AEO Optima Docs
Features

Trends

Track how your brand's AI visibility changes over time with time-series charts across visibility score, mentions, rank, and sentiment.

Overview

The Trends page gives you a longitudinal view of your brand's performance across AI engines. Where Analytics shows a point-in-time snapshot, Trends shows how those metrics move over days, weeks, and months so you can spot regressions, measure campaign impact, and report progress to stakeholders.

What you'll see

ChartWhat it tells you
Visibility Score over timeThe percentage of snapshots mentioning your brand, plotted day-by-day. Rising is good.
Mention count over timeRaw mention volume across all LLMs. Spikes often correlate with newly published content or campaigns.
Average rank positionWhere your brand appears in numbered lists. Lower is better (1 = first mentioned).
Sentiment trendPositive / neutral / negative ratios over time, so you can see whether AI engines are "warming up" to your brand.
Per-LLM breakdownThe same metrics split by ChatGPT, Claude, Gemini, Perplexity, etc., so you can see which engines are moving.

Time range controls

Use the time range selector at the top of the page to switch between:

  • 7 days — Recent week, best for tracking active campaigns
  • 30 days — Monthly view, the default for most reporting
  • 90 days — Quarterly trends, useful for detecting long-running shifts
  • Custom range — Pick any start/end date

All charts update in sync when the range changes.

How to read the charts

  • Solid lines show metric values per day. Days with no captures are emitted as null values, and Recharts draws a continuous line through the gap (connectNulls) — so sparse schedules don't create misleading visual breaks, but the gap is still visible on hover as "no data."
  • Hover any point to see the exact value, the number of snapshots that contributed to it, and the breakdown per LLM. Gap points explicitly show "No captures on this day."
  • Zero visibility days are different from gap days — a zero means captures ran but no model mentioned your brand. Check Schedules if you expected data on gap days.

Segment filtering

The Trends page has a segment toggle at the top that filters every chart by prompt type: All, Branded, Non-Branded, or Competitor. The selection syncs to the URL (?segment=non-branded) so the filtered view is bookmarkable and shareable.

  • Switch to Non-Branded to see your true organic AI discoverability trajectory.
  • Switch to Branded to see how AI describes your brand on explicit mentions.
  • Switch to Competitor to see how often you co-appear with or lose to competitors over time.

Dynamic period comparison labels

Trend indicators like "vs previous period" adapt automatically to your selected date range:

  • 7-day range → "vs prior 7 days"
  • 30-day range → "vs prior 30 days"
  • Custom 45-day range → "vs prior 45 days"

Labels are never hardcoded, so the comparison is always apples-to-apples for whatever window you pick.

A sudden drop in visibility usually has one of these causes:

  1. A competitor published new content that displaced your brand in AI responses. Check the Competitors page to see if their mentions rose.
  2. An LLM model was updated and its retraining shifted what it knows about your category. Look for drops that only affect one engine.
  3. Your website changed in a way that removed signals AI crawlers were using. Run a GEO Audit to compare before/after.
  4. Your prompts drifted away from what your audience actually asks. Review the prompt list for stale entries.

Best practices

  • Capture on a consistent schedule. Weekly or daily schedules produce cleaner trend lines than ad-hoc captures. See Schedules.
  • Don't over-interpret short windows. LLM responses have natural variance — a single day's dip is usually noise. Look for trends that persist across 3+ captures.
  • Cross-reference with Sentiment. A flat visibility score with declining sentiment is often a leading indicator of a deeper problem.
  • Export to PDF using the Reports feature when sharing trends with non-platform users.

On this page