Profound’s $180M Raise Says SEO Is Finished as You Knew It
Google rankings are no longer enough. Profound just raised $180 million because brands now need to win the answer ChatGPT gives before a buyer ever sees a link.
Google search is not dying. But the bit that made a lot of marketers feel useful — fiddling with rankings, keywords and dashboards while calling it strategy — is getting belted.
Profound raised $180 million at a $1.8 billion valuation on September 15, because the next battleground is not whether your company appears on a page of links. It is whether ChatGPT, Gemini and the rest mention you when a buyer asks a question. ([bloomberg.com](https://bloomberg.com/news/articles/2026-09-15/profound-hits-1-8-billion-value-to-boost-brands-in-ai-search?trk=article-ssr-frontend-pulse_little-text-block&utm_source=openai))
That should make every founder and marketing chief slightly uncomfortable. It should also make them stop treating AI search as a side project run by the intern who used to do SEO.
The core story: $180 million for being recommended by a machine
New York-based Profound announced a Series D led by Sequoia Capital and Kleiner Perkins. The company says it helps brands understand how they appear in AI-generated answers, then helps them create and manage the marketing work intended to improve that visibility. It was founded in 2024. ([tryprofound.com](https://www.tryprofound.com/newsroom/profound-raises-usd180m-series-d-at-usd1-8b-valuation-to-build-the-ai-platform-for-marketing-teams?utm_source=openai))
The valuation is the bit that matters. Profound was valued at $1 billion less than seven months earlier when it raised a $96 million Series C. Now it is worth $1.8 billion, according to the latest round. That is not normal venture-capital froth over another chatbot with a trendy logo. It is a very expensive wager that product discovery is moving from search results to conversational recommendations. ([techcrunch.com](https://techcrunch.com/2026/09/15/aeo-startup-profound-hits-unicorn-valuation-raises-180m-series-d-7-months-after-last-round/?trk=public_post_comment-text&utm_source=openai))
Profound says it has more than 1,000 enterprise customers, including Comcast, The Estée Lauder Companies and Walmart, and that revenue tripled over the previous six months. Those are company-reported numbers, so don’t tattoo them on your forehead. But they are still a decent signal: serious companies are already paying to find out what AI systems say about them. ([techcrunch.com](https://techcrunch.com/2026/09/15/aeo-startup-profound-hits-unicorn-valuation-raises-180m-series-d-7-months-after-last-round/?trk=public_post_comment-text&utm_source=openai))
Here is the brutal truth: when someone asks an AI assistant, “What is the best accounting platform for a 20-person business?” or “Which tequila is worth buying for a decent margarita?”, there may be no glorious first page of ten blue links for your brand to compete on.
There may be one answer. Maybe three names. Maybe a follow-up question. And if you are absent at that moment, your gorgeous brand campaign can be about as useful as a billboard in the desert.
This is not SEO with a new haircut
The industry is calling this answer-engine optimisation, or AEO. Fine. Give it whatever acronym makes the budget committee happy. The commercial change is bigger than the label.
Traditional search asked a consumer to choose among options. AI search increasingly does the choosing before the consumer arrives. The machine summarises reviews, product pages, editorial coverage, Reddit threads, retailer data, pricing signals, comparison articles and whatever else it considers relevant. Then it gives the user a neat little verdict.
That changes the job of marketing.
For years, too many businesses treated marketing as a traffic-generation department. Buy attention. Rent clicks. Push people toward a landing page. Measure the conversion rate. Repeat until the finance director starts swearing.
That model is still useful. It is not sufficient.
The new job is to make your company understandable enough, credible enough and consistently represented enough that an AI system can confidently include you in an answer. Not merely find you. Recommend you.
Profound began with analytics around how brands appeared in AI answers. Its current pitch is broader: an AI marketing agent that scans brand data and AI responses, identifies opportunities and uses sub-agents to do work; plus tools designed to keep brand context current as positioning changes. ([tryprofound.com](https://www.tryprofound.com/newsroom/profound-raises-usd180m-series-d-at-usd1-8b-valuation-to-build-the-ai-platform-for-marketing-teams?utm_source=openai))
That expansion matters. Analytics is a feature. Workflow ownership is where the serious money sits.
Why the old playbook is about to get expensive
There are three groups who should be worried.
First: businesses that have built their growth on generic, SEO-farmed content. You know the stuff — “17 Best [Thing] in 2026,” written by someone who has never touched the thing, padded with enough headings to make Google temporarily tolerate it.
AI systems can still ingest that material. But the more generated sludge enters the web, the harder it becomes to build genuine authority from keyword volume alone. You will need proof: useful original content, clear product data, genuine customer evidence, respected third-party references and a brand position that does not read like it was approved by a committee of beige people.
