportunities here, and AI is foundational to both. On the homeowner side of the marketplace, many homeowners visit Angi needing to get a job done, but they do not get the right pro for the job. To address this, we are building AI-powered experiences that help homeowners find and book the right pro and encourage them to return to Angi every time they need a job done. On the pro side, too often pros miss opportunities to turn leads into completed jobs. We are building AI-powered tools that help pros respond to homeowners, follow up and book appointments at much higher success rates, and I can tell you the early tests here have been very promising.
The better we convert homeowners' needs into service requests, the happier our users will be, the more business we will bring our pros, and the more profitable revenue we will book for Angi. And the more jobs we help our pros win, the more valuable our leads become. This should support better retention, more capacity, and pricing that maximizes the value for pros and for Angi. These improvements can have compounding benefits for our marketplace and unlock tons of profitable revenue. The work you have heard about previously, including our replatforming, Pro CRO, and the homeowner agent work, continues to make exciting progress. Fourth, we will supercharge the enterprise with AI. When I was recently CEO of TEGNA, we introduced AI and automation across our largest departments and identified approximately 2 million hours of annual productivity improvement opportunities. That unlocked cash flow and productivity.
It freed up people to focus on delivering better work and expanding our offering. AI capabilities have only multiplied in the months since. I am excited to build a team here of elite AI-enabled talent that can multiply their impact by managing teams of agents. Fifth, as a leader in our space, we will ensure we are plugged into the LLMs that will provide new sources of both free and paid traffic. We are already making big strides here. In AI answers, our internal data show that our share of voice is the highest in the category. Our apps are live in ChatGPT and Gemini, and we are building out ChatGPT ads as a growing and promising marketing channel. You will see another big announcement on this front later today. Now, with all that, let me talk about the financial outlook.
I'm not ready to give you a detailed financial forecast for 2027, but I expect EBITDA less CapEx to grow meaningfully in 2027, and I believe we can run Angi at structurally higher margins from now on. My first priority is going to be maximizing and growing cash flow. Our third quarter performance demonstrates what's possible. We delivered our highest quarterly Adjusted EBITDA margin since 2018, reflecting the cost actions already taken and more disciplined marketing spending compared with earlier this year. That was before the savings from our latest organizational changes. Due to efforts the team has been making to remove unprofitable or lower-quality marketing and sales, revenue comps remain challenging for the next few quarters, and the year-over-year impact will continue into 2027, even as those decisions improve the fundamental economics of our business.
That said, everything I've seen so far suggests the team's outlook for a return to growth in the back half of 2027 makes sense. In the meantime, profitability will be greater than expected. Most of the savings from our recent reorganization will flow to the bottom line, and I expect EBITDA less CapEx for 2027 to be substantially better than the $50 million floor my predecessor previously provided for this year. Going forward, I plan to provide you with regular updates on the priorities I shared today, along with clear and consistent KPIs that demonstrate the impact of our work on Angi's marketplace and financial future. I'll be back on our next earnings call with greater fidelity on the 2027 outlook. Let me close with this. I'm excited to be working with all of you and the team here at Angi.
Angi is a great brand, a large network of pros, and enormous opportunities ahead. I'm proud of how our team has responded to the changes we're making, and I'm excited about what we can accomplish together. We're going to build better products that delight homeowners and pros. We're going to run this company at better margins. We're going to return to growth as a much stronger, more valuable company. Thank you, and I'll turn it over now to Julie to discuss our financial results in more detail.
Thanks, Mike. I'll briefly review our third quarter financial results, the key driver of revenue and profitability, and close with our outlook. Third quarter revenue was $242 million, down 8.9% year-over-year. Adjusted EBITDA increased 6.3% to $42.2 million. That's our highest quarterly gross margin since 2018, as Mike mentioned. Let's start with revenue. Our marketplace revenue is driven by three factors. First, we generate homeowner service requests. Second, those service requests turn into leads through matching. Third, pro capacity. Service requests represent homeowner demand. Matching determines how many leads we can deliver against pro budgets to generate revenue, and pro capacity represents the budgets available to purchase those leads. Starting with homeowner demand, service requests declined 2.4% year-over-year to about 4 million, compared with 5.5% decline last quarter.
