Eugene Levin, who ran most of Semrush's acquisitions and was the company's president when Adobe agreed to buy it for $1.9 billion, says the businesses he now wants to acquire must grow at least 10% a year, and that whether they make money matters far less to him.
Levin set out those terms on the Ahrefs Podcast, in an episode published on YouTube on November 10, 2026. The host was Tim Soulo, chief marketing officer of Ahrefs, the SEO software company that Soulo describes as Semrush's main competitor. Soulo today promoted the episode in a thread on X that opened with Adobe's $1.9 billion purchase and then added a line of his own: "And yet Adobe never called us."
The conversation runs just over 64 minutes across eight chapters. It moves from the Adobe deal, to the ten companies Semrush bought while Levin was there, to the holding company he is now building to buy small software businesses. Along the way Levin described a 2021 purchase he considers a mistake, a German crawler Semrush expected to quadruple, and the valuation method Semrush used for Backlinko, in which the founder's personal brand was counted as worth nothing.
In Short
Eugene Levin, the former Semrush president who led most of its acquisitions, told the Ahrefs Podcast what he looks for now that he is buying marketing software companies on his own: $3 million to $10 million in yearly recurring revenue, growth of 10% or more, and access to the seller's billing records. It matters to founders of small, bootstrapped tools because Levin said those companies are harder to buy than a large public firm like Semrush, since so much depends on a few people, and because he said such businesses now trade at two or three times their recurring revenue, while some listed software firms have been cheaper still. The practical message from the interview is that a business valued on what keeps running without its founder sells more easily than one built on the founder's name.
Why Adobe went to Semrush
Adobe's interest, as Levin told it, arrived unannounced. A couple of months before the first conversation with Adobe's corporate development team, Semrush had held a strategy session. "We knew what to do for next two years. We had our own acquisition roadmap," he said. Bill Wagner had just joined as chief executive, and Semrush's AI optimization products were growing quickly.
"It was completely out of nowhere inbound interest," Levin said, adding that the process is documented: "If you know where to look, there are filings where you can see the whole process."
His view of what followed supplied the line the episode opens with. "You never wanna be looking to sell. You wanna have a plan," Levin said. "Follow the plan, and then if someone offers you something that makes sense, of course you need to work on that." For a chief executive, he argued, refusing to engage is not an option: "That's an obligation. It's a fiduciary duty."
PPC Land reported the deal on November 19, 2025: $1.9 billion in cash, or $12 per share, for a company whose market value had been slightly above $1 billion before the announcement. Adobe completed the purchase on April 28, 2026. Today is the first anniversary of the announcement.
| Adobe-Semrush deal | Figure |
|---|---|
| Announced | November 19, 2025 |
| Deal value | $1.9 billion, all cash |
| Price per share | $12 |
| Semrush market value before the announcement | Slightly above $1 billion |
| Expected closing at announcement | First half of 2026 |
| Completed | April 28, 2026 |
| Semrush ARR at June 30, 2025 | $435.3 million |
| Deal value as a multiple of June 2025 ARR | About 4.4 times (PPC Land calculation, not adjusted for cash or debt) |
Sources: PPC Land coverage of the deal, the completion and Semrush's second-quarter 2025 results.
Why Adobe wanted the company at all is, in Levin's account, a story about one category. Adobe, he said, is the largest martech company in the world but assembled much of its portfolio through acquisitions such as Marketo and Day Software. What it lacked was a position in the field that the industry variously calls agentic search optimization, answer engine optimization or GEO. "No matter how you call it, I think it is the hottest thing in MarTech right now. I think it's the fastest growing vertical," he said. "And so Adobe didn't wanna miss the party."
Adobe was also building its own tools; PPC Land covered the general availability of Adobe LLM Optimizer in October 2025, a month before the Semrush agreement. Levin said the company was "not shy" to build and to buy at the same time.
He also pointed out that some of Adobe's best-known products did not start inside the company. "Even some of the most famous Adobe products, like Photoshop, those were acquired as well," Levin said. "It just happened so long ago that nobody remembers."
