Find out what AI actually thinks about your company.

Whether you're just starting with AI search or already investing and can't tell if it's working, the first step is the same: research into how AI models see your category, your company, and your competitors.

Book a conversation

Your buyers are asking AI who to consider.

More software buyers now start their research by asking ChatGPT, Claude, or Gemini. The assistant doesn't hand back ten links. It reads what it can find, weighs the options against what the buyer said they need, and suggests a shortlist. That judgment happens before anyone visits your website.

51%

of B2B software buyers say they start research with an AI chatbot more often than with Google.

54%

say AI chatbots influence which vendors make their shortlist.

Source: G2, The Answer Economy, a survey of 1,076 B2B decision makers, March 2026.

Counting mentions is a start. Understanding them is the work.

AI visibility tools track how often your company appears in answers to a set of prompts. That's useful, the same way web analytics is useful. But analytics never told anyone why a visitor left, and mention counts won't tell you why a model recommends your competitor to your best-fit buyer.

Visibility tools answer

How often are we mentioned?

Which sources get cited?

Is our share going up?

Construal Research answers

What does the model believe about us, and where did it learn it?

Which buyers does it send elsewhere, and why?

What can we change, and what should we work with instead?

Most teams will want both. We're the research.

How an engagement works

  1. 1

    Audit

    We map how AI models understand your category: the criteria they use to compare vendors, what they know and believe about you, and where that picture is wrong or incomplete.

  2. 2

    Strategy

    We build controlled experiments based on your goals, so our recommendations are specific to your needs instead of best-practice checklists.

  3. 3

    Report

    You get the findings, a brief for execution, and a measurement plan with a baseline.

    your first deliverable
  4. 4

    Execution

    Your team runs it, or we bring in a partner and coordinate the work.

  5. 5

    Measurement

    We re-run the research against your baseline each month and report what changed, what's newly possible, and what to do next.

Sometimes the best move is to stop arguing with the model.

Changing what AI believes takes time, and some beliefs won't move. But knowing what a model already thinks about you is valuable on its own. If it keeps recommending you to a certain kind of buyer, you can sharpen your landing pages, sales story, and positioning around that buyer today. Part of the work is telling you which beliefs are worth changing and which are worth using.

From Insights

All Insights

Michael Koach

I'm Michael Koach. I got into marketing because I'm fascinated by people. Since 2023, I've been asking the same questions about AI: what it believes, why it recommends what it does, and how that shapes what people buy.

If you're working out what AI means for your company, I'd love to hear about it. Let's start with a 30-minute conversation about your category.

Michael Koach, Founder

Book a conversation Get Insights by email