This improvement reflects a more stable demand, while lower affiliate and organic search traffic compared to last year continued to pressure results. Turning to leads declined 17.6% year-over-year to 4.5 million. This reflects the wind-down of our legacy advertising product and the overall reduction in pro capacity from our reduced sales organization. On pro capacity, average monthly active pros in the U.S. declined 14.4% year-over-year to approximately 101,000. That reflects the purposeful reduction in our sales organization and our shift toward large pros who have bigger budgets, convert more leads, and stay longer. Average pro capacity increased in the low double digits, partially offsetting the decline in active pros, and as a result, total pro capacity declined in the mid-single digits.
Revenue per lead increased 9.8% to $47, as we moved out of the heavily discounted legacy advertising product, and large pros became a bigger part of the mix, partially offsetting the decline in lead volume. Large pro revenue grew 20% year-over-year, while revenue from smaller pros declined 17%. Sorry, revenue from smaller pro declined by 13%. In summary, service requests are more stable. Leads per SR has declined year-over-year, partially offset by improved revenue per lead, and we're continuing to shift our pro capacity mix to large pros. Turning to profitability, Adjusted EBITDA increased, primarily reflecting lower fixed expense and more disciplined marketing spending compared to earlier this year, but before any benefits from the new cost action Mike described. Fixed expense declined $14 million, or 24.4% year-over-year, reflecting the workforce reduction announced earlier this year in Q1.
Sequentially, consumer marketing declined $13 million from 41.4% of revenue in Q2 to 37% in Q3, while fixed expense were about flat. As Mike discussed, our latest organizational changes are expected to generate approximately $40 million-$50 million in annualized operating expense saving and reduce CapEx by approximately $10 million, so that's $50 million-$60 million of total annual savings. We'll also have approximately $18 million-$30 million of one-time restructuring costs, primarily related to severance expense. We expect to reach the full annualized savings run rate by the end of the first half of 2027. These savings, they are incremental to the $70 million-$80 million that we announced earlier this year in Q1. Some of those savings will be reinvested, but the majority will fall to the bottom line and increase profitability and cash flow.
Now on to cash flow and our balance sheet. Our operating cash flow for the quarter was $28.5 million compared with $21 million last year. CapEx were $14.1 million, resulting in free cash flow of $14.4 million compared with $4.9 million last year. We ended the quarter with $202 million in cash and about $400 million in debt, with no borrowing and our revolving credit facility. Lastly, closing with the near-term outlook before we turn to questions. As Mike noted, due to efforts the team has been making to remove unprofitable or lower-quality marketing and sales efforts, revenue comp remains challenging for the next few quarters, especially with unusually high marketing expense in Q4 2025. At the same time, we do expect profitability to benefit from the cost discipline already underway.
While we're not providing detailed financial guidance for 2027 today, we expect the action we've taken will drive higher Adjusted EBITDA with stronger cash flow. We focused on improving pro capacity, increasing leads, and operating at highest quarterly gross margin. This will result in stronger and more profitable operating business. With that, we'll open it to questions.
We will now begin the question-and-answer session. To ask a question, you may press star then one on your telephone keypad. If you're using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star then two. At this time, we will pause momentarily to assemble our roster. The first question today comes from Sergio Segura of KeyBanc Capital Markets. Please go ahead.
Great. Thanks for taking the questions in the morning. Mike, appreciate the comments upfront and how you're viewing the business and breakdown of the business. I'm curious, just as you've come into Angi and developed your initial framework on strategy, how we should think about the priorities that were laid out over the past few quarters from previous management, and are the core pillars largely intact, or do you see any areas where you think you're going to make more meaningful changes?
Thanks, Sergio. The core pillars are intact. From my time on the board, the time I've spent with Jeff and with the leadership team here since I joined, I believe we're very much on the right track. Taking advantage of the power of AI agents in our homeowner agent product and in our Pro CRO product can ensure many more people who need to get jobs done, get those jobs done, and our pros can recognize most gross capacity, and it brings you pros who often have a lot more resource to put against converting leads to happy, satisfied customers.
Again, good for the marketplace and good for revenue. Finally, the work that the team has been doing and the great expertise we have built internally to plug into the LLMs and make sure that this trusted Angi brand and the value we bring can open up this new channel for service requests through LLMs is right on the money. I would just add that I bring a high-velocity operating system that insists on leadership being close to the action, making fast data-driven decisions, and the RIF that we were talking about today, while sad, I do believe for those who are staying with Angi now, sets this organization up to move with alacrity and to allow talented people to do their best work.
