Zeta Global (ZETA) Q4 2024 Earnings Call Transcript

Image source: The Motley Fool.

Zeta Global (NYSE: ZETA)
Q4 2024 Earnings Call
Feb 25, 2025, 4:30 p.m. ET

Contents:

• Prepared Remarks
• Questions and Answers
• Call Participants

Prepared Remarks:

Operator

Greetings, and welcome to the Zeta fourth-quarter 2024 earnings conference call. [Operator instructions] As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Matt Pfau, Senior Vice President of Investor Relations. Thank you, sir.

You may begin.

Matt Pfau -- Senior Vice President, Investor Relations

Thank you, operator. Hello, everyone, and thank you for joining us for Zeta's fourth-quarter and full-year 2024 conference call. Today's presentation and earnings release are available on Zeta's Investor Relations website at investors.zetaglobal.com, where you will also find links to our SEC filings, along with other information about Zeta. Joining me today on the call are David Steinberg, Zeta's co-founder, chairman, and chief executive officer; and Chris Greiner, Zeta's chief financial officer.

Before we begin, I'd like to remind everyone that statements made on this call as well as in the presentation and earnings release contain forward-looking statements regarding our financial outlook, business plans and objectives and other future events and developments, including statements about the market potential of our products, potential competition, revenues of our products, and our goals and strategies. These statements are subject to risks and uncertainties that may cause actual results to differ materially from those projected. These risks and uncertainties include those described in the company's earnings release and other filings with the SEC and speak only as of today's date. In addition, our discussion today will include references to certain supplemental non-GAAP financial measures, which should be considered in addition to and not as a substitute for our GAAP results.

Don’t miss this second chance at a potentially lucrative opportunity

Ever feel like you missed the boat in buying the most successful stocks? Then you’ll want to hear this.

On rare occasions, our expert team of analysts issues a “Double Down” stock recommendation for companies that they think are about to pop. If you’re worried you’ve already missed your chance to invest, now is the best time to buy before it’s too late. And the numbers speak for themselves:

• Nvidia: if you invested $1,000 when we doubled down in 2009, you’d have $328,354!*
• Apple: if you invested $1,000 when we doubled down in 2008, you’d have $46,837!*
• Netflix: if you invested $1,000 when we doubled down in 2004, you’d have $527,017!*

Right now, we’re issuing “Double Down” alerts for three incredible companies, and there may not be another chance like this anytime soon.

Learn more »

*Stock Advisor returns as of February 24, 2025

We use these non-GAAP measures in managing our business and believe they provide useful information for our investors. Reconciliations of the non-GAAP measures to the corresponding GAAP measures, where appropriate, can be found in the earnings presentation available on our website as well as our earnings release and our other filings with the SEC. With that, I will now turn the call over to David.

David Steinberg -- Co-Founder and Chief Executive Officer

Thank you, Matt. Good afternoon, everyone, and thank you for joining us today. 2024 was a record year for Zeta, capped by a strong fourth quarter, where we once again exceeded expectations. Marketing is at the front lines of the AI revolution, driving an unprecedented replacement cycle across the marketing technology ecosystem.

At Zeta, we've consistently skated to where the puck is going. Our early investments in AI and first-party data are resonating with customers and prospects, fueling our record fourth-quarter results, and contributing to our market share gains. We believe these investments will propel us to over $2 billion in annual revenue by 2028 as outlined in our Zeta 2028 plan. In the fourth quarter of 2024, we generated record revenue of $315 million, up 50% year over year with record adjusted EBITDA of $70 million, up 57% year over year, both well ahead of our guidance.

For 2025, we are guiding to our sixth consecutive year of 20-plus percent growth, alongside another year of free cash flow margin expansion. Our Zeta 2028 plan forecasts a similar trajectory, 20% organic revenue CAGR between 2024 and 2028, adjusted EBITDA margin improvement of 580 basis points to at least a 25% margin and free cash flow margin expansion of 700 basis points to at least 16% over this time frame. 2024 was an exceptional year for Zeta, and as we enter 2025, our momentum is accelerating. The business has never been stronger, and the opportunity has never been bigger.

Our vision of an all-in-one marketing platform with AI and data at its core is resonating more than ever. While most AI products either drive efficiency or drive revenue gains, we believe Zeta's AI does both, creating a powerful competitive advantage and delivering measurable impact. In Q4, the dollar value of our RFPs reached a record high, up close to 40% year over year with the total pipeline growth of almost 60%. Our U.S.

NPS score increased 8 points year over year to 55. We saw significant traction with our One Zeta initiative, which focuses on expanding customer use cases, extending our solutions to become more indispensable. A prime example is a Fortune 500 retailer who is an existing retention-based CDP client that added a customer acquisition use case, which should more than double their investment with us. Now let's talk about AI and agentic AI adoption across our platform.

As a reminder, our GenAI products empower customers in 3 ways: productivity, automating marketing tasks with intelligent AI agents; personalization, enhancing customer experiences with AI-driven audience insights; and precision, enabling marketers to interact conversationally with Zeta's marketing platform for faster, more complete answers. But AI isn't just a tool. It's an extension of the marketing team. Our agentic AI performs specialized roles within an enterprise, seamlessly integrating into workflows to drive efficiency and superior performance.

For example, operations specialist agents. We are working with multiple financial institutions to automate their quality assurance processes for marketing campaigns. These AI-driven QA agents replace a manual 10-step process, integrating with QA systems of record to track completion, agent ownership and results, ensuring compliance and improving efficiency. Your own virtual data scientist agents.

Clients can bring their own predictive models into our platform, but aligning those models with ingested events and data structures can be complex. Our agents automate the data mapping process, ensuring seamless integration by intelligently matching fields, normalizing formats, and resolving discrepancies. This removes manual effort and accelerates time to value for clients. Creative agents.

Our Visual Composer acts like a world-class creative director, helping marketers to go from a blank slate to a fully designed campaign in minutes. It leverages existing templates and uses AI to generate images, body content, subject lines, and content blocks, reducing iteration cycles and accelerating campaign deployment. While for some companies, AI is just a press release or an add-on, for Zeta, it is foundational to our platform. We have been investing in AI for more than 7 years, not 7 months, and we have the results to prove it.

In 2024, AI adoption across our platform surged. 126 brands adopted our Data Cloud AI within its first year of launch. AI agent usage grew nearly 200% sequentially in Q4 as more enterprises embedded automation. Over 1,000 behavioral taxonomies were created using AI alone, doubling the number of taxonomies in our platform.

We are truly shaping the future of AI-powered marketing. Rather than charging separate fees, we monetize our AI products through increased consumption, accelerating adoption across our customer base. Although we are still early in AI adoption across our platform, we saw a meaningful impact to our consumption revenue, which increased over 40% in 2024, a significant acceleration from 2023. And unlike other software companies, we do not need to pivot our business model to monetize our AI innovations.

