The problem with a flat list
Every AI-visibility tool can tell you which prompts you're not cited on. The hard part is deciding which of those forty gaps to work on this week. Most tools hand you the list sorted by nothing in particular and call it 'insights.' Aim's job is to make that decision defensible: every gap gets a score, and the score is the product of four factors you can inspect individually.
Severity: how often a competitor wins the prompt without you
For each prompt, we look at its recent runs across every engine and count the runs where a competitor was cited and your brand was not. Severity is that count divided by total runs, so it ranges from 0 (a competitor never beats you here) to 1 (they beat you every single time). It's deliberately about competitor wins, not just your absence — a prompt where nobody gets cited is a much weaker opportunity than one where a rival is already the answer.
Priority: the worker's own high / medium / low call
When the tracking worker generates an action, it assigns a priority based on the nature of the gap. That becomes a simple multiplier — 3 for high, 2 for medium, 1 for low. It's coarse on purpose: it encodes the kind of gap, while severity encodes how badly you're losing it.
Recency: old gaps get a mild urgency bump
An action that has been sitting open for 30 days multiplies its score by up to 2×, scaling linearly from 1× on day zero. This stops a stale item from rotting at the bottom of the list forever — but it's capped, so age alone can't push a low-severity gap above a high-severity one.
Demand: prompt volume nudges, never dominates
Where we have a demand signal for a prompt, it adjusts the score between 0.7× (no measurable demand) and 1.3× (very high demand). Where we don't have the signal, the multiplier is exactly 1 — a missing lookup can never skew the ranking. Demand is a tiebreaker between similar gaps, not the reason a gap is on the list.
What happens when we haven't measured yet
If a prompt has no runs recorded yet, we genuinely don't know its severity. Guessing a middle value would let an untested action outrank one we've measured as nearly resolved, which defeats the point of an evidence-based ranking. So we fall back to priority alone and mark the item as 'not yet measured' in the interface — a visible 'we don't know' instead of a fabricated percentage.
Closing the loop
When you mark an action done, Aim snapshots your overall Share of Voice at that moment. Every time you open Aim after a new tracking run lands, any completed action whose result is older than the latest run gets re-measured automatically. There's no separate scheduler — the measured impact of your work simply keeps catching up with reality on its own.
See this on your own brand
Run the free check — same scoring the dashboard uses, no account needed.
Run free check