EverQuote (EVER) 2026 年第三季財報電話會議：營收成長 25%，AI 投資擴大

重點摘要

• EverQuote 公布 2026 年第三季營收為 2.181 億美元，年增 25%，高於公司指引區間的高標。
• 調整後 EBITDA 成長 41% 至 3,550 萬美元，調整後 EBITDA 利潤率為 16.3%。營業現金流達 4,050 萬美元。
• 車險營收成長 21% 至 1.904 億美元，而住宅險營收成長 70% 至 2,770 萬美元。管理層預期中長期內住宅險仍將是成長較快的垂直領域，但警告不要將 70% 的成長率直接外推。
• 可變行銷支出 (VMD) 增加 25% 至創紀錄的 6,260 萬美元。可變行銷利潤率 (VMM) 為 28.7%。
• 展望 2026 年第四季，管理層預估營收介於 2.13 億至 2.23 億美元之間，調整後 EBITDA 為 2,900 萬至 3,200 萬美元。由於 EverQuote 投資於代理型 AI（Agentic AI）流量管道與新產品，預計 VMM 將呈季減。
• EverQuote 推出原生的 AI 產險比價平台 Luva。該公司也正在開發與 Meta、OpenAI 及 Anthropic 等平台的整合，惟仍需經各平台審核與批准。

核心財務業績

GAAP 淨利下滑主因所得稅較去年同期增加 730 萬美元，且 EverQuote 認列了一筆與廣告服務供應商合約相關的 400 萬美元一次性解約金。

業務與營運表現

需求依然廣泛，前十大保險公司客戶中有八家呈現季成長。管理層將此有利環境歸因於保險公司的核保獲利能力強勁，以及對保單成長的持續需求。

住宅險是成長最快的垂直領域。公司表示，專屬資源、針對住宅專屬工作流程的產品調整，以及高於預期的保險公司需求，共同支撐了 70% 的成長。管理層預計中長期住宅險營收將繼續超越車險，但也指出該領域是在相對較低的基期上擴展。

EverQuote 繼續向保險公司及在地代理人推廣其基於 AI 的競價產品 SmartCampaigns。管理層表示，先前公布的控制個案研究顯示，相較於客戶傳統的競價方式，其效能提升了雙位數或約 20%，具體取決於保險公司的起始水準。

在內部方面，採用代理型 AI 程式碼開發流程後，工程週期時間季減 36%。該公司還提及在非工程部門廣泛應用 AI 工具。

Luva 專為偏好研究且熟稔 AI 的投保消費者設計，將採用混合變現模式，可能包括佣金、基於績效的保單媒合及轉介費用。管理層計劃追蹤保險公司參與度、產品功能、採用率以及來自對話式 AI 與代理型管道的流量。

EverQuote 在第三季結束時無任何債務，並擁有 2.291 億美元的現金及現金等價物。在前一份實施至第二季的計畫以 5,000 萬美元買回 250 萬股後，董事會授權了新的 5,000 萬美元庫藏股計畫。管理層的資本配置優先事項仍是維持強健的資產負債表、進行有針對性的策略併購與投資，以及買回股票。

管理層指引

管理層預期保險公司的健康需求將持續至 2027 年，並根據 EverQuote 投資人簡報中引用的產業研究報告，認為產險軟性市場週期可能會延長至 2029 年或更久。

公司重申年營收達到 10 億美元的目標。管理層目前預計將在未來一至兩年內實現該目標，同時持續產生強勁的現金流並投資於 AI 相關的成長計畫。

風險與觀察重點

• 代理型與 LLM 驅動的流量佔比仍微不足道：目前 EverQuote 的流量中來自 LLM 的比例不足 1%，代理型消費者的貢獻則更低。
• 新代理型流量來源在早期階段的效率較低。管理層表示第四季的 VMM 可作為 2027 年的合理初步參考指標，同時強調最終表現將取決於這些管道的演進方式。
• 與 Meta、OpenAI 及 Anthropic 的整合仍需經各平台的審核與批准程序。
• 保險公司仍在評估代理型 AI 的保單分銷模式。管理層強調了相關疑慮，包括頻繁重新比價、保單壽命縮短、費率透明度、逆向選擇、法律問題以及監管合規性。
• 目前的 AI 比價工具僅能存取有限的保險產品集，且要完成整個比價流程可能需要龐大的算力成本。
• 住宅險 70% 的成長率得益於較低的比較基期以及高於預期的保險公司活動，這使得單季成長可能呈現波動。