We are already doing so as evidenced by our 2024 results. Zeta's leadership in AI continues to attract top-tier talent. Pam Lord recently joined us as President of CRM from Oracle, where she ran their B2B and B2C marketing cloud businesses. Ed See, our new Chief Growth Officer, was previously a partner at McKinsey & Company's marketing and sales practice.

We also continue to enhance both our team and our platform. We are happy to report that we completed the LiveIntent integration in early January, well ahead of our schedule, and it's already delivering incremental value to our customers. We have already launched our first synergistic product from the acquisition, Zeta Direct. This solution combines LiveIntent's premium publisher network with the Zeta Data Cloud, enabling marketers to deliver personalized ads directly within newsletter emails.

This innovation enhances publisher monetization while driving a higher ROI for marketers and Zeta, a true win-win-win. As I reflect on nearly 4 years as a public company, Zeta's trajectory has never been clearer. We have beat and raised guidance for 14 consecutive quarters, outperformed our Zeta 2025 plan a year ahead of schedule and strengthened our AI data leadership. We remain confident in our ability to be a Rule of 40 business for many years to come.

We are truly building a one-of-a-kind company. As always, I would sincerely like to thank our customers, our partners, team Zeta and all of our shareholders for their ongoing support of our vision. Now let me turn it over to Chris to discuss our results in greater detail. Chris?

Chris Greiner -- Chief Financial Officer

Thank you, David, and good afternoon, everyone. We have a lot of exciting information to share on our 2024 results and Zeta 2028 plan. But before I get into the details, I wanna lead with 3 main themes: first, consistency. Zeta has been incredibly consistent, beating and raising guidance for 14 consecutive quarters and increasing revenue 20% or greater while also expanding our free cash flow margin for 4 straight years as a public company; second, momentum.

Our investment in an all-in-one marketing platform with AI and data at its core is creating accelerating momentum in our business, driving our fourth-quarter record results, and positioning us to target continuing to increase revenue at a 20% organic compound annual growth rate over the next 4 years; and third, rarity. There are over 500 public U.S. technology companies. Of those, only 23 are expected to increase revenue 20% or greater annually from 2021 to 2025.

Of those 23, only 8 are expected to also expand their free cash flow margin annually from 2021 to 2025. Zeta is 1 of those 8, next to other great companies like Cloudflare, GitLab and Samsara included in this list. And as you will see with our Zeta 2028 plan, we expect to continue to do this for the next 4 years. Let's get into the fourth-quarter and full-year results.

In 4Q, we delivered revenue of $315 million, up 50% year to year or 31% excluding LiveIntent and political candidate revenue. The full-year revenue was just above $1 billion, up 38% year over year or 30% excluding LiveIntent and political candidate revenue. This exceeded our initial 2024 guide of $875 million by $131 million or 15%. Total scaled customer count grew to 527 as of December 31, 2024, up 17% year over year and 52 sequentially.

LiveIntent added 34 customers to our scaled customer count, and excluding this contribution, our scaled customer count increased 9% year over year. Super-scaled customers of 148 as of year-end were up 13% year over year and up 4% sequentially. LiveIntent added 3 customers to our super-scaled customer count. And as a reminder, we count each customer spending at least $1 million with us over the trailing 12 months as 1 super-scaled customer regardless of how many brands they are using us for.

The number of brands spending at least $1 million with us over the trailing 12 months increased 28% year over year. Although customers using us for multiple brands does not benefit customer count, it does have a positive impact on ARPU. Scalable customer quarterly ARPU of $577,000 increased 27% year over year as reported and 32% when removing the impact of LiveIntent. Net revenue retention for the year was 114%, at the high end of our 110% to 115% range, an increase from 111% in 2023 and our highest level as a public company.

From an industry perspective, in 2024, 7 of our top 10 industries grew faster than 20% year over year with automotive, consumer and retail, insurance, political and advocacy, and technology and media growing the fastest. We ended the year with 180 quota carriers, an increase of 32% year over year, partly driven by our LiveIntent acquisition. Excluding LiveIntent, quota-carrying reps increased 20% year over year. Our direct mix in the fourth quarter climbed to 74%, up from 70% in the third quarter and slightly higher than 73% in the fourth quarter of 2023.

For the full year, our direct revenue mix was 70%. Our GAAP cost of revenue in the quarter was 40%, a 20 basis point improvement from the fourth quarter of 2023 and up 60 basis points from the third quarter of 2024. For the full year, GAAP cost of revenue was 39.7%, up 210 basis points from 2023, mostly driven by a higher mix of integrated revenue due to our agencies initially adopting the social channel prior to increasing spend to direct channels over time. Leverage in other areas of our operating expenses resulted in our 16th straight quarter of expanding adjusted EBITDA margins year over year.

In the fourth quarter, we generated $70.4 million of adjusted EBITDA at a margin of 22.4%, 110 basis points higher year over year and $4.5 million better than the midpoint of our guidance. We exceeded our adjusted EBITDA guidance despite incurring $2 million of additional expenses related to defending against the short seller report. To this point, our Audit Committee oversaw a review of the allegations, which involved engaging an independent forensic accounting firm to evaluate our shared customer and vendor accounting practices and internal controls. Additionally, we hired a leading data and privacy firm to assess our data and privacy practices.

The reviews corroborated that our accounting practices were consistent with U.S. GAAP and that the data and privacy allegations in the short seller report were without merit. Additionally, the findings reinforced the strength of Zeta's internal controls and data privacy practices. For 2024, adjusted EBITDA was $193 million, representing a margin of 19.2% and a 49% increase year over year.

In the fourth quarter, we achieved positive GAAP net income for the first time as a public company. Our fourth-quarter GAAP net income was $15.2 million, which translated to earnings per diluted share of $0.06 in the quarter and a loss of $0.38 per share for the full year. Finally, fourth-quarter cash from operating activities was $44 million, up 62% year to year with free cash flow of $32 million, up 74% and representing a margin of 10%. This translated to a free cash flow to adjusted EBITDA ratio of 45%.

It's worth noting this includes a $22 million working capital headwind driven by growth with agencies and the industry's longer payment cycles. Absent this, cash flow conversion would have been 76%. This dynamic can be seen on Slide 26 in our earnings supplemental. For 2024, our free cash flow was $92 million at a margin of 9.2% and up 69% year over year.

Now let's get into the details of our Zeta 2028 plan and 2025 guidance. I'll start with the Zeta 2028 plan, as many of the growth drivers and margin levers we will discuss for our medium-term plan are also applicable to 2025. Slides 15 through 20 in our earnings supplemental provide additional details. For revenue, we're targeting over $2.1 billion, which equates to at least a 20% organic revenue CAGR over the next 4 years.