Understood. Maybe just a follow-up on the large pro strategy. Can you just give us a little bit more detail, go under the hood there on the traction you are seeing there and the opportunity that you see going forward?
Yeah. So, to reiterate from my script, with the SEO, unbranded SEO declines and the affiliate declines that the team has self-imposed, if you would, for the benefit of the marketplace, we needed less pro capacity to consume those service requests, which gave us the opportunity to both reduce the size of the small business sales team, but also to shift resources to larger pros. Shifting those resources to larger pros does two things for us. One, these pros come with two orders of magnitude greater budget than the average small pro. Secondly, they often have these teams and sophisticated technology that they put to work to make sure an inbound lead turns into a job. We saw better than 20% growth in the large pro segment this quarter. It is our second consecutive 20% growth quarter.
The team is scaling up with people who are skilled in enterprise sales, and I am watching those new members of the team ramp in their productivity. So, to me, it all looks very encouraging.
Understood. Thanks for the color, Mike.
Our next question comes from Eric Sheridan of Goldman Sachs. Please go ahead.
Thanks so much for taking the questions. Maybe two following up on the prepared remarks. With respect to the cost savings you mentioned, is there a way to conceptualize or quantify the mix between cost saved dropping into margin and free cash flow relative to what needs to be reinvested? Maybe the broader question would be how you think about broadly investing or reinvesting in the platform over the medium to long term. How would you characterize the difference in what you were trying to accomplish with this restructuring as opposed to what took place earlier this year? Thanks so much.
Yeah. Thank you, Eric. I will reiterate, we expect $40 million-$50 million in annualized operating expense savings, plus $10 million in reduced CapEx. I expect at least half of that to fall to the bottom line, and any reinvestment is going to be reinvestment in the strategic priorities I outlined earlier. I am looking forward to future earnings calls where I can give you more fidelity on the 2027 numbers as I get my arms all the way around the business. But for now, that is how to think about the positive financial impact of the restructuring the company. Secondly, this is the second RIF this year. The first one impacted roughly 350 roles, this one roughly 300 roles. The way that I would say this one is different is it is very much focused on my operating principles of having leaders close to the action.
The average manager should be able to manage between six and 12 people. At Angi, the number of people managed by the average manager was three. With this change, we now put more responsibility into our managers. It also allows us to take down some support functions that were out of line with our investments in the company, in the core work of building, marketing, and selling. In my experience, I've seen this work, when you put your leaders closer to the action, especially at a moment where you're bringing new technology into the organization, where you're bringing new features and capabilities in your products, it is so important and valuable to be close to the action that the restructuring that we've just done, while positive for cash flow, is most important for the velocity it can unlock in our organization.
Thank you.
The next question comes from Dan Kurnos of StoneX. Please go ahead. Mr. Kurnos, could your line be unmuted?
Yes, apologies. Mike, good morning, and different sector, but great to be working with you again.
Nice to be working with you again, Dan.
A couple for you, and look, you've given us some great color already. I think there's clearly some of the ad wind-down is going to be impacting this, and I appreciate that you're going to refocus us on the KPIs that truly matter in your view going forward, which I think has been a little jumbled in the past. I will ask two kind of questions since we have what we have today. I think they're related on this. You guys did address a little bit the pro churn. It did increase pretty significantly year-over-year and quarter-over-quarter as you've shifted more towards larger and more tenured pros.
I guess we kind of know what's driving that, but I'm sort of curious if you want to give more color there and just where you think kind of the bottom is, how much more internal decision you have to make around either reallocating resources or focusing more of those resources on large pro. Then, as kind of a byproduct, you also have leads per service requests have fallen, but you have seen that revenue per lead increase. Again, I think that's primarily large pro, but is that a trend that continues? We've had some commentary in the past from prior management about shift towards smaller jobs. The economy is kind of all over the place. So, if you could just help us think about how we should think about sort of those two trends, given your now strategic priorities, would be super helpful. Thank you.
