Philip Ghezelbash

Yes, agreed. It is a bit of an art. It should include a disclosure that it isn’t intended to be 100% accurate (which is impossible anyway), and that valuations will differ on a case-by-case basis, similar to most calculators provided by brokers or marketplaces. That said, it could work well if it incorporates data from generally accepted valuation models along with data from successful sales on TrustMRR to generate recommended ranges based on factors like MRR, profit, churn, industry, and other relevant metrics.

@philghz @marclou @riteshkv_ The problem is that you have both quantitative and qualitative data to measure and determine the valuation. Then, multiples also change based on previous exists. Valuation is a bit of art.

what would you say is the sale price for this on TrustMRR? I'm thinking $70k @marclou 👀

@edwardeachday Super ironic

ohhhhhh myyyyyyyy gossshhhh WOW SEO just surpassed its recent ATH in google trends interest this is insane

@timsoulo It just comes down to perceived growth potential. I mean, look at the stock market. There are countless companies with irrational valuations. That's just how it goes.

So Profound was valued at $1B… while Similarweb is valued at just ~$158M ($229.58M market cap minus $72M in cash). Make it make sense. One is essentially a commodity product, competing with a horde of startups offering very similar tools: AirOps, Peec AI, Scrunch, Otterly, Promptwatch, UseHall, LLMrefs, LLMpulse, Promptingcompany, Xfunnel, Evertune… and many-many more. The other has a strong moat built on proprietary data... some of which they allegedly (and ironically) license to Profound. Something clearly doesn't add up in these two valuations. But most importantly... Since Ahrefs' Brand Radar is a much better product than Profound (at least in my assessment)… does that mean we get to add +$1B to Ahrefs' (hypothetical) valuation? 🙃

@gaganghotra_ Ahrefs is superior in every way, bootstrapped, revenue probably much higher, quality and experience of the team, better and bigger data. Not even close. Perhaps you could make an arguement for the design of Profounds website which is nice.

per LinkedIn Ahrefs - 194 employees Profound - 168 employees Profound probably in next few months will overtake Ahrefs for head count 👀 Ahrefs do over $100M ARR I wonder how much ARR Profound is doing? #SEO

@alex_buraks AI for sure, I misread and thought "Google vs other LLMs". Still, I prefer Google in all cases, as if the AIO doesn't answer my question, I can quickly scroll to classic results. How about you?

@philghz Think we need to decide about AIO - is that classical SEO or AI-SEO?

If you need to buy a new product but do not know exactly what you need, where would you go first, classic Google or AI search?

@sekhsoyebali @framer Super exciting for Framer

Big News for the @Framer crowd 👀 Looks like dynamic CMS filtering might finally be coming natively. Framer University just showed a working filter setup built without any code components or overrides, which strongly hints that an official dynamic filter feature is on the way. If this lands, it's going to make CMS-driven sites way more powerful and flexible. I'm seriously excited to see this roll out. Who else has been waiting for this. 👀

@gaganghotra_ Why? Not saying I disagree

don't use AirOps AI content workflows specifically if you are a medium size business or a large Enterprise you will get wrecked in search rankings

@_JoshuaSEO I used to think this but it’s not true. I’d change it to “if an SEO agency can’t show you recent successful case studies, don’t hire them.”

If an SEO agency can’t rank their own site, don’t let them touch yours. I don’t care how many certifications they have or which agency they worked at. If they can’t rank their own website, why would you trust them to rank yours?

@lilyraynyc Lol had to pick an acronym that’s hard to rank for. Classic

lol, chatGPT still doesn’t fully get what “GEO” means in this context, even with web search enabled. Almost like it’s a dumb name for this 🤭

@piotrkulpinski Let’s go!

Spotted in the AI summary as well 🎉 Domain switch has been a great call so far

@Charles_SEO Pretty much this. The best case scenario is that you create momentum for your brand where other people start to do this for you at scale when creating their own content and recommendations.

AI SEO isn't THAT difficult! and if you're not showing up when someone asks ChatGPT, Gemini or Perplexity about your niche.. you're already behind and it's only going to get worse! AI models don't rank pages, they cite sources, and the sources they cite are usually predictable - They pull from the same trusted, weighted content clusters (or literally just set search results) every time. Which means you can reverse engineer exactly WHERE AI models get their answers and inject your brand into those sources. Here's the framework: - Run hundreds of prompt variants related to your niche across ChatGPT, Claude, Gemini, Perplexity (or use a tool to do it for you, we use Amplitude & Ahrefs, no aff) - Extract which URLs, brands, and domains get cited repeatedly - Map the source clusters, you'll notice the same 10-50 pages usually appear across 80%+ of prompts - Get your brand mentioned, linked, or featured on those exact sources, or domains that are closely related to them (or even better, both) - PR placements, UGC posts, Reddit threads, Medium articles, niche directories, whatever it takes to get into the citation pool This is seeding your brand into the training data and grounded sources that AI models actively pull from - We did this for PressWhizz. 10 months old and it was already a top 5 recommended link building marketplace inside ChatGPT. That didn't happen by accident! Google AI Overviews work the same way. If you're in the sources they pull from, you show up. If you're not, you don't exist. Traditional SEO gets you Google rankings. AI Visibility Seeding gets you recommended by the AI itself. The companies doing this now will own the next decade, the ones waiting will be asking "what happened to our traffic?" in 18 months.