To put this in perspective, our revenue CAGR between 2021 and 2024 was 30%. Importantly, there's more than enough runway in our core business to achieve this target as we estimate we only have about 1% of our existing customers' marketing and advertising spend, and there's room to increase this penetration to 5% to 15% or more over time. We also have multiple new growth levers, some of which include creating new GenAI capabilities. More GenAI features should drive additional usage of our platform, which we monetize through our consumption revenue as highlighted by David earlier.

Leveraging the Publisher Cloud. This provides us with an opportunity to better monetize our extensive publisher relationships. Introducing new channels. We're still very early in mobile and plan to continue to enhance the product while also introducing new channels.

Extending our vertical expertise. Through vertical-specific functionality, we can better penetrate verticals we're underrepresented today, such as CPG, healthcare, commerce, and travel. And expanding our partnership ecosystem. We believe this will drive pipeline growth and be beneficial for margins.

The KPIs to achieve our 2028 plan look very similar to our previous midterm plan. We're targeting scaled customer count growth of 4% to 8% versus our 3-year CAGR of 14%. The biggest factor influencing this metric is our agency business, where we expect agencies to continue to add brands to our platform. However, regardless of how many brands and agency is leveraging Zeta for, it's only counted as one customer.

From a brand count perspective, our growth will likely be around our historical scaled count trend. For scaled customer ARPU, we're expecting growth of 12% to 16%, in line with our 3-year CAGR of 15%. Our One Zeta initiative should positively impact channel and use case expansion, aiding ARPU growth. Further, the aforementioned agency dynamic will also have a positive impact on ARPU.

We continue to expect net revenue retention to be in the range of 110% to 115%, in line with the 111% to 114% range that we achieved since 2021. And we expect our direct revenue mix to be 70% to 75%, roughly in line with the 70% to 77% annual range we've been in since 2021. Much like our initial medium-term plan, we're targeting significant margin improvement with our Zeta 2028 plan. Our 2028 adjusted EBITDA target of at least $525 million implies a 25% margin, an increase of 580 basis points or an average of 145 basis points per year.

We expect to leverage across all areas of our business. Cost of revenue should improve by 100 to 300 basis points. We believe this will mostly be driven by a higher mix of direct revenue, which carries a lower cost of revenue as compared to integrated revenue. This should be driven by agencies ramping usage of direct channels, our One Zeta initiative driving more adoption of direct channels, and higher margins from GenAI products and consumption scaling faster over time.

Across our other operating expense lines, we anticipate 280 to 480 basis points of margin improvement. One key point of leverage for Zeta has been growing headcount significantly slower than revenue and adding a higher percentage of headcount outside of the U.S. Over the past 3 years, our revenue has increased at a CAGR of 30%, while total headcount has only grown at a CAGR of 15% with U.S. headcount growth even slower.

Implementing AI internally should enable us to gain further headcount efficiencies. Expanding our partnership ecosystem will also create a margin tailwind. We expect future partners to take on some of the professional services we provide customers today, taking additional costs out of our business. From a free cash flow perspective, we're targeting $340 million plus in 2028, which equates to a CAGR of 39% from 2024.

This represents a margin of 16%, an improvement of 700 basis points from 2024. In addition to the aforementioned margin levers, there are 2 additional factors driving our free cash flow margin improvement. First, we expect capex to grow materially slower than revenue. Our capital expenditures and capitalized software development was 4.2% of revenue in 2024, a significant improvement from 5.8% in 2021.

Second, the impact of agencies on free cash flow should lessen over time as that business' growth comes more in line with Zeta's overall growth. Moving on to 2025 guidance. We're guiding to the midpoint of full-year 2025 revenue to be $1.24 billion or growth of 23%. This includes LiveIntent revenue of $96 million.

Adjusting for the impact of LiveIntent and political candidate revenue in the year-to-year comps, our excluding LiveIntent and political growth rate is 21%. A bridge is provided on Slide 25 in the earnings supplemental. As expected, our 2025 revenue guidance assumes there is no political candidate revenue. However, we are modeling for advocacy revenue to be between $20 million and $25 million as compared to $36 million in 2024 and $13 million in 2023.

We're guiding to adjusted EBITDA at the midpoint of full-year 2025 to be $256.5 million or a margin of 20.7%, a 150 basis point expansion year over year. We're guiding full-year free cash flow to be $129.5 million at the midpoint, representing a margin of 10.4%, a 120 basis point improvement year over year and growth of 40%. Finally, on Slide 24 in the earnings supplemental, we've included quarterly 2025 guidance for revenue and adjusted EBITDA, a practice consistent with prior years. I'd like to conclude with this reflection.

Anyone can issue a multiyear plan, but executing against it is another story. At Zeta, not only are we set to materially surpass our original Zeta 2025 plan, but we exceeded one of the key targets, revenue, an entire year ahead of schedule. For our Zeta 2028 plan, we kept the same KPIs and fine-tuned them to the most realistic path to achieve our goals. Just like with our original 2025 plan, we put extensive thought and diligence into our Zeta 2028 plan, which the entire company is focused on executing against and accountable for.

With that, let me hand the call back over to the operator for David and me to take your questions. Operator?

Questions & Answers:

Operator

Thank you. At this time, we will be conducting a question-and-answer session. [Operator instructions] Our first question is from Matt Swanson with RBC. Please proceed with your question.

Matt Swanson -- Analyst

Yeah. Thank you guys so much for taking my question, and congratulations on the quarter. Chris, you covered a lot of ground there going through those three sets of guidance. Maybe pulling in the zoom lens a little bit.

Can you just talk a little bit about what you're kind of thinking through from a macro environment, maybe a more like demand-centric spending environment for both Q1 and for 2025?

Chris Greiner -- Chief Financial Officer

Yes. Look, our approach and the simplest way to put it is we've followed how we've done it historically, which means that it requires the low end of each of our KPIs to get to our guide, which gives us the same level of flexibility we've gone into future years in terms of our guide as well as future long-term models. Given where we are in the first 2 months of the quarter, we've got very good visibility into obviously Q1. We've leveraged our Zeta economic insight data to be able to anticipate macro trends.

As I think about the verticals that have contributed to the strong growth we saw in 2024, I would expect those to be on the higher end of performers as well into 2025. And we're obviously very aware of the macro. We're aware of the inflationary backdrop. We're aware of tariffs.

We're aware of other items, but we feel like we've put appropriate conservatism into our guide to account for that. David, anything you'd add?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. And Matt, let me just say that we're not seeing any challenges from any clients at this point. So we're feeling very good about where we are versus the macro environment as it exists today.

Matt Swanson -- Analyst

I appreciate that. And then, David, great to hear that you already have LiveIntent integrated as quickly as January. Can you just talk a little bit more about kind of what you're hearing from customers in terms of their initial use cases and initial adoption? And then to broaden it out, in the 2028 plan, we have some new verticals and new products that are some of those other up drivers. What do you see from when you launch new products that gives you confidence in that long-term ARPU expansion just in customers' desire, I guess, to take more from Zeta?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. And to be clear, we've been experiencing that type of ARPU growth. So we don't need to grow our ARPU growth as a percentage to hit this plan. We can just continue to do what we're doing as we did to get to the 2025 plan from a revenue perspective by the end of 2024.

