Be the answer, not the link.
See whether your client appears when people ask AI engines about their category. Review the prompts, mentions and cited pages, then decide which content needs attention.
Review results across ChatGPT, Perplexity, Gemini, Copilot, Claude and Google AI Overviews.
Read the evidence behind the harvest.
The Harvest Index is our way of framing progress in AI visibility. Assess it alongside the actual prompts, brand mentions and cited URLs. A sampled visibility score is not a measure of all AI answers or a guarantee of traffic.
- Keep a consistent set of relevant client questions and target markets
- Record the engine, prompt and scan date alongside each result
- Separate brand mentions from links that cite a client page
- Compare repeated observations and check search traffic and conversions separately
Why AEO matters in 2026
When a prospect asks an AI engine instead of typing into a search box, the answer either mentions you or it doesn't. AEO is the discipline of being the cited answer — and it is now as important as ranking.
What we track (the breakdown behind the index)
Review the available scan results for your tracked keywords and compare like-for-like prompts over time.
Share of Voice
How often the brand appears in the sampled responses. Interpret the result against the prompts and engines included in the scan.
Citations
Which URLs appear as sources in the sampled answers. A cited URL is different from a brand mention without a link.
Competitor mentions
See which competitors appear in answers about the client category and inspect the sources behind those responses.
Sentiment
Review how the sampled answers describe your brand, alongside the original response and its context.
Tracking Prompts
Custom queries you define; the system runs them weekly across your tracked keywords (up to 5). The actual ChatGPT / Perplexity conversations are surfaced — see what AI said, not just a score.
Engine coverage
Six engines tracked weekly: ChatGPT, Perplexity, Gemini, Copilot, Claude, and Google AI Overviews — bundled in the standard plan, not behind a higher tier.
Four surfaces. One workflow.
Use the visibility dashboard to investigate results and decide what to improve.
Overview
Review visibility trends for the active client workspace and investigate which engines and prompts changed.
Citations
Which URLs AI engines actually cited — and which articles still aren't being picked up. The action list behind the index.
Competitors
Who else gets mentioned in your category, where the gap is widest, and which clusters need supporting depth to close it.
Recommendations
Concrete next moves, ranked: schema completeness, content depth, entity clarity, citation-worthiness fixes. The 20-point AEO audit translates into a queue.
20-point AEO audit
We score your site on 20 dimensions that determine AI-engine visibility — schema completeness, content depth, entity clarity, citation-worthiness, and more. Categories include:
- Entity & brand clarity
- Structured data coverage
- Answer-readiness of content
- Citation-worthiness
- Topical authority signals
- Competitive share of voice
Track — and act
Writegarden tracks visibility and suggests how to improve it. It surfaces concrete recommendations; you stay in control of what ships. It does not silently auto-optimize your site.
A weekly digest you'll actually read
Every Monday: where you gained or lost AI visibility, which competitors moved, and what to do about it — with concrete recommendations.
Content and visibility in one workflow
Connect the content you publish with the search and AI visibility evidence you review for each client.
Coming to AI Visibility
All six engines ship today. Next: deeper per-engine prompt-strategy customisation, expanded tracked-keyword sets beyond five, named-entity tracking across categories, and competitive intelligence reports for Enterprise. We ship features as they harden — and we'd rather name the roadmap than blur the scope.
What buyers ask about AEO.
How should I interpret the Harvest Index?
Use it as a starting point for reviewing AI visibility. Inspect the underlying prompts, sampled responses, mentions and citations before drawing conclusions. It is not a measure of every AI answer or a guarantee of rankings, traffic or leads.
Six engines — including Claude and Google AI Overviews?
Writegarden includes tracking for ChatGPT, Perplexity, Gemini, Copilot, Claude and Google AI Overviews. Check the engines and completed scan dates in the dashboard when interpreting a result.
How often are scans run?
Scans run on a weekly cadence for tracked keywords. Compare the actual scan dates and covered engines when reviewing changes.
Can I see the actual AI conversations where my brand appeared?
Yes — the Citations tab surfaces the actual prompts and the engine response with the citations highlighted. Much more visceral than a score in isolation; you see what AI said about your category, not just where you landed.