@ctwtn Yea this is facts, actual advice that works

There are two types of SEO. Traffic SEO: → High volume → Easy to publish → Looks great in reports → Rarely drives pipeline Revenue SEO: → Pricing pages → “[Product] vs [Competitor]” → Alternatives pages → “Best [category] for [use case]” → Industry-specific pages These aren’t blog posts. They’re buying decisions. Most teams start at the top of funnel because it feels safer. But if you want pipeline impact… Start BOFU-first. Traffic is optional. Purchase intent isn’t.

@forgebitz Ah true, like Bubble? Makes sense then

@philghz i see website builders a bit different compared to no-code saas tools

ironically the first thing ai killed was no-code everyone is now coding

Remote100K has been acquired! 🎉 In 2025, I built remote100k.com to solve a personal problem: I couldn't find a dedicated website for remote positions paying $100K+. I was looking for a remote role myself, so I built it as a side project to test SEO strategies while building something useful. What happened: - Idea and planning in May 2025 - Soft launch in June 2025 - Public launch in July 2025 - Grew to 60k+ monthly active users in 7 months - $0 advertising spend (100% organic) - Fell in love with @framer (huge shout out!) - Caught the attention of a buyer What I learned: - Nothing beats building from scratch. To get better at SEO/organic growth, there's no shortcut — just start something and grow it. - Solving your own problem often solves it for others. The site's in great hands with its new owner, and I'm excited to see where they take it. What’s next: I'm exploring new opportunities and projects in the SEO/growth space. If you want to connect, you can find me at philipghezelbash.com Thanks for your support!

@ViperChill Yep! Definitely interesting, a monthly update would be great. Thanks for putting this together. Partnerships aside, what do you think causes certain publishers to perform disproportionately well in specific LLMs?

These are the publishers currently dominating AI search. A few things stood out to me when putting this chart together. 1. Condé Nast performs much better in ChatGPT than other platforms. (They didn't show up at all in Copilot or Gemini's top domains.) On the surface it seems like striking a deal with OpenAI significantly helped, but perhaps not, as Valnet didn't seem to need one. 2. These ten publishers were most prominent in Gemini (123 of the top 1,000 domains), followed by ChatGPT (117 of the top 1,000 domains). 3. Ziff Davis' ChatGPT visibility is much lower than the brands they're surrounded by. Then again, they're suing OpenAI, and also blocking their bots in different ways across various sites. 4. Some companies that just missed the cut include News Corp, Internet Brands, The Arena Group, and Axel Springer. 5. People Inc. (formerly Dotdash Meredith) is clearly the most prominent publisher overall. I actually track more sites for some other brands, so it's not just about their scale. ** If you've followed me for a while, you've probably seen me cover many of these companies before, sharing how they dominate Google's traditional search results. Outside of any deals you can strike up and possible new advertising partnerships, you have to imagine there's significantly less value in dominating AI search. Please let me know if this is interesting and I'll see if I can get the green light to make it a monthly update. 🙏 P.S. I added partnership icons to hopefully make the chart more insightful. That said, I should make it clear I don't know exactly what each of these entail outside of news reports covering them.

@DavidGQuaid Nice post @DavidGQuaid! Not much to add, but I will say that @rustybrick is definitely worth following for anyone who wants to stay up to date with search! 😁

LLMs.txt isn't new. Its symptomatic. People want to believe in a magic key, a secret button, a hidden switch that "makes" you visible. SEO isn't complex. SEO isn't impossible. SEO is a struggle, especially if you cannot separate reality from what needs to be done SEO can be hard work. SEO can be routine. @rustybrick does a great job in highlighting this in his quote from @johnmu about consistency being critical in Technical SEO.... SEO can be a challenge - from time to time. Thats why its important for SEO thought leaders to help SEO newbies understand theory from practicality. Its why we have to protect communities from branded disinformation and propaganda. Things like saying AI/LLM tools are search engines and have the ability to have their own search ranking model is complete nonsense. So is making up SEO myths and superstitions. So is making up SEO checklists when we're clearly in a system. I'm not here to predict the future of SEO. Thats up to search engines like Google, bing, DDG, whoever. But I will challenge SEO and GEO FUD. cc @edwardeachday

The goal of SEO isn't to increase DR. So many people new to SEO make this their goal. "I've increased my DR!" "My DR went from 10 to 30!" SEO is a marketing channel that you can use to drive business outcomes: signups, demos, revenue, etc. It is so easy to take a website from 10 to 60 DR within a few days by paying someone on Fiverr to build some links to the site (not recommended lol). Is it probably better to get a link from an article on a DR 90 blog vs. a DR 30 blog (all else equal: relevance, quality, traffic, etc.)? Probably yeah! But there are many examples out there where domain rating is increasing while impressions, clicks, and most importantly, sales are falling. I'm not saying DR is totally useless; I like it and I really like Ahrefs. The metric has its place, but it shouldn't be the primary metric for "SEO success".