But Zeta Direct is a really cool product, Matt. What we did was we took their inventory, which embeds into newsletters, which was traditionally served based on the content the consumer was consuming. So you're getting this newsletter from this publisher, might be a daily update from The New York Times. It might be an automotive update from The Washington Post or so on and so forth.

We added our Data Cloud on top of that, and we can now directly target to a deterministic individual in addition to the content they're consuming. That has already shown an increased return on investment for the marketer and higher revenue to the publisher with us sitting in the middle of the transaction. So we garner higher revenue as a percentage of that. So it really is a win-win-win, and I think it's a perfect example of taking 1 plus 1 and equaling 4.

We, once again, Matt, feel very good about where we are with our existing customers. And if you look at the agency business, which has been among our fastest growing, their ARPUs are going up dramatically because they keep adding more brands, and it just counts as one super-scaled customer.

Operator

Our next question comes from DJ Hynes with Canaccord Genuity. Please proceed with your question.

DJ Hynes -- Analyst

Hey. Good evening, guys. Nice set of results. David, maybe I could have you expound on that last point, which is the agency business.

Clearly, the ARPU gains suggest there's lots of momentum there. Maybe just talk about what you're hearing from those folks? What kind of visibility you have into the pipelines that they have for pulling in new brands? How are you thinking about the growth opportunity with the agencies for '25 and beyond?

David Steinberg -- Co-Founder and Chief Executive Officer

Thank you so much, DJ. I appreciate the compliment. What I would tell you is that right now, one of the reasons I think we are drinking out of the fire hose in our agency business is because something that most people don't, I think, understand is we are the most profitable partner on average that the agencies work with bar none because most of our competitors who have had some challenges over the last number of months, they're charging a meaningful upcharge to data expense. We bundle the data in as a part of the activation and are generally lowering the cost to the agency or increasing their profit depending on how they book it by 25%.

So what we're seeing is the agencies are moving more and more brands and more and more volume to us. What we're seeing out of agreements that have already been executed as minimum amounts for this year gives us a tremendous amount of comfort in the projections that we're putting out there.

DJ Hynes -- Analyst

Yes. Makes sense. Maybe a follow-up to that. It was interesting, one of the agents you called out was kind of a push into creatives.

And I'm curious, like how deep does Zeta wanna go into creative? Does that create any channel conflict with these agency customers? Are they asking you to do that? Just talk about the opportunity there and whether that's meaningful to Zeta.

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. So let me be clear. We do not wanna compete with the agencies as it relates to creative. That is not in our road map.

What we are doing is taking their creative and optimizing it for any screen size, would be a perfect example. So you're opening it on an iPhone, an Android device, an iPad, a laptop, a TV commercial. The dynamic agent is able to optimize that creative for that. Now we have some enterprise clients that are using very basic creative tools versus they might not have an agency.

But any time we're working with an agency, and we are not putting out as a stand-alone product a creative product, we very much wanna support the creative assets of the agencies, and we wanna utilize our tools to help optimize that creative, look at what has the highest conversion rates by creative and the best return on investment for the client.

Operator

Our next question comes from Jason Kreyer with Craig-Hallum. Please proceed with your question.

Jason Kreyer -- Analyst

Hey. Thanks, guys, and congrats on another strong quarter. So look, you guys are working with some of the largest marketing companies. You talked about 44% of the Fortune 500.

You're just running at a scaled customer ARPU of $2 million today. So what are the key elements of the strategy that help you increase wallet share, so it's not $2 million but $20 million or $40 million or $80 million over time?

David Steinberg -- Co-Founder and Chief Executive Officer

So great question, Jason. Well, I mean, first, let's start with One Zeta, which is consolidating all of our use cases into one sale, and we're starting to see more and more Zeta one customers. It's interesting because, today, the scaled customers we have spent over $100 billion a year in marketing, and in the last year, we captured 1% of that spend. As we look forward to the next few years, our goal is to get to 2% of that spend.

It's not -- we don't have to get to 5% or 10%. But what I would tell you is we have a number of enterprise clients where we are above 5% of their wallet share, which gives us a road map for how to move other clients up. And as you move the top of the range up, you bring the mean up. And as you bring that mean up, you can move from 1% last year to 1.25% or 1.3% this year and then go up from there to 200 basis points by 2028.

Chris Greiner -- Chief Financial Officer

You can -- Jason, you can see this really clearly in the earnings supplemental on Slide 10, where if you look at our greater than 1 million scaled customers, which on a brand basis grew 28% year over year, but on an ARPU basis, to your question, those greater than $1 million customers on average are already spending almost $7 million. Now if you compare that to our $100,000 to $1 million customers, the amount of leverage we have, there, call it, just shy of $500,000 per, we see a substantial growth and ramping as they spend more time on the platform. And then if you reference Slide 11, which is a slide that we produce annually, you'll see we continue to make significant inroads within scaling those customers within the first year, almost now at $1 million per compared to around $600,000, $700,000 in last couple of years in that first year.

Jason Kreyer -- Analyst

Perfect. And then just a follow-up to kind of something we talked about in December. But you spent a lot of time with your biggest customers over the last handful of months here. Just wondering how those conversations continue to progress.

And what is it about those conversations that drive them to wanna do more with Zeta?

David Steinberg -- Co-Founder and Chief Executive Officer

I think a lot of it goes back to efficiency and effectiveness and superior revenue growth, right? So when you think about artificial intelligence, the vast majority of enterprises that are out there are using it for efficiency and automation, which, of course, we're doing as well. But the ability to drive meaningful revenue growth per dollar invested into the Zeta Marketing Platform is causing them to move budget from other partners to us. And we're seeing that in the results, right? We grew the business 50% in the fourth quarter, 40% excluding political. And we're really seeing the AI's implementations.

We also talked about a 200% sequential growth in AI adoption from Q3 into Q4. That's also flowing through the results. It's driving more efficiency, and it's driving superior revenue growth to our competitors.

Operator

Our next question comes from Ryan MacDonald with Needham & Company. Please proceed with your question.

Ryan MacDonald -- Analyst

Hi. Thanks for taking my question, and congrats on a great close to 2024. David, you talked about this concept of One Zeta, where you're consolidating all of your use cases into one sale. Can you just talk about the challenges of sort of that concept given that you're now integrating sales forces on new products that you've recently acquired, adding in generative AI to that? And then are you seeing early signs of the benefits of that on, say, like the RFP flow in terms of just larger or more comprehensive RFPs as you mature in that motion?

David Steinberg -- Co-Founder and Chief Executive Officer

First of all, thank you, Ryan. We're super proud of the quarter and the year. The answer is absolutely yes. So we're starting, for the first time, to see meaningful RFPs coming in as One Zeta.

