← Back to Blog

The first 14 days: what to do each day of an AI visibility trial

A day-by-day plan for the 14-day trial that follows the emails you will get: setup, the first result, the page to change on day 2, questions on day 3, the re-check, the decision at three days left, and the extension.

The trial is 14 days at Pro capacity, no card. Most of the work is in the first three days and the last three. This plan follows the emails the product sends, so each day's task matches what arrives in your inbox. The trial covers 3 websites, 75 tracked questions, 25 GEO topics, 10 competitors, weekly checks and 15 on-demand re-checks. Google AI Mode is the one Pro feature not included.

The plan at a glance

DayWhat arrivesWhat to doTime
0First result, about five minutes after setupAdd the site, edit the suggested questions, read the first result20 minutes
1The weekday of your first weekly check, and one question a rival already winsAdd the named rivals as competitors10 minutes
2The one page to change this weekChange it, or request the listing30 minutes to a day
3Connect Search ConsoleConnect it; replace advice-only questions10 minutes
4Your last check is ageing: re-checkRe-check the question you changed5 minutes
5 to 6NothingKeep re-checking once a day; start the second fixAs needed
7First weekly results, and one answer where a rival was namedRead what moved; do the one thing before next week15 minutes
8 to 10NothingSecond fix; share the dashboard; set an alertAn hour
11Three days left: where you standDecide on the plan15 minutes
13Your trial ends tomorrowPick a plan, or plan the extension5 minutes
14Trial ended; account read-onlyExtend once by seven days if you need more data1 minute

Day 0: setup and the first result

Add your website. We read it and suggest the questions buyers ask AI about businesses like yours. Edit them; the questions are the whole product. Keep the ones a buyer would type before knowing your name and drop the ones that contain it, because branded questions inflate every number. The first check runs in the background and the first result arrives in about five minutes: named in X of N answers, who was named instead, the sources, one quoted answer and one fix. How to read it is its own post. If you ran the free audit first, the question you used is already in the account.

Day 1: the weekday and the rival

The day 1 email tells you which weekday the weekly check runs, and shows one question a competitor already wins, with the answer. Add every brand named in your answers to the competitor list, up to ten. Share of voice and the weekly "who is above you" line are built from that list, and the decision on day 11 is easier with them in it.

Day 2: the one page to change

The day 2 email names one page: either a page you own that the engines read without taking a recommendation from it, or a page you are missing from. It is chosen from the sources behind your answers, ranked by effect. Do this one before anything else. Across one account's 34 questions the engines cited 380 domains, and four of the ten most-cited pages were roundup lists (what those citations showed); the page in your email is usually one of those. The Opportunity report for the question, under Improve, is generated once per question on request and gives the detail.

Day 3: the questions

The day 3 email asks you to connect Google Search Console. It is read-only and takes one click, and the tracked questions are rebuilt from searches that actually reach your site. Also replace any question marked advice-only: those get a how-to answer that names nobody, are left out of the score, and carry a Replace action. Fifty questions buyers ask by name beat seventy-five generic ones.

Days 4 to 6: make the change, then re-check

Publish the fix, or send the listing request. Then re-check that one question, once a day. Day 4 brings a short email saying the last check is ageing, which is the prompt to do exactly that. A third-party page has to be updated, crawled and retrieved before an answer changes, so most re-checks in the first week show no change, and that is information too. Watch for a citation first; engines read a page as a source before they repeat what it says.

Day 7: the first weekly results

The weekly email carries the score with its ± range, the change against last week, questions naming you, what was gained and lost, who was named instead or who is above you, and one thing to do before the next check. Read a move inside the range as normal variation: the run-to-run flip rate on the platform is 3.4% (September 2026), so one row changing on its own is not yet a trend. Around the same day a separate email points at one answer where a competitor was named and not you, with the quote.

Days 8 to 10: the second fix, and the team

Take the next item on the improvement plan. Share the dashboard with whoever will decide on the plan; the trial runs at Pro capacity, which includes team seats. Set the Slack or Teams alert if you want one. It fires only when a score moves by at least your threshold and the two ranges no longer overlap, so it stays quiet in a normal week.

Day 11: three days left

The three-days-left email is the decision email: where you stand in AI answers with its sample, how many questions name you, the most-named rival, one quoted answer, the price and one button. If nothing has moved since a date, it says so rather than implying news. Plans are $39, $99, $249 and $799 a month. A paid plan keeps the same questions re-checked every week with the history unbroken, and the plan-started email lists the plan's limits so nothing is a surprise.

Days 13 and 14: the end, and the extension

The trial-ends-tomorrow email shows the last reading. On day 14 the account becomes read-only: every question, answer, source and the improvement plan stay visible, and nothing is deleted. If you need more data before deciding, a seven-day extension is offered once, on the expiry screen. After that, an expired account still gets a free monthly check of 3 questions across the five engines, with an email of what moved, so the record continues at a smaller scale.

What a good trial looks like

Not a high score. A good trial ends with the questions right, the rivals listed, one page changed, one re-check that shows a citation, and a weekly email the team reads. The score follows the sources, on the engines' schedule, not the trial's. The methodology page explains why a first score is usually low and what the range around it means.

Start with the free check

Type your domain into the free audit. We ask two engines one question your buyers ask and show you who they named. Free, no account, about a minute. Then start the trial: five engines, your questions, every week, for 14 days, no card. Monthly counts of answers read, cited and named will be published on the Named Report page.

See what the engines say about you.

Presence, position, citations and sentiment — across ChatGPT, Gemini, Claude, Perplexity and Grok. Free for 14 days, no credit card.

Start Your Free Trial →