Troubleshooting setup

Almost every first-day problem in RankX AI is one of five things: onboarding was never completed so the trial never started, a screen is empty because the run that fills it has not happened yet, a Google connection is not in the one state that yields numbers, a rate is blank because no check has a known verdict, or a spend was refused and the message says which of four reasons it was.

Find the symptom below.

"I have no credits and no trial"

Cause. The trial does not begin at signup. The clock and the credit grant both start when the onboarding package completes, which means the three platform-funded steps have run and you have finished the last required step. An account that signed up and stopped halfway has no trial running, and correctly shows no credits.

Fix. Go back into the product and finish onboarding. Nothing is lost, nothing needs resetting, and the grant lands the moment you complete it.

Why it works this way. The onboarding package is genuinely free and it costs RankX AI real money to run. Attaching the grant to completion rather than to signup is what makes it possible to give it away without a card. See the free trial.

"A screen is completely empty"

Empty is a schedule, not a fault, and which schedule depends on the screen.

The schedule is slower than you may expect on a trial. RankX AI evaluates scheduled runs nightly at 02:00 UTC, but a Website only runs when it is due, and while on trial the fastest cadence is once every seven days on every check family and every plan. So the first scheduled rank check on a trial account lands about a week after setup, not the next morning. See the free trial.

One case that is not a schedule. If AI Overviews is empty and you have tracked keywords with recent checks, the likely answer is that no AI Overview rendered for those keywords. AI Overview data is collected by the rank tracker on tracked keywords, so a Website with no tracked keywords will never populate it.

"A number is blank rather than zero"

That is deliberate and it is the most important convention in the product.

A rate divides by checks with a known verdict. If nothing has a known verdict yet, the rate is unknown, and RankX AI shows unknown rather than 0%. Zero would be a claim about your brand; blank is an honest statement about the sample.

The same rule runs everywhere: "not in the top 20" is different from "never checked", an AI Overview that was detected but not captured is "unknown" rather than "you were not mentioned", and a Website Audit check that could not run is "not assessed" rather than "passed".

Null is not zero explains where each of those states comes from and how to read a screen carrying them.

"Google Search Console shows nothing, or the last two days are missing"

The last day or two missing is correct. Search Console publishes on a two to three day delay at Google's end, not in RankX AI's sync. A healthy connection legitimately has no data for yesterday.

Nothing at all means the connection is not in the one state that yields figures. RankX AI classifies a Google connection into four states and only the last one returns numbers:

RankX AI refuses rather than reporting zero in the first three, and the message names which one you are in. Why traffic data is missing covers each state in full, and troubleshooting connections covers the connection itself.

"Something was refused for credits"

There are four refusal reasons and they have four different remedies. Conflating them wastes an afternoon, which is why the product keeps them apart:

The third only exists on the agency track, and it is worth reading carefully: the agency wallet may be full while a client is stopped, because a per-client budget is a ceiling on the shared wallet rather than a separate pot. See per-client credit budgets.

"The audit says a check was not assessed"

That is a real verdict and it is better than the alternative. A Website Audit runs on three different paths, and not every check can be produced by every path:

• Full crawl covers most checks.
• Instant page check runs a browser and sees things a plain crawl cannot.
• Sitemap analysis answers the sitemap questions and nothing else.

A check whose path did not run is reported as not assessed rather than as passed. Seven checks once rendered confident green ticks for checks that had never run, and this is the fix. What "a check did not run" means lists which paths can produce which verdicts.

"I cannot skip ahead in onboarding"

Correct, and deliberate. The order is enforced on the server because later steps read earlier answers: competitors feed the topic-cluster and prompt generators, and clusters are what a prompt attaches to. Navigating to a later URL sends you back to the step you are actually on.

"I picked the wrong track"

The Direct and Agency tracks are a property of the account and cannot be switched in settings. Contact support rather than signing up again: a second account gets a second wallet, and the free onboarding package is granted once per verified root domain, so the new account will not receive one.

Still stuck

Two things make a support conversation fast: the Website you are on, and the exact wording of the message you saw. Every refusal and every unavailable state in RankX AI is written to name its own cause, so quoting it usually skips a round trip.