The integration thesis, as Levin described it, is a closed loop. "The vision is really to close the loop and to combine the AI optimization and SEO power of Semrush with their content management platforms and their personalization products," he said. Semrush would act as an engine of insights about which pages "get citations, which pages gain mentions or backlinks", and a website could then adjust itself, for example by changing internal links almost in real time. That matters little for a 20-page site, he said, but a great deal for e-commerce or classifieds sites with tens of thousands of pages. On distribution, he was blunter: "If you wanna sell to Fortune 500 brands, there is no better place to be."
Three or four companies, and the one that was not called
Adobe had been watching to see who would emerge as the leader. "At the end, there were maybe three, four companies that really stood out in this space, and Semrush was one of them," Levin said.
Soulo questioned the count. "It's just Semrush and Ahrefs, and we are about three times smaller than you are," he said, conceding that "there's also Similarweb that's quite sizable." He told Levin that he did not believe Adobe had approached Ahrefs at all: "I don't think Adobe even reached out to us." Ahrefs, he said, is bootstrapped, and "famously Dmitry doesn't even want to sell", a reference to the company's founder.
Levin said his "three, four" included newer, venture-funded companies focused on parts of the problem such as prompt tracking, which he declined to name. On Ahrefs, he hedged: "I don't know for sure, but I am sure that you were one of the players that were considered." He described Ahrefs' refusal to sell as good for the market. "I think the world needs companies that run businesses differently and not go and just follow, you know, VC playbook," he said, contrasting Semrush, which he said had been "a bit corrupted by the size of the opportunity".
Ahrefs has been in PPC Land's coverage on its own terms this year. In February its researchers found that AI Overviews cut clicks to top-ranked pages by 58%, and in October the company said Google was poisoning 5% to 20% of the results served to SEO data tools.
Why a small company is harder to buy
Soulo's instinct was that Ahrefs, with slightly over 100 people and no acquisitions of its own, would be simpler to acquire than Semrush, a public company with more than a thousand employees. Levin disagreed, and the reasoning became the thread that runs through the rest of the episode.
"Smaller companies have much higher dependency on key people, right?" he said. "And so, if you're a buyer, you at the mercy of those people staying." A sale makes those people wealthy, he noted, and the money removes their need to work for a new owner. "It's very hard to make some of those situations work."
Semrush, in contrast, had a layer of professional managers. "We just sold the company and I'm not there anymore, and business still runs well," Levin said. "But that took us five years to really build those processes and put everything in place and get the right people, build the right incentives for them." At a company of 30 to 50 people, a buyer has to work out who is essential and "make sure that, let's say, R&D team doesn't just walk away with the founder", or pay "crazy incentives for people to stay, which makes transaction more expensive."
Soulo summarised it in his X thread: "Turns out, being small, private and founder-driven is actually riskier for a buyer compared to acquiring a public company with a thousand employees."
Semrush's acquisitions as capital allocation
Soulo estimated that the companies Semrush bought cost between $2 million and $15 million each, and between $50 million and $70 million in total. Levin did not confirm those figures. He did give the count: "So we bought seven software companies, but we also bought three media companies." The episode introduction lists Datos, Prowly, Brand24, Ryte and Exploding Topics; Backlinko comes up later. Semrush's media purchases had already been in the news: PPC Land reported in October 2024 that Semrush acquired Third Door Media, the publisher of Search Engine Land.
The approach was, in Levin's words, "a little bit unorthodox". "We saw it more as a capital allocation exercise rather than anything else," he said. He contrasted it with transformative deals such as Facebook's purchase of Instagram, which he called one of the best deals ever, and with Yahoo's acquisitions, which he said never worked. Deals of that kind, he argued, require a home run: "So it's more like venture capital investing than capital allocation in a more traditional sense."
Semrush was cash flow positive most of the time, Levin said, and as a public company it was restricted in what it could do with surplus cash. Investing in public markets at scale would have required registering under a different regime. "If you're interested, it's called Investment Company Act, right?" he said. "And without that, you can only go and buy bonds."
His personal test for every deal was whether he would buy the asset with his own money. "I never brought the deal to the board that I was not willing to go and do myself," Levin said.
Some acquisitions replaced products Semrush had built. Its own social media tools, he said, "were never, you know, quite good", while the Polish companies Prowly and Brand24 "tremendously improved our portfolio". In digital PR, "most of our portfolio is acquired."
