SEO Is Dead – Welcome to AI Engine Optimization

01 /The Concept

Unusual has built an AI publicist – one that, perhaps surprisingly, pitches its clients' products not to human journalists, but to AI chatbots like ChatGPT.

The objective, though, is classically PR: change what AI systems think and say about a client's products, so those products show up in chatbot recommendations and generate more qualified leads.

The workflow breaks into two phases.

In the first phase, the AI publicist audits what chatbots currently say about the client's products – whether they mention them at all, and if so, what strengths and weaknesses they highlight.

To do this, it asks chatbots a range of questions relevant to the client's target market and product category – for example: "Name the best CRMs for small businesses and explain your choices."

It then analyzes the responses to identify blind spots that can become opportunities. If ChatGPT cites competitor case studies published on Reddit, the platform will recommend publishing similar case studies and specify exactly which Reddit communities to target.

More broadly, the AI publicist pays close attention to the sources chatbots draw on. It catalogs all sources that chatbots treat as authoritative for a given topic, along with the specific quotes they tend to use.

It also runs a full competitive analysis – mapping what chatbots say about the client's competitors – and produces a comparison table showing positive mention share across products, giving the client a clear view of its relative standing as perceived by AI.

Once the diagnostic is complete, the platform shifts to phase two: creating content designed to make AI systems take notice.

The first content type targets the external sources chatbots already treat as authoritative in the relevant category. Content here is crafted to match the formats and specific language patterns that chatbots use when citing competitors.

The second type goes into owned media – blogs, publications the client controls – where the brand can make stronger, more direct claims than it could in third-party channels.

The third type is a dedicated set of pages hosted on a separate subdomain of the client's site, deliberately kept out of the main navigation. Why? Because this content is written exclusively for AI consumption, not human readers.

The distinction matters. Conventional marketing copy is built around emotion – because humans buy on feeling. This AI-targeted content is built around logic: data points, analysis, structured conclusions. A human reader would find it dry. An AI chatbot treats it as a reliable basis for inference and citation.

The whole cycle – analysis, content creation, monitoring – runs continuously, and the platform shows clients simple charts tracking how their positive mention share across different chatbots changes over time.

Unusual went through Y Combinator last fall with an entirely different product – a platform for dynamically personalizing website content based on how a visitor arrived at the site, [covered here](/review/kak-uvelichit-konversiju-na-300). That product was apparently still active just a couple of months ago. But Unusual has now publicly launched this new AI PR platform via YC's website.

02 /Why It Matters

The B2B sales funnel has changed substantially.

A few years ago, it typically started with a Google search. Traffic converted into leads via forms and landing pages, and then a CRM-powered sequence of calls and emails took over.

Today, the funnel increasingly starts differently. Buyers take products they've seen mentioned on social media and ask ChatGPT which tools it recommends – assembling a preliminary shortlist from the response.

From there, they visit the websites of shortlisted products and explore demos if they exist and are easy to access.

Then they build a final shortlist and reach out directly to vendors to negotiate details, speak with existing customers, and make a decision. More than 80% of B2B buyers now follow this pattern – the same buyers who used to be targeted with unsolicited cold calls and email blasts.

Those blasts haven't stopped, but they're generating diminishing returns. In 90% of cases, corporate buyers end up purchasing a product that was already on their shortlist before any sales outreach occurred.

Which means AI chatbot recommendations are now a critical early-stage filter in the purchase process. Getting recommended by ChatGPT isn't a vanity metric – it directly affects pipeline.

Unusual is building squarely in this gap. And it's not alone – several startups have emerged around the same problem.

Otterly.AI ([related review](/review/optimizirovatsja-nuzhno-uzhe-ne-pod-gugl)) launched a platform for AI search optimization and gained visibility on Product Hunt late last year.

Evertune ([related review](/review/poiskovaja-optimizacija-uzhe-proshlyj-vek)) raised $4 million for a similar platform last October, then followed with another $15 million after the original review.

03 /Opportunities

The broad direction: build platforms that support the new sales funnel.

Unusual and similar tools address one layer – AI chatbot optimization. But interesting platforms are emerging for other stages of the funnel too.

Rocksalt ([related review](/review/kak-naladit-pravilnyj-smm-vedushhij-k-prodazham)) raised $3.5 million in July on a platform that helps companies get their products mentioned in social media before potential buyers even think to ask ChatGPT. The platform recommends which channels to show up in and what kinds of posts and comments should feature company experts and their products.

Brand Stori ([related review](/review/pravila-jeffektivnyh-prodazh-izmenilis)), which launched in June, built an AI engine that analyzes B2B product websites and recommends how to make them more commercially effective.

Supersonik ([related review](/review/kuj-prodazhi-poka-gorjacho)) raised €4.2 million in early September for a platform that creates personalized product demos on the fly – tailored to each individual prospect, generated instantly, with the AI even able to build out specific demo segments in real time to answer live questions about how the product works.

There's still a lot of open space across this category. Pick the stage of the funnel that fits your strengths, study what's already been built, and take the best of what exists as a starting point to build something better