And I will tell you, the addition of Pam and the addition of Ed are gonna be really game changing for our ability to do that with the relationships they both have and the ability to cross sell across the entire company with one group. Whereas in the past, as I think you know, we had been siloed, where you would have to create a sale, then you would come in. You try to upsell. You'd have -- the hunter would close the deal and then the farmers would try to spawn out.

What we're doing now is we're leading with the One Zeta. And it's been very, very exciting to break through those silos.

Chris Greiner -- Chief Financial Officer

If you look inside the sales pipeline, Ryan, just as a point of evidence, what we're seeing is that the average deal size is up 35% year to year, so your point of bringing all 3 of those use cases together.

Ryan MacDonald -- Analyst

Appreciate all the color on that. And then obviously given all the success that the company has had, it's obviously gonna, I think, bring in new competition into the market. We've recently seen sort of reports about Meta trying to go to the agencies to do more on AI-based advertising. Can you just talk about what you're seeing maybe in terms of newer movement or any changes competitively and whether or not some of this moves from the walled gardens creates any concern in the near term?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. Thank you, Ryan. No, we're not seeing any competition from guys like that. In fact, we continue to grow our business with Meta pretty materially and have a very deep and meaningful relationship with them.

They tend to focus on very few, very large partners. You are seeing Google and Meta going at it a little bit against each other. But that has, in no way, shape or form affected us. Quite frankly, I think that it's, in some ways, benefited us.

Because we're such good partners with Meta, we've been able to drive more business and build a deeper and more meaningful relationship. I would also comment on the fact that we haven't seen any small up and comers either. And I think a lot of that has been the inability for start-ups outside of pure AI to get financing over the last 3 to 4 years. So in a normal world, you would get to our scale and our growth rate and our size, you would have some new start-ups coming after you.

We have not seen that either. So it's been a really unique opportunity as we've brought everything together where you would traditionally need upwards of 17 different vendors to put together what the Zeta Marketing Platform can deliver. We're not seeing any one organization trying to replicate that at this point.

Operator

Our next question comes from Terry Tillman with Truist Securities. Please proceed with your question.

Terry Tillman -- Analyst

Yeah. Congratulations from me as well. Hi, David, Chris, and Matt. First question might be a multiparter, and then I had a follow-up.

But you all provided some interesting stats in terms of the pipeline and the value of the pipeline kind of ending the year. What I'm curious about is how do we think about timing of that converting into like meaningful or material revenue. And if some of these are kind of One Zeta deals, does it take a little bit longer before we start seeing that in the model in '25? And then I had a follow-up.

Chris Greiner -- Chief Financial Officer

Yes, we're not seeing any change in our deal cycles, and that's been something that we've been saying now for, frankly, years. So without a doubt, some of the bigger deals that come in through the RFP process will be, call it, 7 to 12 months, in that range. And they can move much faster than that by the way. But still the vast majority of the deals we're closing, and it doesn't mean that's to take a long amount of time, are still pilots and proof of concepts.

That can be on all 3 use cases. So it doesn't necessarily stop us from having deals in the pipeline take longer if you will. I don't expect the conversion of the pipeline to be any different in '25 than we've seen in '24 or 2023.

David Steinberg -- Co-Founder and Chief Executive Officer

And remember, Terry, that 7 or 8 months, many of those guys entered our pipeline 7 or 8 months ago. So you're actually seeing a consistent movement from what I would say is the pipeline to RFPs and pilots to converted to clients to the ability to scale to scaled clients and then ultimately super-scaled clients. So I think that if you look at our confidence in our business, putting forth a 2028 plan that continues to show a minimum of a 20% organic CAGR, we're feeling very bullish about where the business is right now.

Terry Tillman -- Analyst

Yes, that's definitely clear, David. I guess it seems like forever ago, but it really wasn't that long ago in Zeta Live. So much has happened since then, but there was a lot on the mobile side. So just maybe a quick update on that.

And potentially could this start becoming meaningful in '25 in terms of a key revenue channel?

David Steinberg -- Co-Founder and Chief Executive Officer

We've always said we thought mobile would be a big driver into 2026 and beyond, and it could happen in 2025. We have a number of clients who have adopted it, and we're very, very proud of the product we've built, and it's doing quite well. I mean it's growing at a massive rate, but off a 0 base, right? So the one challenge about getting bigger and bigger is it's harder for new products to really drive the needle. But if you look at products like connected television, they're still growing above 100% a year.

So now that that's a meaningful revenue, it's starting to really impact what we're seeing as a business. I expect mobile to do that in the years to come.

Chris Greiner -- Chief Financial Officer

It's a very natural selling motion for our sellers, right? It's not as though they have to learn a new capability. It's really an extension of how they're selling today as a new channel.

Operator

Our next question comes from Arjun Bhatia with William Blair. Please proceed with your question.

Arjun Bhatia -- Analyst

Perfect. Thank you, guys. Congrats on the strong close to the year here. Maybe if I can switch back to agencies for a second.

Chris, I'm curious what role you see agencies contemplating in that 2028 model. We know they're kind of on the upward part of the S curve in terms of growth right now. But kind of how are you anticipating the growth might shape out over the next couple of years? And then the other piece just in terms of mix, like where are you thinking the agencies are in the pace of adopting digital channel -- or, sorry, direct channels? Is that -- are you starting to see that pick up in late '24, early '25 here? And just how do you think that might trend over the coming years?

Chris Greiner -- Chief Financial Officer

Yes. So we think of the agency first off, which reached, call it, right around 20% of revenue this year, which was obviously expected and substantial growth to be an even bigger part of the pie as we go into -- deeper into 2028. But we see it as part of the core being able to get to that growth plan by itself just by accessing more and more wallet share. And the brand strategy and the brand go to market is one of the fastest ways that we can do that.

Obviously, we highlighted a number of other growth drivers incremental to the core that, frankly, could drive us north of the $2.1 billion, which is why we said it's an at least. In terms of where we are on getting direct mix, we made really exciting improvements this year. We're still not quite at like roughly 50% direct to 50% integrated, so there's still upside there. And I think as those agencies continue to grow in brands, grow in their channel adoption toward direct mix that that's also gonna be accretive to our gross margins in addition to the work we're doing on One Zeta.

And obviously, as more and more of our consumption-based revenues generated by our generative AI, that also brings with it higher margin profile business as well.

Arjun Bhatia -- Analyst

OK. Understood. And then just a quick one to put a finer point on it. I know we talked about this a little bit late in 2024, but kind of the fallout from -- or the customer conversations that you've had following up with some of your customers post the short report, has there been any impact or any change in those conversations since that's happened? How are your customers reacting? And what are you seeing from them post some of the events late last year?

David Steinberg -- Co-Founder and Chief Executive Officer

Well, I think if we had seen material changes, Arjun, we wouldn't have grown by 50%. What I would tell you is no. We have seen no material changes from the last update we gave, which was we have not lost any client directly over this, and we continue to see that. I was super excited that we were able to announce that our Audit Committee had done a full review.