分析師問答重點

AI 驅動的保險比價：管理層預期隨著時間推移，保險搜尋將更具代理性 (agentic)，但表示保險公司可能會審慎行事並限制對費率存取的權限。EverQuote 旨在連結消費者、AI 平台與保險公司，同時保留保險公司在定價、產品及品牌展示上的控制權。

2027 年成長驅動因素：除了有利的保險公司環境外，管理層指出 SmartCampaigns 的更廣泛採用、針對在地代理人的新產品、拓展滲透率較低及上游行銷漏鬥的流量管道、AI 搜尋、住宅險成長以及 Luva 均為潛在的貢獻因素。

獲客經濟效益：管理層表示，更高的比價頻率可能會創造更多可變現事件。更佳的 AI 輔助研究也可能提高轉介轉化率與變現能力，儘管最終的經濟模式仍具不確定性。

Google 曝險：EverQuote 表示其流量組合多元化，且主要集中於付費程式化廣告，而非自然搜尋；管理層認為自然搜尋是受 Google 調整影響最大的領域。

資本配置：該公司表示，現金產生能力使其具備彈性，能維持資產負債表健全、評估可擴大總可定址市場 (TAM) 的合作夥伴關係、投資與併購，並買回股票。

法說會逐字稿全文

---

完整財報電話會議逐字稿

管理層陳述

Operator

Hello, everyone. Thank you for joining us, and welcome to the EverQuote Third Quarter 2026 Earnings Call. After today's prepared remarks, we will host a question-and-answer session. [Operator Instructions]

I will now hand the conference over to Sara Buda, Vice President of Investor Relations. Please go ahead.

Sara Buda

Thank you. Good afternoon, and welcome to EverQuote's Third Quarter 2026 Earnings Call. We will be discussing the results announced in our press release issued today after market close. With me on the call this afternoon are Jayme Mendal, EverQuote's Chief Executive Officer; and Joseph Sanborn, EverQuote's Chief Financial Officer and Chief Administrative Officer.

During this call, we may make statements related to our business that may be considered forward-looking statements under federal securities laws, including statements considering our financial guidance for the fourth quarter of 2026.

Forward-looking statements may be identified with words and phrases such as aim, expect, believe, intend, anticipate, plan, will, may, continue, upcoming and similar words and phrases. These statements reflect our views only as of today and should not be considered our views as of any subsequent date. We specifically disclaim any obligation to update or revise these forward-looking statements, except as required by law.

Forward-looking statements are subject to a variety of risks and uncertainties that could cause the actual results to differ materially from our expectations. For a discussion of those risks and uncertainties, please refer to our SEC filings, including our annual report on Form 10-K and our quarterly reports on Form 10-Q on file with the Securities and Exchange Commission and available on the Investor Relations section of our website.

Finally, during the course of today's call, we refer to certain non-GAAP financial measures, which include adjusted EBITDA and adjusted EBITDA margin, variable marketing dollars and variable marketing margin, which we believe are helpful to investors. A reconciliation of GAAP to non-GAAP measures was included in the press release we issued after the market closed today, which is available on the Investor Relations section of our website.

And with that, I'll now turn the call over to Jayme.

Jayme Mendal

Thank you, Sara, and thank you all for joining us today. We delivered exceptional performance in Q3. We generated record results across our key financial metrics, exceeding the high end of our guidance range. Our marketplace continues to fire on all cylinders as carriers maintain strong underwriting profitability and appetite for growth, and our team is executing well to meet this demand.

In Q3, we continued to diversify our revenue across carriers, product offerings, and verticals, punctuated by outstanding performance in our home vertical, which grew by 70% year-over-year. Our market position remains strong, and we are growing and deepening customer relationships with both carriers and agents.

Furthermore, we believe we are making great progress as we begin to lead the P&C insurance industry into the AI era. First, we continue to use AI to drive productivity. In Q3, our engineering teams reduced cycle time by 36% quarter-over-quarter through adoption of our agentic coding pipeline. We have also achieved broad adoption of AI tools across non-engineering functions, allowing us to innovate and launch new products faster while growing adjusted EBITDA by over 40% year-over-year.

Second, AI remains a reliable driver of our consistent marketplace growth. We have deployed SmartCampaigns to new carriers and to more local agents. This AI bidding solution drives benefits, including higher conversion rates, which in turn allows us to access more budget so we can maximize highly targeted policy volume at customers' performance targets. We also rolled out new AI-native features to local agents in the quarter, including a new AI-based reporting tool that surfaces insights to help them understand and optimize account activity.

Third, and perhaps most importantly, AI underpins our strategy to broaden our value to customers and expand our long-term addressable market. We are at the forefront, leading P&C insurance distribution into the AI era.

