Questions, answered straight

Real public reviews, and the source depends on the adapter. A local business is read from its Google reviews via the Places API; a D2C brand from its Amazon product reviews; a mobile app from its App Store and Google Play reviews. ryvl resolves the subject you name, gathers its competitors, and reads everyone's actual reviews. There is no mock data anywhere in a report.

Yes. ryvl is one engine with three input adapters. Point it at a local business (a Google Maps link), a D2C brand (an Amazon ASIN), or a mobile app (an App Store or Google Play link), and the same pipeline produces the gap map, quotes, and the marketer and operator outputs, in the right voice for that market.

Yes, and it is enforced. Every quote shown is checked to be an exact substring of a real review before a report can ship. A single mismatch fails the build, so a report can never show words a customer did not write.

The scores and themes are read from the review text, and a subject is rated on a dimension only when its reviews actually speak to it. When they are silent, the cell stays blank rather than guessing. The bigger limit is volume, and it depends on the source: Google returns only about five reviews per business, while Amazon and the app stores return many more. Either way, treat a report as a sharp read on recent sentiment, not a census.

It depends on the source. The Google Places API caps local-business reviews at about five, with no pagination, which is still enough to surface the loudest recurring patterns. The Amazon and app-store adapters read a much deeper window, often fifty to a few hundred recent reviews per subject, so those reports key on richer detractor patterns. Anything that needs trends over time for the local adapter waits until a denser Google source is in place.

The core of the product. ryvl finds the single sharpest weakness in a competitor and shows it two ways on one screen: a marketing campaign with ad copy, and a tactical action plan with clear next steps. Both halves are pinned to the same weakness in code, so they never drift onto different problems.

No. The review replies are drafts only. You read them, tweak them, and post them yourself. Auto-posting to Google needs a longer approval process, so it is deliberately left out for now.

A local or Amazon run is around two minutes. An app run takes longer, often five to ten minutes, because it reads a deeper review window. While it runs the page shows staged progress, and signed-in users can have the finished report emailed.

All three adapters are real and live: local business (Google), D2C (Amazon), and mobile apps (App Store and Google Play), each with real committed reports and a run-your-own path. The engine and every module are live. The workflow connectors (Jira, Slack, posting replies, exports) are on the roadmap and labeled as such on the Integrations page. Pricing checkout is not wired up yet either.

A live run keeps a short-lived job record so the page can poll for the result, and it expires on its own. Reports are built from public reviews. You can create a free account to save the businesses, products, and apps you track; sign-in is by Google or an email magic link.