They brought in a full forensic accounting firm that went through every single one of the transactions that could be deemed a client and a vendor and found no issues whatsoever in there. And then we brought in a law firm that is one of the top data experts in the country. They vetted everything that was in the report, and they found no merit whatsoever in it, and they reported that to our Audit Committee, which was very happy to put this issue behind us.

Operator

Our next question comes from Kelly Valentini with Goldman Sachs. Please proceed with your question.

Unknown speaker -- -- Analyst

Hi. Thank you for taking my questions, and congrats on the quarter. Wanted to walk through the comments you made on getting those budget penetration to 5% to 15%. Curious like in the typical customer you see that expansion, how much of that would you expect to be taking share from walled garden versus taking share from other technology providers?

David Steinberg -- Co-Founder and Chief Executive Officer

Thank you, Kelly. I appreciate the compliment, and it's a great question. We don't traditionally take business from the walled gardens. We traditionally partner with them.

So what we're able to do by using our data, we're able to build an attribution model that can go into a walled garden, or it can go outside of a walled garden. So we see the walled gardens more as our partner than our competitor. Now there are a number of companies out there that obviously we're taking meaningful market share from. Hard to grow the business at these rates if we weren't.

They're traditionally, I would say, last-generation marketing clouds or last-generation DSPs, where they haven't fully integrated data and AI as native to the application layer. So because of their tech debt, they're not able to get to the type of speed to intelligence that our platform can, which allows for substantially superior return on investment. So if the market itself is growing 10%, 15%, and we grew 40%, not including political, that would infer everything above that would have been taking market share from competitors.

Unknown speaker -- -- Analyst

That makes sense. And then as a follow-up, just curious, anything you can give on what assumptions you're making on AI revenue growth and kind of the 2025 and 2028 targets?

Chris Greiner -- Chief Financial Officer

Yes. We didn't -- we haven't broken it out in the 2028 in a line item fashion. We did say it will be 1 of the 5 new levers we feel like we have in addition to the core growth business. But I think if you look toward 2024 and our consumption-based revenue, which accelerated by north of 40% year over year, and that's, call it, roughly half of Zeta's revenue, and that was an acceleration from 2023, certainly, a portion of that is driven by our early stage generative AI products, which are intended to drive higher outcomes, which then drive more usage, which then drive more spend with the brand annually over time.

David Steinberg -- Co-Founder and Chief Executive Officer

And what we're seeing, Kelly, is the ability to meaningfully drive consumption is a substantially better return on investment for us as a company than charging an all you can eat or putting sort of a small price on the AI, although we will be rolling out AI products in the near term that will be have to be paid for as it relates to sitting on top of the Zeta Marketing Platform.

Operator

Our next question comes from Elizabeth Porter with Morgan Stanley. Please proceed with your question.

Katie Keyser -- Morgan Stanley -- Analyst

Great. Thank you so much. This is Katie on for Elizabeth Porter this afternoon. Wanted to hit on the verticalization piece here.

Zeta has had a lot of success given kind of diversification of the business across industries. And some of those key verticals you called out in commerce and CPG do have some kind of vertical-specific competitors in there. So what gives Zeta the right to win in these verticals where you're more kind of underpenetrated today? And when can we expect that vertical expansion contribution to layer into the model?

Chris Greiner -- Chief Financial Officer

Yes. No problem, Katie. Thanks for the question. We've already started to build.

In fact, in a lot of our virtual demos that we'll do with investors, we showcase a CPG real-world demo today, where we're actually using customers and being able to show them what their customers are spending on with their brand as well as with their competitors. So it comes down to in each one of those verticals where we think we're underrepresented today, where there's a massive TAM and amount of marketing and advertising spend, all comes down to the type of data that we can put in front of the marketer and how actionable we can make it for them in a totally different way than what our competitors can do. So it all starts with the data and then the software and the analysis that we put on top of it and the recommendations that they can then perform on the platform coming out of it.

David Steinberg -- Co-Founder and Chief Executive Officer

And by the way, a lot of the "vertical competitors" really are more vertically focused from a sales perspective than they are from a platform perspective. So we've invested over the last year pretty heavily in our sales force, as you've seen, growing to 180 quota carriers, easy for me to say. And we're expanding out into new verticals with salespeople that we think will really benefit us this year and the years to come.

Katie Keyser -- Morgan Stanley -- Analyst

Helpful. And then just one quick follow-up on the NRR. In the long-term model, not looking to inflect too materially from the 114% today. Obviously, that's a pretty impressive stat already.

But maybe just expand on the different drivers there. Is there any limitation on the upper bound to NRR around expansion activity as you kind of land with these higher ARPU customers?

David Steinberg -- Co-Founder and Chief Executive Officer

Well, there's no limitation on where it can go, although we continue to guide to 110% to 115%, which has been where we have. We've been sort of between 111% and 114%. What I would say is, first, you've got the agency holdcos, which as their ARPU grows dramatically, that's a big benefit to NRR. And we just don't lose a lot of customers, which is also a very good thing, right? So I think we can continue to keep it in that range.

And then in the years to come, I think we can begin to look at getting it above that range. But in the short run, we feel very comfortable with 110% to 115%.

Chris Greiner -- Chief Financial Officer

And I think there's a really good -- Katie, on Slide 11 in the earnings supplemental, we show the life of the cohorting of a customer set. And if you look at those that are spending -- that have been on the platform less than a year, they're spending an average of $900,000, which is up from $600,000 last year. That 1- to 3-year cohort continues to accelerate to $1.2 million. And then those customers that are with us 3 or more years, they're continuing to grow.

They're now at an average of $2.6 million compared to $2.1 million last year. So we feel like that 110% to 115% is rooted in deep analysis by cohort. And it's also consistent with our growth algorithm, which has historically been about half of our growth coming from our existing customers and half from new.

Operator

Our next question comes from Jackson Ader with KeyBanc Capital Markets. Please proceed with your question.

Jack Nichols -- Analyst

Hey, guys. This is Jack on for Jackson Ader. Wanted to ask on what your political revenue assumptions are implied in your outlook. And do you expect to gain market share with political campaigns?

Chris Greiner -- Chief Financial Officer

So there's two pieces -- thanks, Jack. Two pieces of the answer there. There's political candidate revenue, which was, call it, $44 million in 2024. And the reason why we broke that out throughout the year in our guide and in our reported results was it's always been our expectation that, in 2025, there shouldn't be any political candidate revenue of any materiality certainly.

Then there's the advocacy portion, which is where Zeta is working with political action groups as well as NGOs. In 2024, that was in line with our expectations. It was $36 million. For that piece of our business, which we're putting more quota carriers into, building new capabilities, leveraging relationships that come out of the political cycle, our expectation is that business goes to $20 million to $25 million in 2025.