In Q3, AI agents took center stage when Meta and OpenAI launched their personal agents. We have long anticipated personal agents as the next step in the evolution of insurance digital marketing and distribution, from web-native to LLM-native to agentic.

For months in advance of the launches of Muse and Dots, we were in discussions with carriers about agentic commerce to help educate and prepare them for this next wave of change. In fact, during the quarter, we published a thought leadership white paper on agentic shopping and P&C insurance distribution, which has led to an active dialogue with major carriers as they themselves contemplate how to adapt. This topic has become a major focus area for our customers, and we are partnering with carriers to innovate solutions that help them navigate this fast-developing landscape.

Last week, we launched Luva, a separately branded AI-native P&C insurance shopping platform for consumers that will be integrated with LLMs and AI commerce agents. We believe EverQuote can emerge as a key player in the AI ecosystem by bridging the needs of consumers using LLMs and AI agents to research and shop insurance with insurers' need to access these consumers while maintaining control over how their pricing, product, and brand are represented.

We are rolling out apps across LLM and agentic platforms like Meta, OpenAI, and Anthropic. These integrations are subject to the platforms' review and approval process, but we'll build on each solution over time in close partnership with carriers.

It is clear that our carrier and agent customers will need help adapting as agentic AI plays a larger role in P&C insurance distribution over time. We see tremendous opportunity in this dynamism. With our unique data, technology, and industry relationships developed over the last 15 years, we are ideally suited to help steward P&C insurance distribution into the AI era. Recent and forthcoming product launches position us to capitalize on this market evolution.

In summary, it has been an exciting quarter at EverQuote, and we are proud of maintaining our track record of sustained profitable growth while making investments in our future. As evidenced by our Q3 performance, the business is healthy and the team is executing well. We continue to make progress across our strategic priorities, and we are accelerating new product innovation to lead our industry into the agentic AI era.

I will now turn the call over to Joseph, who will discuss our financial results and outlook.

Joseph Sanborn

Thank you, Jayme, and good afternoon, everyone. We delivered record results in Q3 and exceeded the top end of our guidance across all key financial metrics. We grew revenue 25% and Adjusted EBITDA 41% year-over-year. We also drove record Variable Marketing Dollars, or VMD, and Adjusted EBITDA, while at the same time generating strong cash flow. Our AI investments are already yielding results, and we will continue to invest to extend our market leadership position even further.

Now, let's turn to the details of the third quarter. Total revenue increased to $218.1 million. Revenue from our auto insurance vertical grew to $190.4 million in the third quarter, an increase of 21% year-over-year.

Our home insurance vertical grew 70% in Q3 to $27.7 million, as carriers continue to lean into home as a growth opportunity, and as we align our solutions and resources to scale this vertical and diversify our business.

We are executing well in a healthy environment. Looking at customers, 8 of our top 10 carriers grew sequentially, underscoring that demand continues to be broad-based. VMD increased to $62.6 million in the third quarter, up 25% from the prior year period. Variable Marketing Margin, or VMM, was 28.7% for the quarter.

Turning to operating expenses and the bottom line. In the third quarter, GAAP net income was $18.2 million, down from $18.9 million in the prior year period, due primarily to income taxes increasing $7.3 million year-over-year and a one-time $4 million termination fee related to a vendor contract for advertising services.

Q3 adjusted EBITDA increased 41% from the prior year period to $35.5 million, representing a 16.3% adjusted EBITDA margin. Cash operating expenses, which excludes advertising spend and certain non-cash and other charges, were $27.2 million in Q3, up slightly from Q2.

We reported operating cash flow of $40.5 million for the third quarter. We ended the period with no debt and cash and cash equivalents of $229.1 million.

We recently announced that our Board authorized a new $50 million share repurchase program. As a reminder, under our previous program, which concluded in Q2, we purchased 2.5 million shares totaling $50 million. The recent authorization is a component of our 3-pronged approach to capital allocation, which includes: one, maintaining a fortress balance sheet; two, pursuing targeted M&A and investment opportunities that align with our growth strategy to expand our TAM; and three, returning capital to shareholders through stock repurchases.

The continued strength of our operations and expectations for ongoing growth and profitability give us confidence that we have the flexibility and capacity to execute these priorities.

Finally, as Jayme mentioned, we are focused on broadening our offerings and leading our sector into the agentic AI era. This includes new solutions that capitalize on emerging models for agentic LLM traffic and extend the value we provide carriers and agents as they adapt to new models. These initiatives build on the success of our marketplace and reinforce our position as an AI beneficiary long term. We look forward to sharing more with you in the coming quarters.