Ryte and the German crawler
Ryte was the clearest case of a thesis built on distribution. Semrush's own technical SEO crawler, Levin said, worked best below 1,000 pages and became slow beyond about 10,000. "Almost any crawler that you can buy with a credit card is going to be slow for a 10,000 page website," he said. Rebuilding the backend would have taken time, as it had when Semrush rewrote its rank tracker for large sites.
Ryte had a strong crawler but was largely confined to Germany, where Levin estimated it held about 40% of its market. "Germans can buy any crawler," he said, describing them as sophisticated buyers who were not price sensitive. The calculation was what the same share would be worth in the United States, the United Kingdom and Semrush's other markets. "In our models we're expecting to, you know, eventually quadruple the size of that business through cross-sell," he said.
Without Semrush, he said, Ryte faced structural pressure: research and development costs to stay competitive, and "expensive infrastructure on AWS", without the volume "to get real discounts, so that creates pressure for gross profit margin."
Other deals worked without any cross-selling. Levin cited one business bought at $1.5 million in annual recurring revenue, or ARR, that grew to $5 million as a standalone unit before Semrush added cross-sell on top. He did not name it.
The deal that cost too much
Of roughly ten deals, Levin said, "nine out of 10 were good, one was not that good." The weak one was a 2021 purchase, which he also did not name. The problem, he said, "was not that it was a bad business or bad product, it was bad timing."
Semrush paid about seven times ARR, at a time when, by his account, Semrush itself was valued at around 14 to 15 times ARR. "When businesses trade for three times, two times, and I bought that company for seven," he said, the return no longer looked good, even though Semrush improved the business's profitability considerably. "If I bought it for three times less, I would be fine."
| Semrush deal record, as described by Levin | Detail |
|---|---|
| Number of acquisitions | Ten: seven software companies and three media companies |
| Outcome | Nine performed well, one did not |
| Underperforming deal | Bought in 2021 at about 7 times ARR; unnamed |
| Typical price range | $2 million to $15 million per deal (Soulo's estimate, not confirmed by Levin) |
| Total spend | $50 million to $70 million (Soulo's estimate, not confirmed by Levin) |
| Ryte | About 40% of its German market, by Levin's estimate; Semrush modelled a fourfold increase through cross-sell |
| Unnamed standalone deal | $1.5 million ARR at purchase, $5 million ARR without cross-sell |
| Backlinko | Valued on traffic and expected assisted conversions; founder brand counted as zero |
| Founder who sold twice | Brian Dean (Backlinko, then Exploding Topics) |
What earns a higher multiple now
Soulo opened the chapter on valuations with an anecdote: Ahrefs' founder had fed Slack into an agentic loop, and in a few weeks the agents "built a pretty decent copy of Slack for, I don't know, $20,000 in tokens or something like this." The claim was Soulo's and was not detailed further. "So creating software is super easy," he said, and asked what a seller is still selling.
Levin's answer started with the buyer's alternatives. "Getting higher multiples is very hard because investors have so many great options," he said. Public companies require no full takeover and carry no liquidity worries. "This year, for example, there was a time when Similarweb have traded for less than one time ARR multiple," he said, adding that the figure needs adjusting for cash on the balance sheet and that the company has since recovered.
"And so how will you convince someone that your $3 million, $5 million ARR should trade at a very significant premium to Salesforce, to Adobe, to Similarweb, and so on? It's very hard, right?"
He listed what can still justify a premium: much faster growth, such as doubling year over year at $3 million ARR; unusually high profitability, which he said Ahrefs has; exceptional net revenue retention; and extraordinary sales efficiency, such as a product that spreads by word of mouth. "If you have great sales efficiency, great net revenue retention, growth is guaranteed," Levin said. "Growth is a lagging indicator from those, right?"
Investors, he added, remain "very concerned about AI disruption", which he said has hurt a range of businesses. He named Zendesk as an example of a company whose owners "were very smart to sell when they sold".
The applied-AI holding company
Levin's new project has a working description rather than a public name. "We call it Applied AI Holding Company," he said. The premise is that many companies with product-market fit and defensible revenue are run inefficiently, not because their managers are bad but because they were built under different assumptions. "Businesses that were built, you know, in pre-AI era, they're just built fundamentally differently than the way you would build a new business in a current era," he said.