Regarding churn, our pro churn has increased, and that's due to the launch of our online enroll product, which we started to ramp up in Q4 last year, and that's really scaled up this year. These online enroll pros, they sign up online with no upfront commitment, no upfront payment or commitment, and they have a much lower cost of acquisitions, since they do not talk to our sales team. However, since they have lower commitments and low barrier to entry, these pros are less attached to Angi and have a higher churn rate as a result. To date, we signed up thousands of those pros, and this has driven our overall reported churn rate up. If you look at pros who have signed up through our traditional sales channel, so excluding online enroll, churn is actually meaningfully down year-over-year.
Large pros, they have larger budgets. They tend to stay and increase their budgets over time with Angi when Angi works for them. In addition, we have a smaller sales organization, so that means we are acquiring less net new pros and have more tenure and a more stable pro base with a higher retention rate. We are happy with the progress we are making to improve the pro experience and drive retention. However, the measure that really matters is pro capacity to sell service requests rather than the nominal numbers of pro. Active pro could decline while capacity and revenue rise. You should really think about pro capacity.
Yeah, and I want to double underline that. Pro capacity is the number for us. It is not hard to imagine having two small pros churn out and one large pro join, and the capacity difference to be 100X to the positive. The number that really matters is capacity, not the number of pros or the churn of the total pro base.
Regarding your second question on lead per SR, lead per service request, and revenue per lead. We wound down our legacy advertising product, and that generated a lot of very heavily discounted or, in many cases, free leads. As a result, our number of leads per service request has fallen. But at the same time, revenue per lead has increased as we moved away from that product, and large pro became a higher part of the mix. We are prioritizing the economics and success of the connection, and we are not maximizing the lead volume.
Got it. Thank you both so much.
The next question comes from Stephen Ju of UBS. Please go ahead.
Hi, guys. This is Tarini on for Stephen. First, Google headwinds have been a drag on growth for the past three years. Do you guys see them continuing to have an impact on the business? As a follow-up, because of those headwinds, Angi's stock is cheap, and other companies in your sector are probably cheap too, and you guys have some debt coming due. You could buy back stock, and from an M&A standpoint, you could be a buyer or a seller. With your increased cash flow, how are you guys thinking about capital allocation? Thanks.
Thank you. On the first question, SEO declines, I think, have been the story for consumer internet broadly in the last couple of years. For Angi, unbranded SEO is 5% of total service requests. I think that story is largely behind us. I will note that the advent of the new LLM channels is something that we're really excited about. It's the first what seems to be a significant potential source of new service requests going forward, and our team has us well-positioned for that channel. On capital allocation, I would say, thank you for noting that the stock is cheap. Part of the reason I'm here is that I agree with you at the end. The optionality and opportunity here are substantial.
My first priority as we think about capital allocation is the 2028 notes, and I'm glad that we're going to be approaching that from a position of strength now with improved margins in the business. As for the rest of our capital allocation strategy, the core principle here is this is not our capital, it is our shareholders' capital, and it's our job to deploy it where it creates the most value for our shareholders. All of the levers that you all know well will have our attention, and we will approach this in a way that maximizes shareholder value. Hopefully my track record gives you all confidence. I have a good record of creating value for my shareholders.
Great. Thank you.
Our next question comes from Brad Erickson of RBC Capital Markets. Please go ahead.
Hi. Good morning. This is Khadija on for Brad. Two questions for me, if I may. The first one is, what progress have you made on Pro CRO since last quarter? Are there any updates you can give on that front? Then secondly, can you dig a little more into what drove the Adjusted EBITDA outperformance for Q3? Thank you.
I'll take the Pro CRO one, and if you'll indulge me, I've had a chance to get to know a bunch of our pros in my time here. What I can share with you is that these large pros who are putting up roofs in 20 or 30 markets, or fences or siding, have built these organizations that often look like a technology company that happens to install roofs. They have sophisticated AI that warms leads and helps to convert people to appointments. They have live operators standing by to get their hands on a lead as soon as it's available, and they're very good at turning someone who needs a fence or needs their lawn mowed into an appointment and ultimately into a booking.
The average small pro, whether it is one plumber or electrician in a truck or a small outfit, tend to be spending all of their time and energy on doing the work. They do not have often the same sophisticated capabilities that maybe a private equity-backed or just large multi-state operation can. What we are doing with Pro CRO is giving the small pro the power of a large pro in receiving a lead and turning it into a job that they can do well. What we have seen so far on our Pro CRO work is the replatforming that is foundational to it, that Jeff talked about in our previous calls, has continued on schedule and on budget. The early Pro CRO work that we have in the field started with a small group of test pros. We now have about 1,000 pros in that test.