Turning to guidance for the fourth quarter of 2026. We expect revenue to be between $213 million and $223 million, representing 12% year-over-year growth at the midpoint. We expect VMD to be between $57 million and $60 million, representing 19% year-over-year growth at the midpoint. As we mentioned, we are investing in new agentic LLM channels and developing new solutions to drive long-term growth. As a result, VMM is expected to decrease sequentially from Q3.

Our Q4 adjusted EBITDA is expected to be between $29 million and $32 million, representing 22% year-over-year growth at the midpoint.

Looking ahead, we expect healthy carrier demand into the coming year. Insurer combined ratios remain low, and we are benefiting from operating in the relatively early stages of a prolonged soft market cycle. As illustrated in our investor presentation available on our IR website, industry analysts have forecast this soft market cycle could continue well into 2029 or beyond, as carriers focus on growing policies in force.

Amidst this positive environment, EverQuote is a trusted growth partner of choice for both carriers and agents. Our home vertical is showing impressive growth, and we continue to broaden our customer relationships. Our investments in expanding AI solutions are delivering measurable customer results, establishing a strong foundation for sustained growth.

As a reminder, on our November 2025 earnings call, we shared our goal to become a $1 billion revenue company within 2 to 3 years while also generating strong cash flow. Here we are 12 months later. As evidenced by our results and guidance, we are executing well toward our goal. We are on track to reach the $1 billion revenue mark within the stated timeline, which means that we expect to achieve our goal within the next 1 to 2 years.

At the same time, we are making the right investments today to lead our sector into the agentic AI era tomorrow and scale well beyond our near-term $1 billion revenue goal while generating strong cash flow.

In conclusion, we continue to execute our strategy to be a trusted growth partner for our customers. We reported a record quarter across all key metrics with strong growth and cash flows. We have a focused growth strategy to innovate and expand our offerings, capitalize on new growth opportunities with agentic AI, and broaden our value for our customers. We believe this is an exciting time to be a shareholder of EverQuote as we continue building upon our strong foundation while also innovating for what's next.

Jayme and I will now take your questions.

Operator

[Operator Instructions] Your first question comes from the line of Ralph Schackart with William Blair.

分析師問答

Ralph Schackart

Jayme, maybe just to kind of kick it off, maybe at a high level, can you walk us through how you see the workflow? Let's say I go to an agent such as Muse or something similar, and I don't know, maybe a simple query like, I want to lower my auto insurance, and maybe get a little bit more details. How does that sort of trigger, I guess, maybe your agent or sort of that flow, or how does EverQuote in this evolving world integrate into that? And then I have a follow-up.

Jayme Mendal

Sure. So I think there's how it works today and then how it's likely to evolve over time. Today, Meta or Muse will sort of make a determination as to how best to complete that task. And through our testing, we've seen it try a lot of different things, from going directly to carrier sites to looking at comparison sites like ours.

But if you zoom out, we've been anticipating and planning for agentic commerce for some time now. And I think from the consumer's perspective, these personal agents are going to be helpful across a lot of tasks, including researching and shopping insurance. But as we know, there are certain aspects of the insurance sector that differentiate it from others. You've got the regulatory and compliance considerations, the preference towards rate opacity from a handful of particularly the largest carriers, got material costs associated with accessing rates.

And so while from the consumer's perspective, using an AI agent to compare and shop is likely an attractive option, I think that ultimately the shape that this takes over time will really start to depend on how carriers respond. And Ralph, we published a thought piece on this, and we've since spoken with just about every major carrier on this topic over the last 6 weeks. And so I'll share a few things that we've heard, which I think represents probably the latest in terms of how the industry is thinking about it.

First, I think most carriers are in kind of learning mode right now. So they're just observing what's going on and trying to understand it. Number two is, I think across the board, they do perceive potential risk to their business if agentic commerce is left kind of unmanaged. And the risk that they are concerned about, I think, is continuous reshopping, which could compress their policy life expectancy. I think they do not support rate transparency, particularly if it's out of context or not controlled by the carrier themselves. That can lead to adverse selection. And then you still have a lot of legal and compliance questions that are unanswered. And so I think initially, most carriers are likely to move pretty slowly and impose some guardrails around how their rates are accessed, much like they did in the web era, candidly. But ultimately, I think the carriers will start to want to access this traffic as long as they can access it in ways that align with their requirements.