Rather than build from scratch, the plan is to buy those businesses and rework them. Levin said Semrush had already done some of this with acquired companies: "In some cases, we were able to run millions of dollars in ARR with maybe one or two people, right?" He is starting in martech and data, and said the same approach could later extend to service businesses, including agencies. "I think future of marketing agencies is largely in that direction," he said.
There is no website yet. Levin said the easiest way to reach him is a message on LinkedIn, which he said he always reads.
How Levin's screening works
As Levin described it, his assessment of a company covers seven points.
- Size. "Ideally over $3 million ARR, probably all the way up to 10." He said he would consider companies above $10 million but not yet ones closer to $100 million. "So $3 to $10 million ARR I think is a sweet spot."
- Defensibility. The team applies its own criteria to judge whether the business will still be valuable in five to ten years. Levin did not list them in the episode.
- Billing data. "We usually just ask people to give us their billing information, anonymized, whatever, you know, just transactions." From those records, he said, the team calculates the metrics it needs.
- Growth. Growth of 10% or more. His reasoning is that a company below that cannot keep up with inflation or give staff cost-of-living raises. "So I think 10% growth is a minimal for a healthy business."
- Profitability. "Profitability doesn't really matter that much to us." Running at about breakeven or a 10% margin is "totally fine", he said, and the team expects to take such businesses "to, you know, 30, 40% margins" through growth and changes in how capital is allocated, such as moving spend to more efficient marketing channels.
- Deal structure. Incentive packages for founders and staff who stay, tied to growth: "the faster you grow, the more money you get." In some cases payments are tied to research and development milestones, such as migrating a product to a new platform.
- Founder dependence. Value attached to a founder's personal brand is, in the Backlinko precedent he described, set to zero.
| Levin's criteria | What he said |
|---|---|
| ARR | $3 million to $10 million preferred; above $10 million considered |
| Growth | 10% a year or more |
| Profitability | Breakeven to 10% margin acceptable; target 30% to 40% after acquisition |
| Data requested | Anonymised billing transactions |
| Time horizon for defensibility | Five to ten years |
| Seller type | Bootstrapped or venture-backed, "it kind of doesn't matter to me that much" |
| Earn-outs | No traditional earn-outs paid to the whole cap table |
| Portfolio size | About 30 sizable, uncorrelated bets for adequate diversification |
| Contact | LinkedIn; no website yet |
Bootstrapped founders, lawyers and venture investors
Levin said he has bought companies from both bootstrapped founders and venture-backed teams, and that each brings a different problem.
Bootstrapped founders, he said, often lack experience with standard deal terms. The friction shows up in the section of the purchase agreement covering representations and warranties, which Levin called "reps and warrants": the seller's formal statements that, for example, the financial reporting is correct and all material agreements have been disclosed. "And people get really scared of that, even though this is absolutely standard part," he said.
He blamed some failed deals on advisers. One founder's lawyer, he said, "probably thought that if he would kind of boost their paranoia, there will be more legal fees." After Semrush sent its redlines, the other side would "revert everything", and Semrush walked away. "We said, 'We're done.' You know, we're out."
He was still complimentary. "I think they have much better operational discipline," he said of bootstrapped founders, describing businesses where "every dollar is accounted for". Semrush itself, he said, "was largely bootstrapped": it raised venture money but "didn't really touch that money" and still held the cash when it went public.
Venture investors bring structure and, ultimately, want to sell, which helps transactions close. But Levin recounted deals in which investors agreed a valuation and then backed out at the end because "this other company in completely different space just raised at this crazy valuation, so now we don't wanna sell." He expects that to change. Record amounts of capital went into venture capital and private equity in 2021 and 2022, he said, and those funds now need to return it. "So I think you'll have a tremendous number of VC-backed companies trying to sell."
Why a portfolio rather than one company
Soulo asked why Levin would not simply buy one strong company and turn it into the next Semrush. Levin said the model he follows is the holding company, naming Berkshire Hathaway as the best-known example and Bending Spoons as a newer one in the consumer space. The method is repetition: buy, improve operating efficiency, and use the new cash to buy more. "So it kind of rinse and repeat," he said. A single purchase, in his view, "would be more of a PE play", typically with a private equity sponsor and borrowed money.