We launched the alpha migration of the replatforming on the pro side on schedule earlier this week, all of which is giving us a lot of confidence in our ability to build these features. In the test group, and it is still early, we are seeing that the conversion rate to an appointment is in line with the conversion rates that you would expect when a human reaches out and talks to the homeowner. That is very encouraging for what it can mean longer term for the value of our leads. It is early, but this is the kind of work I was excited to come here and dig into with the team, and what they have achieved so far gets me very excited.
Yeah. Regarding the Q3 Adjusted EBITDA performance, two items contributed to the outperformance. First, we had lower fixed costs reflecting the workforce reduction we announced in the first quarter. Second, we had more disciplined marketing. As we said last quarter, we pulled back on TV and other offline spend where the returns were not there, and we shifted dollars to higher return channels like paid social. As a result, consumer marketing was down from 41% of revenue in Q2 to 37% in Q3. That came before the action we just announced. The new savings are on top of that. On whether this carries forward, we do think that the structural pieces do carry forward over time, and we believe that this should be a structurally higher margin business prospectively.
The lower fixed cost base is permanent, and we will continue to run marketing to a breakeven return on the last dollar. The majority of the new savings should reach the bottom line. However, I would caution against extrapolating the Q3 margin quarter by quarter. Marketing is variable and will adjust to returns. In addition, regarding Q4, seasonality is softer, and we do have a difficult revenue comparison against the higher marketing Q4 last year. As we said on the Q2 call, Q4 usually comes down a bit. In summary, quarterly results can still vary with seasonality, channel mix, and investment timing. Our goal is a sustainably more efficient cost structure, not by simply maximizing one quarter margin, but we do expect the cost action to drive higher Adjusted EBITDA and stronger cash flow over time.
Great. Thanks for the color.
The next question comes from Bryan Smilek of JP Morgan. Please go ahead.
Yeah. Hey, guys. This is Nishant for Bryan. Thanks for taking my question. I guess you kind of touched on it, but just looking at the tech replatforming as a kind of key milestone to getting more agentic features to market, could you talk a little more about the status of the project and maybe what you kind of expect from this product in the future? And then second, real quick on accounting, I guess it seemed like the SBC depreciation and tax rate kind of had a notable step up. If you can share any color on that, would be helpful. Thanks.
Yeah. So, I believe everyone has this background, but the tech replatforming is bringing together the technology stack that drives our European marketplaces and the technology stack that drives our U.S. marketplace, I say international and U.S. And there are advantages to each, and bringing them together should give us the best of both and an easier platform to manage and maintain, and a platform on which we can innovate and ship features a lot faster. The progress has been good here so far. I'm really proud of the work I've seen in my short time with the company from Kris Boon and the rest of the organization. Tech replatforming are hard, and these guys are on it, and they have been hitting their milestones so far. The homeowner side, we've seen a large portion now of the homeowners and revenue moved to the new platform.
On the pro side, the alpha product with a handful of pros launched earlier this week. We'll continue to ramp these up incrementally. We want to test and make sure it has the right impact on the marketplace performance before you scale it all the way. I've got a lot of confidence in this team, and I'm optimistic that this will continue to click along on schedule.
Regarding your accounting questions, taking this in two pieces. On stock-based compensation, it was higher in Q3, but that was due to a one-time acceleration of stock-based compensation expense related to the outgoing CEO separation agreement terms. So it's really a one-time item. Depreciation expense that has been running a little higher for the last few quarters. This is linked to our one platform work. We've been capitalizing more internally developed software costs, which we deprecate over two years. So that's showing up in a higher depreciation as we place the new platform into service. We also have some assets that are being retired related to one platform and also resulted in some accelerated depreciations. Last on tax, as pre-tax income decrease and is closer to breakeven, the effective tax rate movements are magnified.
In Q3 specifically, the rate is elevated because the Q3 pre-tax income, excluding discrete item, was significantly higher than Q1 and Q2. So a larger share of the year's tax expense was recognized in Q3, including a catch-up from applying the higher effective tax rate to prior quarters. Year to date, the rate is negative, and that's not meaningful because of the Q2 goodwill impairment, which is largely a non-deductible item. So the pre-tax loss generated little tax benefits.