So you've got the consumer that will probably want to apply this use case. You've got the carriers who are still working through it. And then you've got the third category, which is what you asked about, which is the actual AI agents themselves, or the platforms that operate them. And what do they want? Well, they want a streamlined path to access products, right? They want inexpensive, they want fast, they want ultimately effective. And so we've done a lot of testing. Based on this testing, I think today, they're pretty far off from getting the job done well. They can only access a relatively small subset of available products. And it costs quite a bit to actually run one of these shopping cycles today in terms of compute.

So this is going to develop over time. We see it as a really significant opportunity for EverQuote, given our position in the value chain, our scale, our technical acumen, carrier relationships we have. But we do see an evolution towards more agentic discovery over time, same way traditional advertising shifted to digital. And we think EverQuote can play a significant role in this next phase of distribution to help solve for the best outcome, not just for consumers, but also for carriers and for the AI platforms themselves alike.

Ralph Schackart

That is super helpful. Just maybe one quick one. Just looking at homeowners growth of 70%, obviously very strong this quarter. Joseph, I think you talked about just carriers leaning in, but any more color on this in terms of the drivers? And then how should we think about this growth rate going forward? Thank you.

Joseph Sanborn

Yes, we were pleased with how the vertical did this quarter. I mean, just to give some context, Ralph, we made some decisions in the first part of last year to really start to put some new execution expertise around the home vertical. We have talked about how we saw this opportunity emerging. First, the carriers focused on getting the right pricing in auto, now they are switching to home. And we saw that coming, we got ready for it. And so our teams have been doing a really nice job. The first part of this year, we had 30%, 35% growth in Q1 and Q2. Q3 was even stronger.

What I would say what drove it in Q3, I think, would be a couple things. One is, I think our teams are doing a nice job scaling the offering, but we are also putting some dedicated resources on there as well. We also see the carriers at the same time leaning into the home opportunity broadly. And so we have been working with them, and we have been helping, as we have been putting more resources on, we have been helping them say, how do we set up this vertical with the specific attributes and characteristics in the workflow that are required for home? We have been doing that. So we have been very pleased with it. I would say it was the over-performance on home was in part driven by carriers coming on even stronger in the quarter than they anticipated when they told us at the start of the quarter how they were looking. So we are pleased by that.

As we look ahead, we see home continue to be our stronger grower than auto in the medium term. I would say 70% was very impressive, and I would not put that into your model, Ralph. But I would say having outsized growth from the home vertical, we were in the 30s for first part of -- 30% year-on-year growth for Q1 and Q2. We sort of see the home verticals continue to grow nice strongly going ahead.

Just be mindful on the numbers. It is off a relatively small base. So in any given quarter when you have a few things go really well, that can sort of amplify the year-on-year growth, which is what a little bit has happened in Q3. But again, really strong execution by the team. We are really pleased, and we are bullish on home's opportunity as we look ahead.

Operator

Your next question comes from the line of Maria Ripps with Canaccord.

Maria Ripps

Great. Congrats on the strong quarter. Clearly, you see strong momentum here exiting the year. So as we look ahead to 2027, can you maybe talk about the key growth drivers, sort of in addition to a favorable industry backdrop? And how should we think about sort of their relative contributions to growth?

Jayme Mendal

Can you repeat that last part of the question, Maria? You broke up a little. The contributions?

Maria Ripps

Yes. As we think about 2027, how should we think about sort of key growth drivers in addition to a favorable industry backdrop and sort of relative contribution to overall growth?

Joseph Sanborn

Right.

Jayme Mendal

Yes.

Joseph Sanborn

Yes, sure.

Jayme Mendal

So you mentioned the favorable backdrop, but we do expect that to persist with carrier underwriting as healthy as it has been, and I think most analysts seeing this soft market cycle extending at least into 2029. And so we feel good about the backdrop for next year.

Beyond that, I think we always think about it, there is distribution. And so how do we drive more carrier budget and improve pricing? And that typically is through, over the last couple of years, it has been driven through our AI bidding product, SmartCampaigns, which we continue to improve and drive greater adoption of. Most notably, this last quarter, we expanded it to a critical mass of local agents, and we will continue to roll that out into next year.

Then within the agent population, there are a number of incremental products that we are rolling out as we work to consolidate to be the one-stop growth partner for the local agent. And so we see those starting to get some scale moving into next year.

Then over onto the traffic side, as we get more budget, we are able to push that back and compete more aggressively in the traffic landscape. And we continue to expand into more channels, into un- and under-penetrated channels, some higher funnel channels, AI search, we expect to grow into next year. So we get more traffic to feed that distribution budget.