Now if you compare the -- how that was in an off political year back in 2023, that's up considerably from around $13 million. So that's how we think about the contribution of political candidate and advocacy in 2025.

Jack Nichols -- Analyst

Got it. Very helpful. And as a follow-up, how much revenue today is coming from the marketing cloud? And how much do you expect that to be in 2028 as it progresses?

David Steinberg -- Co-Founder and Chief Executive Officer

I mean all of it? All our revenue goes through the marketing cloud, so it's -- I don't -- maybe you could fine-tune that question.

Jack Nichols -- Analyst

I'll follow up with you guys offline on that.

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. Well, I mean we're gonna talk in a bit. But literally, Jack, all of our revenue goes through the marketing cloud at this point. So when we consolidated that a few years ago, everything now goes through one user interface with one reporting infrastructure, and everything is totally integrated.

Operator

Our next question comes from Koji Ikeda with Bank of America. Please proceed with your question.

Koji Ikeda -- Analyst

Yeah. Hey, guys. Thanks so much for taking the questions. Maybe this one's for Chris.

And so Chris, when I look at the third-quarter transcripts, I recall that you said you were very comfortable with 2025 consensus revenue growth of 17%. And so I know that was supposed to be without LiveIntent, and that got us to a number that is a couple of millions below the $1.144 million -- $1.144 billion that you're guiding to today without LiveIntent. But if I use that $1.144 billion number and then compare it to where you ended up at 2024, that implies 14% organic growth versus the third quarter when you said 17%. And I know on the Slide 24, it says 21% growth.

And so I just wanna make sure I'm comparing apples to apples here with that original 17% comment from the third-quarter call.

Chris Greiner -- Chief Financial Officer

Yes. And that 17% comment, when we very carefully walked through what we were presuming, that should have gotten most analysts and The Street to around, call it, $1.2 billion in revenue compared to the $1.24 billion that we're at today. So obviously, you back out $17 million of LiveIntent from the $1.6 billion in revenue we did in 2024 and you back out $44 million of political candidate revenue, you get to a normalized base of $944 million. You then do the same on 2025.

You remove $96 million from LiveIntent -- for LiveIntent, which is what we'd expect that business, and that's a 20% growth rate consistent with what we talked about when we made the acquisition. You get to, call it, $1.44 billion. So it's a 21% growth rate, so roughly 4 points higher than that 17% that we had talked about earlier. We went through the task also on Slide 24 to kind of clearly break out this in steps by quarter, and what you'll find is that for each quarter of 2025, growth is effectively between 20% and 22% when you exclude LiveIntent and exclude political candidate revenue itself.

David Steinberg -- Co-Founder and Chief Executive Officer

And I think, Koji, it's important because we've got to exclude those 2 things to get to the 21%. But you really have to exclude the revenue we picked up from LiveIntent in the fourth quarter, right? So even if you just wanted to do apples to apples, you would pull that inorganic revenue out. If -- you can't pull it out of 2025 without pulling it out of 2024.

Koji Ikeda -- Analyst

Got it. And maybe just a quick follow-up. I did wanna ask if -- the conservatism in the guidance or any sort of consideration in the guide this year, just thinking about all the macro and regulatory and tariff noise that's out there, is there any additional conservatism that you pulled with the guidance this year?

Chris Greiner -- Chief Financial Officer

We tried to be consistent with our historical approach, which is to build in ample conservatism so that we don't need the midpoint or even the high end of our metrics to get to our guidance. As I said kind of earlier in the Q&A, we can get to guidance at the low end of each one of our -- one of our metrics call for. And we obviously made that call based upon our awareness of what's going on in the macro backdrop. David, anything you'd add?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. I would say, Koji, just we've known each other for a while now. We've beat and raised 14 quarters in a row. Our goal is to be here the same time next year and saying we've now beat and raised 18 times in a row.

And we feel we've put the right guidance out to do that.

Koji Ikeda -- Analyst

Thanks, guys.

David Steinberg -- Co-Founder and Chief Executive Officer

Thanks, Koji.

Operator

Our next question comes from Brian Schwartz with Oppenheimer. Please proceed with your question.

Chris Greiner -- Chief Financial Officer

Brian, you out there?

David Steinberg -- Co-Founder and Chief Executive Officer

Brian, are you on mute?

Brian Schwartz -- Analyst

I'm on mute. Sorry about that, everyone. A couple of questions from me. Chris, I just wanted to ask you on the decel in the 1Q guide.

I know the comp is a few points harder here and there's political spending headwinds on the comparable. But is there anything else that you're contemplating in that guide to see that type of organic deceleration beyond those items? Having 1 less day, is that also an impact in terms of the 1Q guide? And then I have a follow-up.

Chris Greiner -- Chief Financial Officer

No, no, that wouldn't be any kind of material impact. As you said, it's 30% all in. The guide is 20% if you adjust for effectively LiveIntent because there really wasn't any political candidate revenue in the first quarter. And just -- it's our conservatism.

We're trying to be consistent with what we've done in the past and just build multiple ways to get to the number and feel like if we do that, we'll be in the same place we've been in the past in terms of what our traditional beats are.

Brian Schwartz -- Analyst

OK. And then the follow-up question I had on the 2028 guide. I know there was some discussion earlier about the agency business, which is doing really well for you. Is it your expectation, Chris, that, that business could double in terms of its percentage of the revenue mix as we fast forward to 2028.

Chris Greiner -- Chief Financial Officer

So the base will keep growing, right? So I think doubling in size, I mean, it's not without -- it's not outside the realm of possibility. I think it will be a bigger and bigger piece of the pie. Like I said, today, it's about 20% of revenue. I think doubling, there are certainly cases for that, but I wouldn't count on it as we sit here today.

David Steinberg -- Co-Founder and Chief Executive Officer

Yes, nor would I, Brian. I think it's gonna double as a business. I don't think it's gonna double as a percentage.

Operator

Our next question comes from Zach Cummins with B. Riley Securities. Please proceed with your question.

Zach Cummins -- Analyst

Hi. Good afternoon, and just adding on my congratulations for the quarter. Just double clicking a little bit more on the agency opportunity. I mean, David, can you speak to just the growth opportunity that you have with the top 5 global holding companies versus maybe your aspirations to expand into the mid-market on the agency side?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. I mean we are adding mid-market agencies faster than I think we could have even expected. And they're all on platform. So that's a really nice thing for margin in the long term.

But what I would say, Zach, is that the opportunity with the 5 large agency holdcos we work with is very, very large. We continue to be their most profitable partner. We tend to be more flexible than our competitors, and we have built deep and meaningful relationships with them. I would expect us to continue to meaningfully grow those businesses and continue to meaningfully grow brands.

And if you look at our sort of projections out through 2028, as Chris, I think, has said a few times, we can be at the low end of our metrics and still beat those numbers.

