New reviews, de-duplicated
Each run collects reviews published since the last one, for you and each rival, and drops ones already seen.
Product · Tools
Product · Intel loops · recurring, in build
Solutions · By problem
Solutions · Online stores
Resources · Learn
Resources · Guides and ideas
Resources · Company
Your rivals' customers publish what annoys them about your rivals. Almost nobody reads it. This loop will collect new reviews on a cadence, sort them into themes, and show which themes are moving, quoted from the reviews themselves.
In build · not on sale Planned on code that already runs inside the reports. The loop itself is not built, does not run for customers and is not on sale.
Review themes, compared with the last run
A loop that collects new reviews for you and the rivals you name, groups them into themes and shows which themes grew or shrank since the last run.
Your reviews tell you what your customers value and what annoys them. Your rivals' reviews tell you the same about theirs. The difference is often the gap you can win on. Reading both by hand, every month, is a job nobody does twice.
The reviews watch will do it on a cadence. For you and each rival you name it will collect the reviews published since the last run, remove duplicates, and place each one in a theme: delivery, price, quality, communication, whatever the reviews themselves are about. Then it will compare against the last run and show which themes grew or shrank, with the reviews that moved them quoted verbatim so you can judge the evidence.
It is built from code that already runs. The reports read reviews today: the Google Places connector, the review-source readers and the review analysis that groups comments into themes. The loop adds what a one-off read does not need: a history of reviews seen, de-duplication across runs, and a comparison of themes over time.
One lesson from earlier work is designed in from the start. Review sites change their pages. A scraper that breaks silently reports "no new reviews" when the real answer is "we could not read the source". Every source will carry a health flag, so an unreadable source is shown as unreadable. This loop is in build: nothing runs for customers and it is not on sale.
Each run collects reviews published since the last one, for you and each rival, and drops ones already seen.
Reviews are grouped into themes, and every theme links back to the reviews in it, quoted as written.
Themes that grew or shrank since the last run, per business, so you can see a problem starting before it shows in the average.
Each review source is checked on every run. One that cannot be read is flagged as unreadable rather than counted as quiet.
What customers praise and criticise, for you and each rival, grouped and quoted.
Which themes gained or lost reviews since the previous run, with the reviews behind the change.
Themes where rivals' customers are unhappy and yours are not: the points worth saying out loud in your own marketing.
Per-source health on every run, so a missing review is a known gap rather than a silent one.
You and the rivals you want watched, matched to their review listings.
How often new reviews are collected and themed, in whole days.
See what moved since the last run and open the reviews behind each theme.
The reviews watch is being built. When it runs in production the pricing page will say which plans include it; until then nothing is charged and nothing is promised.
Not as a loop. The free Insight Report and an Intel loop already read reviews once; the recurring watch is in build. Join the loops list to hear when it ships.
Google reviews first, through the Places API. Other sources only where they can be read reliably, each with a health flag on every run.
No. The Places API returns a selection of reviews per listing, not every review ever written. The watch builds its own history from what it sees on each run.
No. It reads and groups reviews. Replying stays with you.
Nothing is priced yet. When it runs in production, the pricing page will say which plans include it.
Planned on code that already runs inside the reports. The loop itself is not built, does not run for customers and is not on sale.