Joseph mentioned non-autos and our expectations for home next year. And then the other product, which we recently launched, is Luva. It is a new experience we think will allow us to begin to tap into a new segment of shoppers who are a bit more sort of research-oriented, AI-native shoppers that should help with growth next year as well.

Maria Ripps

Got it. Got it. That is very helpful. And then just in terms of sort of AI agents and your white paper. So as insurance sort of, excuse me, as insurance shopping sort of shifts towards AI agents, how should we think about sort of the economics of customer acquisition for your platform? I guess, could greater automation improve conversion and acquisition efficiency, or could we see sort of some pressure on the economics as AI platforms essentially take a larger role in driving demand?

Jayme Mendal

Yes. So I think when we think about the unit economics, I mean, to put it in context, right now it's negligible. It's immaterial in terms of its scale. We get less than 1% of our traffic from LLMs and even less from agentic shoppers. But it is likely to expect it will -- reasonable to expect it will grow over time.

So question number one is, for us in terms of unit economics is, what does it do to volume? Because you have the -- to the extent that agents can shop, reshop with greater frequency and less friction, we actually may see an increase in monetizable events as that happens. And then I think we have some -- I think the jury's out in terms of how exactly the economic model will evolve. But to the extent that these agents are able to do good research and improve the likelihood that the ultimate referral converts into a policy, we could see greater performance from the referrals going out and in turn, greater monetization.

Operator

Your next question comes from the line of Naved Khan with B. Riley.

Naved Khan

A couple of questions from me. Maybe the first one, just on the sort of -- we're hearing some changes to Google Search, in terms of Google increasing the coverage for AI Mode. And it seems like, looking at your numbers, it does not seem to have affected your numbers, or even for the -- looking at the guide for the Q4. So could you just talk about your ability to adapt to Google changes and how you look at the quality of the leads you get from there directly, and also the diversification in your own traffic? And then I have a follow-up.

Jayme Mendal

Sure. I think to begin, we have a very diversified traffic portfolio. Google is a large partner, but we have a very diversified base of traffic to begin with. And I think over the last few years, the segment of Google traffic that has been most impacted has been organic traffic, which, for better or worse, and right now it is for better, we have never had much exposure to organic traffic. So we have always been heavily concentrated with paid programmatic advertising. And we have managed to work through all the changes that have occurred with Google to date, and we expect that to persist moving forward.

Overall, we see shopping activity relatively stable this year at elevated levels, because I think the overall policy value and premiums just remain high. And so people are shopping and reshopping frequently as in response, I think, the cumulative effect of rate increases over time. And we have seen relative stability over the course of the year.

Naved Khan

That is good. Okay. And then I just wanted to touch on this partnership with Waniwani, I guess, is how you pronounce it. But I think there is some kind of exclusivity there. What does it cover exactly? Is it for the U.S.? Does it cover P&C or something broader? And what's the duration there? Also, with respect to any kind of revenue generated through such a partnership, how do the economics compare versus something that you get through other channels?

Jayme Mendal

Sure. So the basic premise of their offering is that they provide technology infrastructure and services to help companies of regulated industries distribute through AI channels. And so they will help build applications and then host infrastructure that provides related services, compliance services, insights, analytical services. And we have partnered with them on our market, effectively. So they serve multiple markets geographically and from a vertical standpoint. And the scope of our partnership is to help them distribute their offering to our customers in P&C insurance within the U.S. market. And then maybe Joseph can add.

Joseph Sanborn

It's a multi-year arrangement. Obviously, there's milestones as the relationship evolves. The economics, I think we were pleased with how the economics work from our partnership basis. As you might imagine, like any partnership, it's not a perfect answer exactly how it compares to current economics, it depends on how much we sell, but I think we feel very good in the opportunity and the partnership.

I think sort of stepping back from the specific arrangement, I think it puts in broader context of our strategy. As we think about bringing more products to carriers, more AI-first products to carriers and agents, how do we go about doing that? Some of those we will certainly be building ourselves when we're doing that. Some of those we'll find companies like Waniwani where it makes sense to partner, and we can do that as well. Sometimes acquisition we involve as well, and investment. So there's a whole range of possibilities, but Waniwani sort of fits into that strategy, just sort of to give you sort of the broader context.

Operator

Your next question comes from the line of Bryan Smilek with JPMorgan.

Unknown Analyst

This is Vignesh on for Bryan. I guess just as you look into 2027, you kind of touched on it, but just a little more color into how the conversations with carriers are kind of shaping up, and if you're seeing any changes in kind of appetite and what that broad-based strength kind of looks like. I know you mentioned 8 of your kind of top 10 customers also grew sequentially. So any color on that would be good.

