Skip to content
Audience Intel
Solutions · by problem

How do buyers actually see us, compared with the alternatives?

Customers praise you once they have worked with you. The buyers who looked and left say nothing. The free report reads your site and your rivals' with five buyer types on one fixed rubric, and cites the points on the page behind each verdict.

Answered today
The sample report's buyer narrative: a first-person walk through the firm's site as one buyer archetype would read it
Sample report · Hadley & Whitmore is a constructed test firm
The problem

What owners say.

  • I don't know how we come across.
  • Everyone says we're great once they've worked with us. Getting them in is the problem.
  • Do we look like a big firm or a small one?
  • What does someone see when they search for what we do?

Composite phrasings of a common problem, not quotes from named customers.

Why it is hard

Why checking by hand fails.

Feedback from customers comes from the ones who chose you. The buyers whose view you need left without a word.

Buyers are not one person. A careful researcher reads a site differently from someone in a hurry who wants a price and a phone number. A site that suits one can lose the other.

Here is the output, from the sample report on a constructed test firm. Crowthorne Legal scores 81 with The Anxious First-Timer. Hadley & Whitmore scores 58. The recurring drag is a missing "How we work" page, and priority action one is to publish it in five steps.

Buyers also meet you on the search results page. The report shows where you sit for the terms that matter. Ask the AIs puts a question about your trade and town to several assistants and shows each answer word for word, unscored.

The limits. The buyer types are archetypes, not surveyed customers. Ask the AIs shows what assistants answer. It makes no claim about how buyers use those answers.

What we read

Each source, and the code that reads it.

Every source names the module in our code that reads it. A source marked in build does not run for customers yet.

  • Five buyer archetypesYour site and your rivals' sites scored by five buyer types against a fixed, published rubric.src/lib/prompt/evaluate.ts
  • Your pages and theirsThe pages a buyer reads before choosing, read as text.src/lib/firecrawl.ts
  • Search results for your tradeWho appears for the terms buyers search, and where you sit.src/lib/connectors/dataforseo.ts
  • Assistant answersThe answers several assistants give to a question about your trade and town, shown word for word and unscored.src/components/shell/tools-meta.ts
What you get

What you get back.

  • A score per buyer type for your site and each rival, on the same rubric
  • Where each buyer type stalls, with the evidence from the page
  • Where you sit in search for the terms that matter
  • Verbatim assistant answers to a question about your trade and town
Today and in build

What runs today, and what is in build.

Live tools run now. Loops in build will answer it again on a cadence you choose. They are not on sale.

FAQ

Questions about “How do buyers see us”

Are the buyer types real people?

No. They are archetypes from a published library, each with the decision factors that kind of buyer weighs. They are a fixed lens, applied the same way to every site. They are not a survey of your customers.

What is the rubric?

A fixed set of questions a buyer needs answered before choosing. The same rubric is applied to your site and every rival's, so the scores compare like with like. The sample report shows it in full.

What does Ask the AIs show?

The answers several assistants give to your question, word for word. It is a reading of what they say, not a score, and we make no claim about how buyers use those answers.

Can I commission research with real buyers?

Yes. Research studies are on the pricing page as an add-on, scoped on a call before anything is charged.

Name the rivals you lose to.

The free report reads your site beside theirs, five buyer types on one rubric. No account and no card to start.