Diversification is part of the logic. "No matter how good you are at buying companies, you know, some of them will fail," he said. Returns vary widely: one company might deliver "20% IRR. It's better than buying bonds, but it's not extraordinary," while another "triples in two years and increases margin five times." He put the useful limit of diversification at "about 30 sizable uncorrelated bets".
He would not name companies he would like to buy. For businesses not for sale, his approach is to stay in touch: "I'm just friendly neighborhood Eugene." And for each deal he does, he said, "to say yes one time in this space, you have to say no 100 times."
Semrush's buying spree, Soulo noted, became something of a meme in the SEO industry. Levin welcomed it. "That's the best PR I could have," he said, adding that the best references came from people who had already sold to him. Brian Dean sold Semrush two companies, and for the second, Levin recalled, Dean told Semrush: "Guys, if you're interested, you know, give me a number. I would rather sell it to you than go shop around."
Earn-outs, milestones and the founder's face
On structure, Levin drew a line between rewarding people who stay and rewarding everyone on the cap table. "We don't like traditional earn-outs that go to the entire cap table," he said, arguing that an investor no longer involved in the business should not be paid for its later success inside Semrush. Earn-outs are sometimes used to bridge a gap between what investors want and what a buyer will pay. "I don't really like it," he said, preferring to price the most probable outcome fairly.
Incentive packages for the team are different, because "growth for me is the most important predictor of value down the road." Their length depends on the founders' plans, whether they want to stay a year, two years or indefinitely.
The hardest case is a business whose value sits in one person. "I think it's very hard to buy a company where most of the value is connected to personal brand," Levin said, and he generally advises against such deals. Semrush did one anyway with Backlinko. "When we bought Backlinko, tons of value were attached just to Brian Dean himself," he said. "Back then we couldn't do AI avatar of Brian Dean to do videos."
Semrush's solution was to ignore the brand in the price. "We valued the asset primarily based on traffic and what we called expected assisted conversions," Levin said, meaning the new customers the asset would bring to Semrush. As for the value of Dean's name: "We had to multiply it by zero and that's it." A separate, contingent agreement had Dean shoot new videos for a year, which he did.
The advice Levin said Semrush gave Dean when it bought Backlinko was specific: "Listen, you do next one, just make sure this thing does not depend on your face." Dean's next company, Exploding Topics, drew its traffic through research content. "People had no idea who's Brian Dean," Levin said, which made the business "much easier for anyone to buy." Soulo's X thread gives the advice as "Don't build your next business around your personal brand", a paraphrase presented as a quotation; the words above are Levin's in the episode.
Asked whether a founder could rent a personal brand to a buyer for a period after the sale, Levin was doubtful. "I think you will find that universe of buyers is very thin."
Why this matters for marketers
For the marketing software sector, the episode is an unusually direct account from the buy side of how one of its most active acquirers priced deals. The numbers Levin gave place most small tools in a weak position. If businesses now change hands at two or three times ARR, as he said, and a listed company such as Similarweb traded for a period below one times ARR, a $5 million company asking for a premium has to show growth, retention or efficiency that justifies it.
The Adobe-Semrush deal also continues to shape the tools marketers use for search and AI visibility. Since the deal closed, Semrush has added a native Claude connector and announced that it would drop Microsoft Copilot from its AI visibility tracking on November 9. Soulo drew his own conclusion from Levin's answers: Ahrefs should not expect Semrush to fade under Adobe's ownership. "So we have to stay competitive," he said.
Levin's comments also put a figure on a concern that runs through SEO and content businesses built around a single expert. A brand that brings in traffic may be worth nothing to an acquirer that cannot keep the person behind it.
What the conversation leaves open
Levin named neither the 2021 deal nor the business that grew from $1.5 million to $5 million in ARR, gave no total for Semrush's acquisition spending, and did not list his defensibility criteria. His holding company has no published name, funding details or website. Whether Adobe approached Ahrefs remains unresolved: Soulo said he did not believe it had, and Levin said he believed Ahrefs had been considered but did not know.