Joseph Sanborn

Sure. So I guess as we -- maybe we'll start with how are we feeling right now? As I said in our prepared remarks, it's a healthy environment, and we expect that to continue not just into next year. We see it's a relatively early stages of a prolonged soft market cycle. And as you know, in insurance, a soft market cycle is good. The research we reference on our IR deck shows us going into 2029, several analysts. So we feel very good about that backdrop.

As you look at carriers right now, we see carriers broadly healthy, right? They want to grow policies in force. We saw in Q3, we had sequential growth in 8 of our 10 carriers, top 10 carriers. So we feel really good about that. So as we look into next year, we see that dynamic continuing. And so we see this favorable backdrop as being a very good time for EverQuote.

Unknown Analyst

Got it. And then I guess one follow-up on the launch of Luva. Can you kind of talk about maybe how you see monetizing the platform and any sort of maybe milestones to track into next year?

Jayme Mendal

Yes. So there's a number of ways we will monetize. Depending on the carrier relationship, some will be treated as commission, some will be paid effectively for performance on the placement of the policy. And there may be some referral-based models as well. So it will be a hybrid from a monetization standpoint, but that's really based on what suits the carrier for this channel.

So we launched -- we're all, what, 5 or 6 days old at this point. But some of the things that we'll be looking at as we progress into next year is obviously further development of the product in terms of the carrier panel, the capabilities, and then adoption, right? So this is really meant to access a segment of shopper that wants a different kind of comparison experience that is built around conversation, education, like really deeper understanding of their coverage and their policy options.

And it was designed from day 1 around the needs of consumers using LLMs and conversational AI, and built on top of infrastructure that can be leveraged to serve agentic commerce. And so we are going to be looking for arrival specifically through those channels to ensure that the sort of the intended segment that we are trying to serve is actually accessing the product and getting the value.

Operator

Your next question comes from the line of Jason Kreyer with Craig-Hallum.

Jason Kreyer

So can you just talk about the progression of the AI partnerships, just where you are at in terms of rolling out the partnerships that you talked about on the call? And then maybe what kind of a pipeline you are developing to extend to additional AI partners?

Jayme Mendal

Yes. Just to clarify, are you talking on the carrier side or on the platform side?

Jason Kreyer

More on the platform side, like you talked about Meta and Anthropic and OpenAI and stuff.

Jayme Mendal

Yes. So we have relationships across the board. I think that each of these platforms is still trying to learn their way into our category, particularly into like the distribution landscape within insurance. And so we've been having good, productive conversations with them to try and educate them about some of the unique characteristics of our market. And our expectation is that we'll be able to kind of build relationships that will sort of give us certain access and ability to really partner with them over time.

Jason Kreyer

Appreciate that. Maybe one for Joseph. So you called out the expectation for a downtick in VMM Q3 to Q4. You said that would be at least in part driven by the move into agentic. Curious what your perspective is over the next couple of years. Do you think that VMM -- like do you think those will remain stable, or is there potential for upside or downside to VMM as agentic becomes more and more a part of workflows?

Joseph Sanborn

Sure. So great question. So let me give you some context. So when you think about the investments we're making on our AI offerings, one of the areas those investments show up is in the traffic line. So again, different than some of the investments we've talked about, the AI investments previously, where they're showing up in driving efficiencies in the OpEx line, and we're able to drive a lot of increase in leverage through those investments in AI. Different with our AI and the investments we're making in agentic traffic sources. Just like any new traffic source, those new channels do not perform as well in the early days. And again, these particular traffic sources are new. They're still evolving for us and for the industry as a whole.

So as we look at that area, we're very bullish on the opportunity. You are seeing in Q4, we are making investments in these traffic channels, and Luva is part of that. Those will continue into the first part of next year. As we look at the VMM margin, what I would sort of say is I would look at Q4 as a reasonable proxy for next year as sort of a starting point. We will see how it evolves from there. But I guess that gives you a high-level view on it from the VMM margin level.

Operator

Your next question comes from the line of Jed Kelly with Oppenheimer.

Jed Kelly

Another nice quarter with a strong free cash flow. I think, ever since I've known you, so it's been almost 10 years, or maybe over 10 years, the company at its core has always been machine learning, AI. So it seems like you've got a pretty big advantage harnessing this technology relative to your competitors. How should we think about -- is there an opportunity here to sort of maybe consolidate the market, gain more distribution, accelerate share gains? How should we see that in terms of using this to maybe drive better unit economics from the carriers or gain more market share? Just can you talk about how we kind of think about your ability to use this solution and just accelerate share and drive better unit economics for the business?

