Gradial’s $65M Bet: 29% of Companies Still Have Nobody Owning AI
Nearly one in three companies has nobody responsible for GEO. That is how a brand can spend millions on marketing yet leave its AI visibility to chance.
Nearly one in three companies has nobody responsible for generative-engine optimisation. You can spend millions on marketing and still leave how your brand appears in AI-generated answers to chance.
That is the marketing story worth paying attention to: not another celebrity campaign, not a shiny rebrand, and not some agency winning an award for making a logo wobble. Seattle startup Gradial has raised $65 million to build AI agents that operate across enterprise marketing systems. At almost the same time, a survey of 1,115 communications professionals found that 73% think generative-engine optimisation matters to their strategy, while 29% say nobody at their organisation owns it.
There it is. Marketing’s newest job is important enough for everyone to discuss and inconvenient enough for nobody to own.
The $65 million is not the story. The operating mess is.
Gradial’s Series C matters because it is not selling the usual AI fantasy: type a prompt, get a mediocre blog post, call it innovation. The company is pitching something much more valuable and much harder to get right: AI agents that do actual marketing operations across the ridiculous stack large companies have accumulated.
Think Adobe, Salesforce, ServiceNow and Databricks. Different systems. Different teams. Different permissions. Different dashboards. Usually, different executives protecting different budgets like they are defending a small island nation.
Gradial says its agents can identify where a brand is missing from AI-generated answers, draft changes, send them through existing approval workflows and publish across systems. That is not glamorous. It is also exactly where the money is.
Most marketing departments do not have a creativity problem. They have a throughput problem. Good work gets stuck in approvals. Product information is outdated in one system and correct in another. The customer-support team hears objections that never reach the content team. Search data sits with performance marketing. Brand owns the words but not the distribution. Sales knows what prospects actually ask, but its insight gets lost in a CRM graveyard.
Then management buys an AI tool and acts surprised when the mess becomes automated at scale.
I have built businesses. I can tell you this much: automating a bad process does not make you advanced. It makes you faster at looking incompetent.
AI search has created a new front door to every business
For twenty years, the internet had a fairly clear bargain. You wanted attention, you fought for it through search, social, email, PR, paid media, retail shelves or direct relationships. Google made the rules of search visible enough that an entire industry grew around gaming, improving and measuring them.
Now customers are increasingly asking AI assistants to narrow the field for them.
That changes the game because an AI answer is not a page of blue links. It is a compressed judgement. It may tell a buyer which accounting platform suits a 30-person company, which running shoe works for flat feet, which insurance policy has the right cover, or which tequila is worth bringing to dinner. If you are not in the answer—or worse, the answer gets you wrong—you do not merely lose a click. You may never make the shortlist.
Axios reported that generative-engine optimisation, or GEO, is becoming a strategic concern as answer engines grow and traditional search adds more AI features. Yet ownership is a mess: only 24% of surveyed professionals said PR or communications owns GEO, while just 11% said marketing leadership does.
That split tells you everything. Everyone can see the new shopfront. Nobody has accepted responsibility for the rent.
The old argument—“Is this SEO, PR, content, brand, digital or product marketing?”—is managerial wallpaper. Customers do not care which internal tribe owns their question. They care whether the answer is accurate, useful and trustworthy.
If an AI assistant describes your product badly, cites an ancient pricing page, repeats a competitor’s framing or skips you altogether, the damage lands on revenue. It does not politely wait for the org chart to sort itself out.
The overlooked angle: this is a data discipline before it is a content play
Here is where I disagree with a lot of the commentary around GEO. The temptation is to treat it as SEO with a new haircut: write more pages, sprinkle in more keywords, make your site look friendly to robots, and hope for the best.
That is too shallow.
The businesses that win will not be the ones publishing the most AI-flavoured sludge. They will be the ones with the cleanest, most consistent and most defensible underlying information.
Can a customer find your current product range, exact pricing logic, returns policy, availability, proof points and category credentials without tripping over contradictions? Are your official pages better than affiliate lists, stale review pages and competitors’ spin? Does your customer-service team use the same language your website uses? Do you have credible third-party evidence that supports what you say?