Second: companies with messy facts. If your website says one thing, your sales team says another, your reviews say a third and your distributor catalogue has last year’s information, don’t be surprised when an AI assistant gets confused. The bot is not sabotaging you. Your business gave it a rubbish filing cabinet.
Third: founders who think paid advertising solves every discovery problem. It might. Platforms will almost certainly build more ad inventory into AI answers because nobody has ever met a high-intent commercial query they didn’t want to monetise. Profound itself is developing an Ads Studio for AI-search campaigns, according to its announcement. ([tryprofound.com](https://www.tryprofound.com/newsroom/profound-raises-usd180m-series-d-at-usd1-8b-valuation-to-build-the-ai-platform-for-marketing-teams?utm_source=openai))
But you cannot buy your way out of a weak reputation forever. Paid distribution gets you the meeting. Brand truth determines whether anyone wants a second one.
The overlooked angle: AI search could make real brands more valuable
Most commentary about AI discovery goes straight to panic: fewer clicks, weaker publisher economics, lost control, robots eating the internet. There is plenty of merit in that concern.
But there is a contrarian upside for operators who have done the boring work properly.
A strong brand is not just a logo, a colour palette and a bloke on LinkedIn saying “we’re building in public.” It is a compressed decision. People hear the name and know what you do, who you serve, why you are credible and what price bracket you inhabit.
That compression is precisely what an answer engine needs.
If your company has a sharp category, consistent language, robust reviews, recognisable experts, reliable product information and a trail of independent validation, you are easier for both humans and machines to understand. You become easier to recommend.
That is why I would not respond by hiring an army to spray 5,000 AI-generated articles across the web. That is the marketing equivalent of putting cheap cologne on a dead fish.
Instead, build evidence. Publish comparisons that genuinely help buyers. Make pricing and availability clear. Get your product data cleaned up. Earn coverage from people whose judgement buyers respect. Capture customer stories with specifics, not the usual “great partner” nonsense. Give your subject-matter experts a real point of view.
While building Agave Finder, I see the same issue in the spirits world. A consumer asking for a good additive-free tequila, a bottle under a certain budget or something available nearby does not need 40 vague lifestyle posts. They need accurate, structured information and a source they trust. The brands that make themselves easy to understand will win more of those moments than brands that merely shout louder.
The investment signal is bigger than Profound
Sequoia and Kleiner Perkins did not put money into Profound because every company needs another dashboard. They are betting that marketing teams will be rebuilt around AI-driven discovery and execution.
That does not mean Profound automatically wins. A $1.8 billion valuation is not a business model; it is a very demanding promise. The category will attract incumbents, SEO platforms, agencies and plenty of opportunists selling snake oil labelled “GEO.”
And there is a genuine limit here: no outside tool fully controls the answer a model gives. The models change. Their source preferences change. Their shopping features change. The measurement is still immature. Anyone promising a guaranteed No. 1 recommendation from ChatGPT is either confused or hoping you are.
Still, the direction is obvious. The cheap era of pretending marketing was just performance spend plus SEO content is ending. The winning teams will combine product truth, brand clarity, data discipline and distribution.
That is harder work than keyword stuffing. Good. Most worthwhile things are.
What this means for you
Do not wait for a giant AI-search budget or a boardroom PowerPoint. Do these five things this week.
1. Run the buyer-question test. List the 20 questions a high-intent prospect would ask an AI assistant before buying. Ask several leading AI tools those questions. Record whether you appear, how you are described, who beats you and whether the facts are wrong.
2. Fix the source material. Make your website, product pages, FAQs, retailer listings, sales collateral and public profiles agree on the fundamentals: what you sell, who it is for, price, proof, availability and differentiation.
3. Create evidence, not filler. Commission original research, publish useful comparisons, show real implementation details and collect customer proof with numbers. “Customers love us” is not evidence. “Cut onboarding from 21 days to nine” is.
4. Own a sentence. Every good brand should be describable in one clean line that a customer, journalist, salesperson and AI system can all repeat without mangling it. If yours needs a paragraph and a diagram, you do not have positioning. You have a hostage note.
5. Treat AI visibility as a board-level metric, not a vanity exercise. Track it alongside branded search, conversion, retention and share of voice. But never optimise the metric at the expense of the business. Being recommended is useful only if customers stick around and tell their mates.
Profound’s $180 million round is not a cue to lose your mind. It is a cue to wake up. The next customer may not search for you. They may ask a machine who to trust.
Make sure the answer is not your competitor.