WTAIWhen They Ask AI RU Free audit
How it works

From a customer question
to a clear work plan.

We turn changing AI answers into a documented process you can inspect, repeat and act on.

01

Ask

We run fixed buyer questions for one service and one city.

02

Record

We preserve every answer, business named and source cited.

03

Compare

We map the public evidence that makes competitors easier to trust.

04

Plan

You receive prioritized changes for your site and public profiles.

The question set

10 questions per trade, fixed in advance and identical for every company we measure, so two businesses in the same trade can be compared and one business can be compared against itself month over month. Three ask the browsing question — best, top rated, around the clock. Seven ask the buying question, naming a specific job. Service-specific questions indicate a concrete need, but this sample does not establish search volume or purchasing intent for any individual.

#Question as typed (plumbing set)Intent
1“Best plumbing company in [city]”Browsing
2“Who is the top rated plumber in [city]”Browsing
3“24 hour plumber [city] recommendations”Browsing
4“Who should I call for emergency plumbing repair in [city]”Buying
5“Who is the most trusted plumbing company in [city]”Buying
6“Affordable drain cleaning [city], which company should I hire”Buying
7“I need a water heater replaced in [city], who should I call”Buying
8“Who can fix a leaking pipe under my sink in [city]”Buying
9“Best company for sewer line repair in [city]”Buying
10“Who does whole house repiping in [city]”Buying

The free check runs the first 3. The paid audit runs all 10. Trades other than plumbing get their own set, written the same way and published before the first run.

Which engine, exactly

Every measurement we run uses Google AI Mode, and only Google AI Mode. The free check, the full audit and any follow-up implementation measurements all use it. We keep the engine consistent so the scope is clear.

ChatGPT and Perplexity are not included in these packages. Other platforms need their own measurement scope and baseline; results from different engines should not be combined into one score.

Every run uses a United States connection with your city named in the question. We record the date, engine, location context and session conditions, and retain the answer text and available source links. An unavailable answer is marked unavailable, not scored as a missing business. Coverage uses completed answers, with any shortfall disclosed. We record three things per answer: whether your company was named, which companies were named instead, and what evidence the assistant quoted for each recommendation — a review count, a rating, a warranty, a price, a license number, a forum thread. That third column is what turns a score into a work plan.

What the answers actually cite

In our September 12, 2026 Las Vegas run, recommendations were justified with published, checkable statements far more often than anything else. The assistant quoted exact review counts and ratings, BBB grades written out in words, founding years, Nevada contractor license numbers, named neighborhoods, concrete offers such as a $59 drain special or $250 off a tankless install, and process details like in-house crews or permits pulled. The sources behind those quotes were the companies' own service pages, Yelp and Google review text, local Reddit threads, and 2026 roundup articles.

On structured data, stated honestly

We publish Organization, ProfessionalService and Service markup for our clients because it is how a business states its identity, services and service area in machine-readable form, and because search engines document it as a supported way to understand a page. We do not claim that any markup type causes an AI assistant to recommend you. No one outside these companies can demonstrate that, and a vendor who tells you a schema block buys you a place in the answer is guessing.

We also do not put your Google star rating into your own review markup. Google's guidance treats ratings a business publishes about itself as self-serving, and they are not eligible for review snippets. Your rating belongs on the page as plain text with its source and the date it was read.

What the optional Growth service changes

1

The facts you publish about yourself

Founding year, license number, BBB grade, warranty terms, pricing policy, service areas by name, and your real review figures written as plain text with the source and date — not as a logo or a banner image.

2

A page per job you actually sell

Water heaters, sewer lines, drains, repipes, emergency leaks. Each with process, options, price range and the questions customers ask. Buying questions are lost on sites that answer them only with a list of service names.

3

Your profiles and listings

Google Business Profile completeness, consistent name, address and phone across directories, and a review cadence that keeps recent reviews coming in and answered.

4

The third-party sources being cited

Legitimate inclusion in local roundups and trade directories, and an honest presence where your customers already discuss trades. No fake reviews, no sockpuppet forum posts, no paid placements disguised as recommendations — those are the fastest route to being dropped by both search engines and communities.

Limits you should hear before you pay

Answers are not deterministic. Re-running the same question can return a different set of companies. In our own Las Vegas runs on two consecutive days, two companies appeared one day and not the next. A single run is a sample, never a ranking.

Attribution is limited. When your number improves we can show what we changed and when, but we cannot prove which change caused it. We report the correlation and the work, not a causal claim.

Placement cannot be bought or promised. We sell measurement and the publishing work behind it. Anyone offering a guaranteed spot in an AI answer is selling a promise they do not control.