If the answer is no, you do not have a GEO issue. You have a business-information issue.
That is why Gradial’s pitch is more interesting than another content-generation tool. Its value, if it delivers, is in connecting discovery to action inside the systems companies already use. Spot a gap in how the market finds you. Create the right update. Route it to the people who can approve it. Publish it in the places that matter. Measure whether the gap closed.
Boring? A bit. Profitable? Potentially very.
The hard truth is that brand has always been operational. Your brand is not your typeface, your colour palette or the expensive video your agency wants to enter at Cannes. Your brand is the pattern of expectations people form after interacting with your company. If your claims, service, product data and public reputation all point in different directions, no volume of clever advertising fixes it.
Creators are becoming distribution, not decoration
There is another piece to this shift that founders should not miss. As AI makes generic content cheaper and search traffic less dependable, trusted human distribution becomes more valuable.
Axios has described creator commerce as moving from experimental tactic to core retail strategy. That should not surprise anyone. In a world flooded with content that looks polished but may have been produced by a machine in 45 seconds, a credible person with a real audience is not just “influencer marketing.” They are a media channel, a trust layer and often a sales channel rolled into one.
But do not make the rookie mistake of treating creators as rented billboards.
The useful question is not, “How many followers do they have?” It is, “Do they have earned attention from people we genuinely want as customers—and can they explain why our product deserves it?”
That is a higher bar. It requires a product worth discussing, a clear offer, decent briefing and enough confidence to let the creator sound like themselves. The brands that try to squeeze every partner into approved corporate scripts get the worst of both worlds: they pay for reach and receive an ad nobody believes.
This is also why authenticity has become such a marketing obsession. Marketing Dive’s review of major campaigns this year pointed to brands leaning into honesty, self-awareness and real consumer voices as fake feeds and AI-generated material proliferate. The lesson is not that every brand needs to become painfully earnest. Please, spare us. The lesson is that manufactured polish has lost some of its power.
People can smell when a brand is performing humanity.
Do not hand the whole thing to the marketing intern
The contrarian point is simple: this should not be owned solely by marketing.
Marketing should lead the customer narrative. Absolutely. But GEO, AI visibility and brand accuracy are cross-functional problems. Product controls what is true. Legal controls what can be said. Customer service knows where the promises break. Sales knows which questions block a purchase. Data teams hold the systems. PR understands reputation. Marketing is the natural quarterback—but the quarterback still needs a team on the field.
If you are a founder, appoint one accountable owner anyway. Not a committee. One person with authority, a budget and a weekly dashboard.
Their job is not to “win AI search,” whatever that means this month. Their job is to make sure your company is accurately represented wherever customers make decisions: AI assistants, search results, review sites, creator content, marketplaces, retail shelves and your own channels.
That is brand management now. Less theatre. More truth maintenance.
What this means for you
If you run a business, do these five things this week.
1. Ask five AI tools what they say about your company and category. Use the questions a buyer would actually ask, not vanity prompts. Screenshot the answers. Look for omissions, errors, stale facts and competitor framing.
2. Name one owner. Give them a 90-day mandate to map how your business appears across AI answers, search, reviews, marketplaces and key creator channels. If nobody owns it, it will become a Slack thread and die there.
3. Build a source-of-truth file. List your current products, prices, policies, differentiators, proof points, customer FAQs and approved claims. Then make sure every customer-facing team is using it. You cannot manage your reputation if your own business cannot agree on the facts.
4. Fix the ugly basics before creating more content. Update stale product pages. Remove contradictory claims. Improve help content. Get real customer reviews. Make your expertise easy to verify. This is less exciting than an AI content factory—and far more likely to make you money.
5. Treat creators as commercial partners, not borrowed cool. Find people whose audience already cares about the problem you solve. Give them a product, a point of view and room to be honest. Track sales, not applause.
Gradial’s $65 million raise is a bet that marketing’s fragmented machinery can be made to move faster. Fair enough. But the bigger bet is on whether companies are willing to stop treating brand as the art department and start treating it as a revenue system.
The companies that do will be easier to find, easier to trust and easier to buy from. The rest will spend more on ads explaining why they should have fixed the basics years ago.