How to Diagnose Which Gap You Have (the Diagnostic Decision Tree)

Last updated: June 24, 2026

Once you've picked a prompt to fix, you need to know which gap is causing the problem before you act. The diagnostic decision tree reads the signals on a flagged prompt and maps them to one of three gaps. This article shows you how to run it.

Read the 8 signals first

For your priority prompt, record one status per signal from the Rankscale dashboard:

Signal

Options

Detection Rate

Yes (all engines) / Partial / No

Position

#1-3 / #4-6 / #7+ / Not Found

Sentiment

Positive / Neutral / Mixed / Negative

Mentions (out of 7 daily runs)

6-7 / 3-5 / 1-2 / 0

Citations (out of 7 daily runs)

6-7 / 3-5 / 1-2 / 0

Stability

Stable / Volatile / Too early (no 7 runs completed)

Engine agreement

Consistent / Mixed / Conflicting

Competitor dominance

None / One wins / Multiple win

Collect first, interpret second. Don't second-guess while you're recording.

The 10 diagnostic rules

Compare your signals against the rules below. The first match is your diagnosis.

  1. Poor detection on a branded prompt → entity signal problem. Go to the SEO Gap.

  2. Poor detection on a non-branded prompt → topic association gap. Check whether you even have content answering this non-branded prompt (Content Gap) and are present on the 3rd party websites/blogs sourced (PR Gap).

  3. No mentions, citations, or sources → full invisibility on the topic. Branded means a crawler or entity problem; non-branded means a content- or PR-existence problem.

  4. Mentions missing, sources present → your pages are sourced by AI Search engines but that doesn't trigger a brand mention in the AI response. Check whether you can place your brand in the content that gets sourced.

  5. Mentions present, sources missing → Trust Gap. You most likely get mentioned because of training data, but your content isn't citable enough. Fix BLUF and justification (Content Gap). Look into the content gap article to find ways to strengthen your content.

  6. Weak position → competitors has content and/or presence on third-party websites AI prefers. Reverse-engineer their page (and improve it based on the content rules) and/or get listed on the third-party website.

  7. Volatile position across runs → source-pool instability; third-party coverage is thin. Go to the Citation Gap.

  8. Negative sentiment → reputation problem. Go to the Sentiment Gap and source-trace where the negative keywords come by clicking on the negative keyword and tracking which source was used to influence this negative representation of your brand.

  9. Engines disagree on sentiment → source conflict. The lowest-scoring engine is pulling from a worse source pool; check its cited domains.


If no single pattern triggers and you just have several weak signals, prioritize content freshness and quality as well as entity consistency.

Document the diagnosis

A diagnosis that lives in your head doesn't ship fixes. For each flagged prompt, write one line: the prompt string, the matching rule, the gap type (SEO / PR-Entity / PR-Citation / PR-Sentiment / Content-BLUF / Content-Justification / Content-Freshness / Content-Evidence), the next action, an owner, and a due date. That turns an observation into a ticket, and it gives you a before-state to measure against when you re-audit in four weeks.

Do this now

Pick your highest-priority flagged prompt, record the 8 signals, and match the pattern. Write down: "Diagnosis = [rule number and name]. Next fix = [SEO / Content / PR]." That's one line of documentation per prompt.

Related Articles

  • How to Prioritize Which Prompt to Fix First

  • Understanding the 3 Gaps Framework

  • Mentions vs. Citations vs. Sources: Spotting the Trust Gap and the Brand Placement Gap

  • How to Reverse-Engineer a Competitor's Winning Page