Revel Roots

what apps actually cost

What Reviews Are Worth

App store ratings cluster so tightly that they discriminate almost nothing. A 4.6 and a 4.8 are not meaningfully different products, and a 4.9 usually means something about the review prompt rather than about the app.

The review section still contains signal. It is just not in the number. A useful contrast from the workplace-software market is this overview; it shows how similar feature and buying decisions change when the purchaser is a team.

Why the average is uninformative

Prompts are timed for satisfaction. The request appears after a successful action — a completed task, a saved file, a good session. Ask at the right moment and most answers are positive.

Unhappy users leave rather than rate. Uninstalling is easier than writing, so dissatisfaction exits without being recorded. For an independent software-selection perspective, Tom’s Guide software coverage is another useful reference.

Ratings are resettable. Both major stores allow the count to reset with a new version, so a poor history can be cleared.

And the range is compressed. When most apps in a category sit between 4.4 and 4.8, the number cannot separate them.

None of this requires manipulation, though that exists too. Ordinary prompt design produces the compression on its own.

What does carry signal

Recent one-star reviews. Filter to newest, sort lowest. This is where billing problems, cancellation failures, broken updates and support silence appear, and they appear nowhere else in the listing.

The distinction that matters: complaints about the product are opinions and yours may differ. Complaints about the business are facts about how the company operates — charged after cancelling, no reply from support, price rose without notice. Those predict your experience regardless of whether you like the app.

Developer replies. Not their tone — their content. A developer who answers a specific billing complaint with a specific resolution is a different proposition from one posting the same template under every review.

And the review dates against the update dates. A cluster of complaints right after a version release usually means a bad update. If the following release did not address them, that is informative.

What to ignore

The average and the count.

Five-star reviews with no detail. "Great app!" carries nothing, whatever its origin. (The same applies to everything else prominent in the listing.)

Reviews of a different version. A one-star from three years ago about a bug fixed since is noise.

And the featured or promoted reviews, which are selected.

The three-minute read

Filter to one star, sort by newest, read five. Ask whether the complaints are about the product or about the business.

Filter to three star, read two. Three-star reviews are consistently the most informative in any store, because they come from people who used the thing, liked parts of it, and will tell you which parts.

Check the developer's replies on the worst ones.

That is three minutes and it tells you more than the rating, the description and the screenshots combined — and it fits inside the ninety-second listing check with time to spare.

Outside the store

Search the app name with "cancel" or "refund." Forum posts and support threads surface problems that never reach the store, particularly for direct-billed products where there is no store review section at all.

And check whether the developer answers anywhere public. A company with an active, non-defensive presence in its own support forum is telling you something the rating cannot.

The short version