Zach Cummins -- Analyst

Understood. And just my one follow-up toward Chris. Can you talk about your plans for quota carrying hiring for Zeta 2028? Obviously, much larger sales force than 3 years ago but just curious of any sales efficiency gains that you're baking into that versus necessary capacity to execute on those targets.

Chris Greiner -- Chief Financial Officer

Yes. Slide 17 in the supplemental lays out what our compound growth rates have been through the first -- if you will, the first long-term plan. The quota carrier compound annual growth rate was 22% from 2021 to 2024. We are baking in some efficiency that we, frankly, had been seeing as of late to where it's our expectation that we can grow the quota carrier base between 10% and 15%.

What we found, though, is we don't want a growth number to force us down the path of just quantity, what we've continued to do well, and this allows for us to focus on that quality element over just throwing bodies at sales.

Operator

Our next question comes from Richard Baldry with ROTH Capital Partners. Please proceed with your question.

Richard Baldry -- Analyst

Thanks. Maybe switching gears. I wanted to look at the balance sheet. You repurchased $31 million worth of shares in the quarter.

That's 3 times what you've done the rest of the year combined and almost matched your free cash flow. Sort of curious your thought process around the buyback on a go forward. Do you think you'll run it at a higher percent of the free cash flow? Do you think you'd be opportunistic and ever like look at taking the actual cash balances down to pursue it more aggressively? Just to give us a backdrop for the year ahead.

David Steinberg -- Co-Founder and Chief Executive Officer

I think, listen, Rich, the stock we thought dropped to a stupid place. So we massively accelerated buying the stock. We still think it's very low, and we'll continue to buy it at an accelerated pace. I don't see any better use for our cash right now than buying our shares back.

And yes, we used about 100% of free cash flow. We expect to drive meaningful free cash flow this year. You could see us do, I would say, at least half and potentially much higher as a percentage of free cash flow as we look at buying the stock back because, once again, it right now is the best investment for us for our cash in the current environment.

Richard Baldry -- Analyst

And the follow-up for me would be, it looks like we probably have a better M&A environment with an administration that might let more things go through. So how are you thinking about growth, the inorganic growth opportunities now and as you look ahead to sort of broaden offerings, get into new markets, etc.?

David Steinberg -- Co-Founder and Chief Executive Officer

Yes. We are seeing more deals now than we have in many years. What I would say, it's still gonna be hard to find deals that match our 4 M&A pillars. If we can find deals that match our 4 M&A pillars, we will act on them.

We have meaningful cash. We have meaningful capacity, and we're in a very unique position with the type of free cash flow we expect to generate to be able to do very opportunistic deals. I will say what I say to you whenever we're together. I believe transformative deals transform both companies for the worse, so we won't be doing anything that "is transformative".

But we will continue to be very, very opportunistic and look at opportunities where we can take 1 plus 1 and equal 4.

Operator

Our next question comes from Ryan MacWilliams with Barclays.

Eamon Coughlin -- Analyst

Hey, David and Chris. This is Eamon on for Ryan. Thanks for the question, and appreciate all the detail today. Can you help us understand the key levers for the free cash flow and EBITDA margin expansion highlighted by your 2028 guide?

Chris Greiner -- Chief Financial Officer

Yes, certainly. So we talked about getting anywhere from 100 to 300 basis points, we think, from cost of revenue, and that really being driven by 3 different initiatives. The first being our One Zeta, where we're being more and more successful at bringing all 3 use cases together, which gives us good synergies and higher margin profile, continuing to drive more and more direct channel usage across the platform, in particular, with agencies. And then as our generative AI and our consumption-based revenue, which tends to have also a higher margin to it, scales, that should be a lever toward the COGS line.

We think we can continue to get great leverage out of the opex line. And that, I think, is really well represented in the supplemental deck where we show that we've had a 30% compound annual growth rate on revenue over the last several years as compared to only 15% compound growth rate in headcount, in U.S. headcount even being at a rate well below that. In addition to those levers, you see we'll eventually start to make up that working capital deficit that we have with the agency as that growth starts to catch up.

That's a significant lever. I mean, this year's -- just the fourth quarter, for example, if we had a neutral working capital position, which we didn't, it was a $22 million gap, would have been a 76% free cash flow conversion. So I think that's a key driver. And then over time, and again, this year's results evidenced it, less and less as a percentage of revenue of capex will be in the business.

So I think we'll get efficiencies from there as well, which puts that 65% conversion out in 2028 really nicely within reach of us. In fact, already there really on a normalized basis.

Eamon Coughlin -- Analyst

Perfect. And then as a follow-up, have you seen any changes in the environment post U.S. election as it relates to customers' willingness to spend?

David Steinberg -- Co-Founder and Chief Executive Officer

We really haven't, Eamon. We've been working with our customers. And as you know, we're very, very close to our customers. I think people always discount what a large percentage of our business we have done when we enter the year.

We already have incredible visibility into what we put out. And as Chris has said a couple of times, we've tried to be incredibly conservative in the projections that we put forth. So we have not had any clients pull back. And thus far, we continue to be full speed ahead.

Operator

There are no further questions at this time. I would now like to turn the floor back over to David Steinberg for closing comments.

David Steinberg -- Co-Founder and Chief Executive Officer

I could not be prouder of this team to have worked through the turbulence that we've seen over the last quarter and put up the type of results that we've put up as an organization. I think it really shows that our strategy of putting AI and data as foundational to our platform, bringing in the world's best people and letting them do their jobs really can ultimately drive to financial results that show how strong our platform is in the marketplace. And we believe we're gonna organically double this business again over the next 4 years. I could not be more proud of our Zeta people, and I am incredibly proud to be running this company.

Thank you very much.

Operator

[Operator signoff]

Duration: 0 minutes

Call participants:

Matt Pfau -- Senior Vice President, Investor Relations

David Steinberg -- Co-Founder and Chief Executive Officer

Chris Greiner -- Chief Financial Officer

Matt Swanson -- Analyst

DJ Hynes -- Analyst

Jason Kreyer -- Analyst

Ryan MacDonald -- Analyst

Terry Tillman -- Analyst

Arjun Bhatia -- Analyst

Unknown speaker -- -- Analyst

Katie Keyser -- Morgan Stanley -- Analyst

Jack Nichols -- Analyst

Koji Ikeda -- Analyst

Brian Schwartz -- Analyst

Zach Cummins -- Analyst

Richard Baldry -- Analyst

Eamon Coughlin -- Analyst

More ZETA analysis

All earnings call transcripts

This article is a transcript of this conference call produced for The Motley Fool. While we strive for our Foolish Best, there may be errors, omissions, or inaccuracies in this transcript. As with all our articles, The Motley Fool does not assume any responsibility for your use of this content, and we strongly encourage you to do your own research, including listening to the call yourself and reading the company's SEC filings. Please see our Terms and Conditions for additional details, including our Obligatory Capitalized Disclaimers of Liability.

The Motley Fool has no position in any of the stocks mentioned. The Motley Fool has a disclosure policy.