Jayme Mendal

Yes. Thanks, Jed. So to your point, since the company's inception, I think one of the value adds we've been able to bring to carriers is being -- we tend to be kind of at the leading edge within our industry in terms of performance marketing and customer acquisition in digital channels, and through the use of whatever the latest technology and data available is. And that sort of represents -- that has driven the growth that has gotten us to this point.

And now, we are now in the early stages of the evolution to kind of the next phase of things. And I think the same pattern is likely to play out again, and we're really leaning into it, right? We are trying to stay very much in close contact with the AI platforms and with the carriers to help start to bridge the gap between what is a very fragmented, very opaque, and nuanced end market, and these platforms that really are just trying to deliver maximum value to their consumers. And I think if we can maintain a strong position in helping bridge that gap, there's going to be just a tremendous amount of opportunity along the way for us to build products and deliver traffic in a way that really benefits both sides of that marketplace.

Jed Kelly

Got it. And then just around capital allocation, I see you're implementing a share buyback. Get that just given the current valuation. But does it make sense maybe given where valuations are now for some of your other competitors, does it ever make sense to explore more M&A versus buying back stock where it's hard to see if the market's going to give you credit for that? Any thoughts on that?

Joseph Sanborn

Sure. Thanks, Jed. So when we think about capital allocation, it continues to be the same as we've said before, which is there's 3 ways we think about it, right? First is we want to make sure we have a fortress balance sheet. As we think about deepening our partnerships with our carrier partners, they will want to see a partner who's got depth on their balance sheet and will be here. And so I think that certainly is important to us, and it always has been, but even more important now as we think about these longer-term relationships.

The second is M&A and investments, right? We see this as ways -- we did the Waniwani as an example of an investment. We think about M&A as an opportunity as well, and M&A is something certainly we're looking at. But what form it will take, we will look. We will see over time. But again, whether it's investments or M&A, it's all about how we will build our TAM over time as the world evolves into this AI-first world. That'd be the second.

And buybacks, for us, we can do all 3. We have the cash flow generation that gives us the flexibility and capacity to all of the above. On buybacks, we see it as, it's opportunistic utilization. It also helps us offset employee dilution, which happens on an annual basis, and we think that's wholesome.

Operator

Your next question comes from the line of Gregory Peters with Raymond James.

Charles Peters

Well, good afternoon. I guess most of the questions that I had in mind have been asked. But maybe you can just remind me, during your commentary, you talked about the vendor services contract that was non-renewed. Maybe you could just remind me of the details behind that, what's going on there?

Joseph Sanborn

Sure, sure. So this was a relationship we entered about a year ago. It was sort of a new traffic opportunity, a branded traffic opportunity, and it was sort of a multi-year arrangement, a 5-year arrangement. As we've progressed into that arrangement, we realized it was not performing the way we liked. It was also distracting from other traffic operations. So we decided to end that particular contract. We still have a relationship with that brand, and we continue to work with them, but the exact arrangement didn't make sense.

And so it was a 5-year arrangement that we cut it short. We took the charge in the period. It was reflected in our adjustment to net income. This was sort of a very unique kind of situation for us. We had not done something like this before. And it was something we tried, and obviously, we reassessed when it wasn't working the way we expected.

Charles Peters

Can I go back to the SmartCampaigns? I guess, I'm trying to come up with this. I believe you're helping improve the conversion rates, and yet you say you're doing it, and your revenue results reflect it. But I don't know if there's some statistics you can add to your disclosure that will help us understand, hey, when the carriers use our SmartCampaigns, their conversion rates improve by 2x or 3x or something like that. But any color you can provide on that would be helpful.

Jayme Mendal

Yes. We've shared case studies in the past, Greg. I mean, it always depends on where the carrier starting point is. But in some of the case studies we've shared in the past, I think some of the numbers we've used have been double-digit or 20% performance improvements as a result of, in a controlled test, as a result of testing SmartCampaigns against their whatever legacy bidding they were doing prior to that.

Operator

There are no further questions at this time. I will now turn the call back to management for closing remarks.

Jayme Mendal

Thank you. And thanks, everyone, for joining us. To close out, I want to first thank the EverQuote team for delivering yet another excellent quarter with record financial performance. We continue executing our strategy, deepening customer relationships, and driving further innovation as we lead P&C insurance distribution into the AI era. We look forward to sharing further updates in the coming months as we build on this momentum. Thanks all.

Operator

This concludes today's call. Thank you for attending. You may now disconnect.