Content
The Sources tab tells you which pages AI engines read before answering. The Content tab is for writing the missing one: an article aimed at a specific prompt you’re losing.
Content is organised by prompt, not by topic. Pick a prompt from the list on the left, and the right pane shows the articles generated for it.
Two article modes
Benchmark blog
Supply one or more reference URLs plus your website URL. The generator reads the references and writes a piece pitched at the same question.
The references worth using are the ones from the Sources tab for that prompt: the pages engines actually cited. Benchmarking against a page nobody cites teaches the generator the wrong target.
Authority blog
Pick an article type, matched to the shape of the prompt:
| Type | Best for |
|---|---|
| Buyer’s Guide | ”best” / “top” / “recommended” prompts |
| Technology / Materials | ”why it works” / “how it works” prompts |
| Use Case / Scenario-Based | Prompts naming a specific audience, profession or scenario |
| Problem-Solving / FAQ | ”why” / “how” / “is it normal” prompts |
| Educational / Data-Informed | Broad informational queries |
Matching the type to the prompt’s intent matters. A buyer’s guide written for a scenario prompt answers a question nobody asked.
Generating, editing, exporting
Generation runs as a background job — start it and leave. Finished articles are listed with title, type and word count.
Every article opens in an editor. Treat the output as a strong draft, not a publishable page: the generator has your prompt, your reference material and your site, but not your product’s edge cases, your legal constraints or your voice. The pages that get cited are the ones a human made specific.
Articles export for publishing wherever your site lives. There is no auto-publishing — nothing is pushed to your site without you.
Where content actually helps
| Panel signal | Content response |
|---|---|
| Your domain barely appears in cited sources | You have no page answering these questions — start here |
| Losing scenario prompts, winning category | Write use-case pieces for the segments you lose |
| A competitor’s page cited across engines | Benchmark against it directly |
| Site Audit flags “no content section” | Any article is an improvement over none |
Content moves the AI half of the score slowly, on the retrieval cycle of each engine — expect the effect over runs, not overnight. Where speed matters more, the lever is distribution: placing content on domains engines already read.
Allowance
Article generation draws on a separate monthly allowance from probe runs. The two ledgers never convert into each other, in either direction. See Plans & usage.
See these numbers for your own brand The free plan probes every engine we support — no card required.
Start free