How to Prioritize Which Search Term to Fix First
Last updated: August 4, 2026
Once several prompts are flagged, the question is where to start. Fixing in the wrong order wastes effort, since some wins compound and others don't. This article gives you the order of attack Rankscale recommends.
How status is set
Prompts get one of three statuses: good, medium, or bad. How that's decided depends on the prompt type.
Branded prompts use fixed thresholds:
Good if Detection Rate ≥ 95%
Medium if Detection Rate is 90–95%
Bad if Detection Rate < 90%
Non-branded prompts are benchmarked against your category depth:
Good if above Standard (the category median)
Medium if between Floor and Standard
Bad if below Floor
The order of attack
Work the list top to bottom. Don't jump ahead.
Bad results first, in this order:
Branded
Non-branded transactional + niche
Non-branded consideration + niche
Non-branded informational + niche
Non-branded transactional + category
Non-branded consideration + category
Non-branded informational + category
Then medium results, in the same order.
Why branded comes first, always
If AI can't find you on your own brand name in your own niche, nothing else you fix will compound. A branded bad result is a signal that your entity isn't trusted by AI engines as a real, verifiable thing. Entity recognition is the foundation for every non-branded win downstream, so fix it before touching anything else.
Why niche comes before category
Niche prompts have fewer competitors, so a small content investment moves visibility faster. You prove results in weeks rather than quarters, and that earns you the budget and headcount to take on broader category prompts later. The rule of thumb: more specific equals faster visibility proof.
Why transactional comes before informational
Conversion potential rises with intent. Transactional prompts sit closest to a buying decision, so winning there drives pipeline directly. Informational wins build awareness, which matters, but awareness without pipeline is hard to defend in a budget review.
Do this now
Open the Brand Dashboard tab in Rankscale, filter based on tags and flag your prompts using the order above. Pick the top prompts from the highest group with a bad result. That's your diagnostic target.
Related Articles
How to Diagnose Which Gap You Have
Benchmarking Your Visibility Score by Category Depth
How to Read Your Rankscale Dashboard
Understanding the 3 Gaps Framework