Key facts
- Show: Ahrefs Podcast (YouTube channel "Ahrefs Podcast")
- Episode title on YouTube: "How to Sell a SaaS Company: What Buyers Look For | Eugene Levin (ex-Semrush)"
- Title shown on the episode link in Soulo's X thread: "What Makes a Company Worth Buying? | Eugene Levin (ex-Semrush)"
- Published on YouTube: November 10, 2026
- Length: just over 64 minutes, eight chapters
- Host: Tim Soulo, chief marketing officer, Ahrefs
- Guest: Eugene Levin, former president of Semrush
- Video: https://www.youtube.com/watch?v=CBgbmz9OKt8
- X thread by Tim Soulo: posted November 19, 2026, https://x.com/timsoulo/status/2123404676163076241
- Adobe-Semrush: announced November 19, 2025; $1.9 billion in cash, $12 per share; completed April 28, 2026
- Levin's new venture: described as an "Applied AI Holding Company"; no public name or website at publication of the episode
| Timestamp | Chapter |
|---|---|
| 00:00 | Intro |
| 01:32 | Why Did Adobe Buy Semrush? |
| 16:30 | What to Look for in an Acquisition |
| 33:04 | What Earns a Higher Multiple |
| 37:47 | Criteria for Buying a Company |
| 45:47 | Bootstrapped vs. VC-Backed Sellers |
| 50:54 | Why a Portfolio, Not One Company |
| 57:37 | How to Structure a Founder Deal |
Timeline
- 2021 - Semrush buys a company at about seven times ARR, the one deal Levin says underperformed
- October 16, 2024 - Semrush acquires Third Door Media, publisher of Search Engine Land
- August 2025 - Semrush reports second-quarter 2025 revenue of $108.9 million and ARR of $435.3 million
- October 14, 2025 - Adobe makes LLM Optimizer generally available
- November 19, 2025 - Adobe agrees to acquire Semrush for $1.9 billion, or $12 per share
- February 2026 - Ahrefs research finds AI Overviews cut clicks to top-ranked pages by 58%
- April 28, 2026 - Adobe completes the Semrush acquisition
- August 26, 2026 - Semrush publishes a native Claude connector
- October 22, 2026 - Ahrefs says Google poisons 5% to 20% of results served to SEO data tools
- November 9, 2026 - Semrush stops tracking Microsoft Copilot in its AI visibility tools
- November 10, 2026 - Ahrefs Podcast publishes the Eugene Levin episode on YouTube
- November 19, 2026 - Tim Soulo promotes the episode in a thread on X, one year after the Adobe-Semrush announcement
Related PPC Land coverage
- Adobe acquires Semrush for $1.9 billion to expand brand visibility tools - The November 2025 announcement of the all-cash deal at $12 per share.
- Semrush drops Copilot from AI visibility tracking on November 9 - A product change after Adobe's ownership, including the April 28, 2026 completion date.
- Semrush gains native Claude connector covering 28.8 billion keywords - Semrush's keyword, backlink and prompt data made available inside Claude for Semrush One and SEO Classic subscribers.
- Semrush posts 20% revenue growth as AI products accelerate adoption - Second-quarter 2025 results, with $435.3 million in ARR.
- Semrush acquires Search Engine Land - The October 2024 purchase of Third Door Media and its publications and events.
- Adobe launches enterprise tool for AI visibility optimization - Adobe LLM Optimizer, the AI visibility product Adobe built before buying Semrush.
- Ahrefs says Google poisons 5% to 20% of results served to SEO data tools - Ahrefs' account of altered search results reaching data providers.
- Google's AI summaries now swallow 58% of clicks that once went to websites - Ahrefs research on the click-through effect of AI Overviews.
Summary
Who: Eugene Levin, former president of Semrush, interviewed by Tim Soulo, chief marketing officer of Ahrefs, on the Ahrefs Podcast.
What: Levin explained why Adobe bought Semrush for $1.9 billion, how Semrush treated its ten acquisitions as capital allocation, why small founder-run companies are riskier for buyers, and the criteria for his new applied-AI holding company: $3 million to $10 million in ARR, at least 10% annual growth, breakeven profitability as acceptable, billing data on request, and no value for a founder's personal brand.
When: The episode was published on YouTube on November 10, 2026. Soulo promoted it on X today, November 19, 2026, one year after Adobe announced the Semrush deal, which closed on April 28, 2026.
Where: On the Ahrefs Podcast YouTube channel, with a promotional thread on X.
Why: The interview gives founders of small marketing software companies a buyer's account of what lowers or raises a valuation, at a time when Levin says businesses trade at two or three times ARR and AI has made software cheaper to build.
Discussion