The marketing industry is changing
Get recommended by AI, not just ranked
Buyers now ask an AI assistant what to buy — and it answers with a short list of names. Agentic endorsement is being the business the AI names. If it can't read you, trust you, and quote you, it recommends someone else.
Find out how you show up today, or build the foundations so AI recommends you.
Why being answerable to AI is the new front door
The answer replaces the list
A buyer used to scan ten links and pick. Now they ask an assistant and get one answer naming a couple of businesses. If you're not named, you're not in the room.
The reader is a machine first
Before a human ever sees you, an AI reads your site to decide whether it can trust and quote you. If it can't parse you, it recommends whoever it can.
Trust is structured, not claimed
Models lean on clear, machine-readable evidence — structured data, answerable copy, citable facts. Adjectives don't move them; parseable proof does.
Two ways to start
See where you stand today, or hand us the business and let the pipeline position you to be endorsed.
Check where you stand
Run the AI-visibility analyzer
Scan your site to see whether AI crawlers can read you, whether your structured data earns trust, how answerable your copy is, and who the AI cites instead of you today — with the fixes to close the gaps. One-time $250.
Analyze my siteAlready running a business
Position my existing business
Tell us your revenue, the pain points you already know about, and your current site. The same evidence-gated pipeline positions you so AI assistants can find, trust, and recommend you — human-reviewed before it lands in your dashboard.
Start the questionnaireThe engine we run on
We practice what we position. This is the actual technology behind itstru.ai — and the answer engines we set you up for.
Our stack
Claude + Mastra.ai
The reasoning in every run is done by Claude, Anthropic's frontier model family. The multi-step pipeline around it — research, evidence gating, positioning, prototyping — is orchestrated with Mastra.ai, an open-source TypeScript agent framework, with evals, observability, and durable execution built in.
The engines we optimize for
ChatGPT (GPT-6 Astra), Gemini, Perplexity, Claude
These are the assistants buyers actually ask — so they are the answer engines we measure against and set you up for AEO success in. Claude wears both hats: it powers our engine, and the Claude assistant is also one of the answer engines we measure.
We are not sponsored by or partnered with OpenAI, Anthropic, Google, Perplexity, or Mastra. We use their tools and optimize for their answer engines.
Frequently Asked Questions
What is agentic endorsement?
Agentic endorsement is when an AI assistant recommends your business by name in its answer. Buyers increasingly ask tools like ChatGPT (GPT-6 Astra), Claude, Gemini, Perplexity, and Google's AI overviews what to buy or who to hire, and the model replies with a shortlist. Agentic endorsement is earning a place on that shortlist — being the name the AI trusts enough to say out loud.
How is this different from SEO?
SEO optimizes pages to rank in a list of blue links a person scrolls. Agentic endorsement optimizes for a machine that reads your site, decides whether it can trust and quote you, and then answers on your behalf. Ranking gets you a click; endorsement gets you named in the answer before the click ever happens.
Why does the marketing industry say this is changing?
Because the first surface a buyer sees is shifting from a page of links to a single AI-written answer, and the assistants writing those answers now run on frontier models — OpenAI's GPT-6 Astra (launched September 2026), Claude, Gemini. When the assistant names two or three businesses and stops, the ones it didn't name are invisible — no ranking, no click, no chance. Being answerable to AI is becoming the entry ticket, not a nice-to-have.
How do I find out whether AI can already recommend me?
Run the analyzer on your site — a one-time $250 scan. It checks whether AI crawlers can read you, whether your structured data lets a model trust and quote you, how answerable your copy is, and who the AI cites instead of you today — then it gives you the fixes.
I already run a business. What's the fastest path to being endorsed?
Start the existing-business questionnaire. It captures your revenue, known pain points, and current site, then runs the same evidence-gated pipeline to position you so AI assistants can find, trust, and recommend you — human-reviewed before it reaches your dashboard.
What technology does itstru.ai run on?
The reasoning in every run is done by Claude, Anthropic's frontier model family, and the multi-step pipeline around it is orchestrated with Mastra.ai, an open-source TypeScript agent framework — with evals, observability, and durable execution built in. We optimize for the answer engines buyers actually use: ChatGPT (GPT-6 Astra), Gemini, Perplexity, and Claude. We are not sponsored by or partnered with OpenAI, Anthropic, Google, Perplexity, or Mastra. We use their tools and optimize for their answer engines.
Not sure which path fits? Talk to us.