TL;DR. Raw visibility scores mean nothing without context. A 17% score is catastrophic in a global category and elite in a niche. Use the benchmark table below so every number lands as "top quartile in your niche," not just a number. Then read all 7 metrics, not just the headline.
Visibility benchmarks by market depth
| Market depth | Invisible (floor) | Standard (median) | Elite (top) |
|---|---|---|---|
| Global category (e.g. CRM software) | 1.0% | 2.8% | 15.0% |
| Niche in category (e.g. SMB-specific CRM) | 5.0% | 24.5% | 35.0% |
| Niche-within-niche (e.g. 10-FTE German SMB CRM) | 12.0% | 38.0% | 55%+ |
| Consumer retail | 1.9% | 4.5% | 17.99% |
| Local services | 1.2% | 7.4% | 45.8% |
The pitch lever: the deeper you go, the higher the ceiling. Tell the truth. If visibility in the global category is 2%, that is median. If it is 2% in a niche-within-a-niche, that is an emergency. The benchmark frame turns flat numbers into urgency.
The 7 metrics that matter
| # | Metric | What it tells you | Red flag |
|---|---|---|---|
| 1 | Visibility Score | Weighted brand presence. Your North Star. | Branded high, non-branded ~0% = false security |
| 2 | Position | Where in the answer you appear (#1–3 = elite) | Position #7+ or 'Not Found' on key prompts |
| 3 | Sentiment | How AI describes you | <55% positive = reputation risk |
| 4 | Mentions vs. Citations | Awareness gap vs. trust gap | High mentions / low citations = no source authority |
| 5 | Detection Rate | % of prompts where the brand was found at all | <90% on branded prompts = entity emergency |
| 6 | Share of Voice | Your mentions vs. competitors | Competitor owns the conversation |
| 7 | Share of Citations | Whose pages AI is quoting | Competitor is the 'source of truth' for the category |
Expert tip: always filter branded vs. non-branded. The blended score lies. Non-branded is where you find out whether the AI actually knows you matter in the category.
Do this now:
Open your Rankscale dashboard. Write down the 7 values for your top prompt group. Compare each against the benchmark row for your market depth. Circle the one that looks worst. That is your entry point into Step 3.
Full course: Module 3: Diagnose.
Start improving your AI visibility today with Rankscale.
Get started