When a buyer asks AI who to hire, are you in the answer?

Why this matters now

A growing share of buying decisions never reaches a results page.

68% of US Google searches ended without a click (SparkToro, June 2026). ChatGPT serves roughly 900 million weekly active users. Google AI Overviews appear on about 30% of searches, and on 74% of problem-solving queries (Authoritas, 2025). Gartner expects a quarter of organic search traffic to shift toward AI assistants.

None of that means Google is finished — it still holds roughly 90% of global search. It means the answer a buyer sees is increasingly assembled for them, and a business that is not in that assembly is invisible at exactly the moment of choice. Being ranked and being recommended have come apart, and most businesses only measure the first.

Step one — measure

We establish what AI says today, with receipts.

Measurement runs on Routeless Radar, our own software. It asks real buying questions for your trade and service area, then records what came back. You get:

You can see the shape of this before talking to anyone. The free scan takes about three minutes, needs no signup, and returns a real report rather than a teaser.

Step two — fix

Measurement is the cheap part. The work is what changes the answer.

A scan tells you that you are absent. It does not make you present. Most of what keeps a business out of AI answers is unglamorous and fixable: the assistant cannot parse the site, the business facts are inconsistent across the places it looks, or there is nothing on the site that answers the question being asked.

You can check parts of this yourself right now. The AI readability checker runs 11 technical checks in seconds, the Google Business Profile audit scores your public profile, and the question generator gives you the ten questions worth testing in any assistant.

Step three — measure again

The same questions, asked after the work ships.

This is the part that separates the service from an opinion. Radar keeps asking the same buying questions on a weekly cadence, so the record shows what your visibility looked like before the work, and what it looks like after. Competitor movement is tracked alongside it, because appearing more often matters less if everyone else did too.

We would rather show you a flat line than explain one away. That is the point of dating and sourcing every observation.

What we will not tell you

Honest limits, stated up front.

AI answers vary between runs. Ask the same question twice and you can get different businesses. That is why Radar measures repeatedly and reports a rate with a confidence range rather than a single snapshot, and why anyone quoting you a one-off result is measuring noise.

Nobody can guarantee a #1 recommendation in ChatGPT. There is no ranking dial, no submission endpoint, and no paid placement. What can be done is to remove the reasons an assistant cannot or will not name you, and then measure honestly whether it started to.

Radar is our own product, so weigh that accordingly. We built it because we needed the measurement for client work and nothing available did it with receipts. You can use the free scan without ever speaking to us.

What it costs

Part of a build, or an ongoing program.

AI visibility is not sold on its own, because on its own it does not work — the fixes land on your website, your profile, and your content. Every project tier includes a Website & Conversion Foundation Radar baseline scan; the Visibility & Intake System engagement ($12,500–$17,500, averaging about $15,000) adds 12 months of weekly tracking plus the search and AI corrections themselves.

After launch, Visibility Management keeps it running at $1,250–$1,750/month — profile management, weekly Radar tracking, ongoing technical improvements, and a monthly piece of content. Smaller one-off work, like a single technical fix or citation cleanup, is scoped per piece from $500.

Full inclusions are in the packages section, and the other systems this connects to are on the services overview.

Questions we get

Before you ask.

Is this different from SEO?

It overlaps and it is not identical. Most of what makes a site legible to an AI assistant — crawlable content, structured data, consistent business facts, pages that answer real questions — is good SEO practice. What differs is the measurement: SEO asks where you rank, this asks whether you get named, and those two can disagree. Google's own guidance is that standard SEO fundamentals apply to its AI features too, so this is not a separate discipline bolted on the side.

Do I need an llms.txt file?

It is worth having, and it takes seconds. An llms.txt file hands AI tools a clean, machine-readable summary of your business — who you are, what you do, and which pages matter — instead of making them work it out from your markup. Customers are genuinely finding businesses through ChatGPT, Perplexity, and other assistants now, so anything that makes your site easier for those tools to read is effort in the right direction. Our free generator will write one for you.

Keep the expectation right, though. Google has confirmed that AI text files have no effect on Google Search, so this is a tidy addition rather than the thing that makes you visible. The heavy lifting is still crawler access, server-rendered content, structured data, and business facts that agree with each other everywhere an assistant looks.

How long before the answers change?

Honestly: it depends on what was wrong. Correcting a bad business fact can show up quickly. Building the content and review base that makes an assistant confident enough to name you is a matter of months, not weeks. The weekly measurement is there so you are watching it happen instead of waiting for a quarterly report.

Can I just run the free scan and fix it myself?

Yes, and some people should. The free report includes up to three prioritized actions with DIY steps, and the free tools cover the technical checks. We are the right call when the fixes reach into the website, intake, and content at the same time, and you would rather that be one accountable scope than four vendors.

Start here

See where you stand before you decide anything.

Run the free Radar scan — about three minutes, no signup, real answers with sources. If the report shows a problem you would rather hand off, check your fit and you will work directly with the engineer who builds it.