AI changes its answer every time. Here’s how we make sense of it.

1. See where you’re missing out

When you sign up, Ansoar reads your website and works out the jobs you want more of, like a home battery install or a roof restoration after a storm (more examples by industry). For each one it writes the questions a customer would actually type, and sets them to the suburb, city or region your customers are in. You can change any of it, or leave it be.

Then it asks those questions once a day on the three AI platforms you pick, and records whether you were recommended, only mentioned in passing, or left out. Those are different problems with different fixes, so they’re counted separately.

The percentage is simply the share of answers that recommended you. It doesn’t tell you how much work each business wins, so treat it as a direction, not a market share figure.

The five platforms you can choose from

They often name different businesses for the same question, so each gets its own results. You can swap platforms whenever you like and keep your history, but sticking with the same three for about a month gives the numbers something solid to compare against.

2. Work out why

Click into any job and you can read every answer behind the number. You’ll see who gets picked instead of you, including businesses you never thought of as competitors, and whether you’re losing everywhere or only on one platform.

Ansoar also lists the websites AI quotes when it picks a competitor over you: review sites, directories, a rival’s rebate guide. They’re worth a look, though being listed on one is no guarantee. Where you do appear, it pulls out what AI tends to praise you for and what it’s unsure about.

3. Fix the right thing first

No 80-page audit. After a few days to a week of answers, Ansoar suggests a handful of jobs in order, each with a rough effort, how many of your searches it relates to, and why it made the list. Where something needs writing, you get a first draft with your prices and guarantees highlighted for you to fill in, because Ansoar doesn’t know them.

Two checks look at your own website. Crawlability makes sure the crawlers behind AI search and Google can reach and read your pages, with fixes your developer can act on. Page Readiness reads a page the way a buyer would for a particular job, and points out what they can’t find.

4. Check if it worked

When you make a change, log it. Ansoar keeps asking the same questions on the same platforms, then puts the weeks before and after side by side, along with what your main competitor did over the same period. Agencies use this to show clients what changed after their work.

If your numbers rose but so did everyone else’s, that’s worth knowing. And because AI answers shift for reasons nobody outside the AI companies can see, Ansoar reports what changed after your work, not that your work caused it.

What you connect

Your website

So Crawlability and Page Readiness can check pages on your own domain.

Google Analytics, optional

To see visits that clearly came from AI platforms, and the leads and sales that followed. You configure how AI visits, leads and sales are tracked, using the events you already measure. Plenty of people who find you through AI won’t show up as AI traffic, so think of it as the minimum. Ansoar only reads reporting data.

Your email

For the Daily Brief, on the days you choose. If nothing changed, it doesn’t send.

What we won’t pretend to know

We can’t tell you exactly why an AI picked someone, and we won’t promise you a recommendation. Nobody outside those companies can.

We also won’t sell you fixes with nothing behind them. A blocked training crawler doesn’t make you invisible in AI search, and there’s no published evidence an llms.txt file helps: Google says Search ignores it, and Ahrefs found 97% of published files got no requests in a month.