Super helpful. Thank you.
Our next question comes from Matt Condon of Citizens Bank. Please go ahead.
Thank you for taking my questions. Michael, welcome aboard. You've addressed this a lot on the call, but you've expressed a lot of confidence in Angi's ability to improve EBITDA and cash flow in the near term. As you evaluate the business, can you just talk about how you balance that? Obviously, a lot of that's going to flow through to the bottom line, but how do you balance that against investments that are necessary to return the business to sustainable growth? Then can you also just talk about your AI strategy more broadly across products, but also in the enterprise as well? Thank you so much.
Thank you, Matthew. The reductions that we described today, we did not reach that number by saying we have to get to a number of cuts. We said, "What is the right organization and the right structure of that organization to deliver on the strategic priorities I laid out for you before?" So that's the way that we think about the balance of investment. The balance is return to growth, execute your priorities, and don't spend a dollar that's not necessary to build, market, and sell that product and deliver profit for our shareholders. Second thing I would just note for you is some companies will stretch their sales and marketing spend from one quarter to the next to hit a revenue or a growth target.
I just want to be transparent with you all that my focus is always going to be on what maximizes the present value of future cash flows. That means we're not going to approach quarter to quarter with a target on what percentage of revenue, for example, we'll spend on marketing. We're going to look at the cost of acquiring service requests, and we'll do it at the point that maximizes cash flow today and over the long term. While it was important to get this restructuring of the organization done and behind us so we can get cracking with the right organization on our priorities, I know better than anyone you can't cut your way to growth.
The growth is going to come from us executing on our agentic future in our products, our agentic future in our enterprise, our agentic future in our partnerships with LLMs, and the shift of the composition of our revenue to accommodate more and more of the large pro capacity. Matthew, your second question was about the AI strategy. Let me touch on it in three parts, each of which you've sort of heard through today's conversation. The first is, when a homeowner comes to us, she has a problem that needs to be solved, and the problem's not always simple. It's not always as simple as, "I need a plumber at zip code 10001," right? People want to describe the problem. They want to describe when they're available, the kind of pro that they think they need.
It could be something more complex like a kitchen renovation or an addition to the home. All of these things are underserved by the historical sort of user experience that people run into on the web or on an app. An agentic partner to that homeowner, however, can much better come to understand their needs and help to match them to the right solution on our marketplace. As I discussed on the pro side, having an agent there, especially for smaller pros who can step in and work with that homeowner and the pro to get you to that magical moment of an appointment, both I think will create a lot of incremental value for the participants in our marketplace and therefore for the marketplace. The second part, though, is bringing AI into the enterprise. I have seen firsthand the power of this.
When you move from being a company that doesn't really use AI, which is where most companies truly are today, to a company where people use AI to support their work, which is probably where Angi sits today, to becoming AI supercharged people who manage teams of agents and who not only use those agents to help them do the work, but who actually teach the agents how to do the work with them. We're already starting to see the results from that in our engineering organization, although it's still early. We've got a roadmap for the ways that AI can take the busy work, the grunt work off the plates of our people and free them up to do the really high impact selling, marketing, and building that's going to drive our growth in the future.
Last part I want to sort of leave us with is these LLMs, in a lot of ways, to me, remind me of the early days when we were all figuring out Google. They would become a source, Google was going to become a source of free traffic and was going to become a source of paid traffic, and the question was, how do you maximize your opportunity there? One way was have a strong and trusted brand that people recognize, and Angi's very fortunate to have that. The second was to have your team very focused on it and to be building ahead of the growth in that platform, and that's what you see Angi doing today. Thirdly, there are technical capabilities.
In the same way that we were building technical capabilities for SEO and SEM 2 and a half decades ago, Angi has built the technical capabilities in-house to know how to partner with and embed with these LLMs to set us up for longer term success as the average user starts to become the power user in an agentic world. I think we're putting ourselves in the right position there, and we'll have another announcement later today that just shows the drumbeat of the work that we're doing here.
This concludes the question and answer session. I would like to turn the conference back over to Mr. Steib for any closing remarks.
Well, I just want to thank you all for your interest in Angi, and the questions were really helpful, and thanks for the warm welcome. The team here is doing an incredible.
The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.