Choosing

Six checks before tapping install, ninety seconds, answerable from the listing. What does it cost annually. What happens in thirty days if I do nothing. Is there a free tier that would do. What am I replacing. What happens to my data if I leave. Will I use this twice.

The prominent parts of a store listing are the least informative. The average rating is prompted at moments of satisfaction and clustered so tightly across categories that it discriminates nothing — a 4.8 tells you the prompt is well timed. The useful parts require scrolling: subscription terms, recent one-star reviews, the last update date, and the data disclosure. A useful contrast from the workplace-software market is this resource; it shows how similar feature and buying decisions change when the purchaser is a team.

Distinguish complaints about the product from complaints about the business. The first are opinions and yours may differ. Charged after cancelling, no reply from support, price rose without notice — those are facts about how the company operates and they predict your experience regardless of taste.

Export is the feature nobody checks and the one that decides whether your data is yours. Test it in week one, before anything is at stake. A free tier that accumulates material with no export is the strongest lock-in in this category, because it is entered without a decision. For an independent software-selection perspective, WIRED app coverage is another useful reference.

And sometimes the answer is no app. The operating system already scans, signs, converts and compresses; a habit is not fixed by a fifth app; paper works; and if you cannot name what would improve, there is nothing to check at the four-week review. For an independent software-selection perspective, monday.com productivity guides is another useful reference.

The comparison that never appears in a recommendation is against installing nothing at all. Doing nothing costs whatever the current friction costs, which is usually smaller than it feels in the moment of irritation that prompted the search — and no listicle can reach that conclusion, because it earns nothing.

• Before You Install Anything

  Six checks that take ninety seconds and prevent most of the recurring charges people later cannot account for at all.

• Your Data If You Leave

  Export is the one feature worth checking before you accumulate anything. Four questions, and what a useless export looks like.

• When Free Is Enough

  Free tiers are designed to be almost sufficient. How to tell whether the gap is real, and the four limits that actually bite.

• One-Time or Subscription

  One-off software mostly disappeared for reasons that are not cynical. When each model suits you, and the break-even arithmetic.

• Reading a Store Listing

  The rating is the least informative element. Six things worth reading, in the order that tells you whether to install.

• Replacing Four With One

  Consolidation saves money and concentrates risk. When it works, when the all-in-one is worse than four specialists, and what to check.

• What Reviews Are Worth

  The average rating is close to uninformative and the reason is structural. Which parts of the review section actually carry signal.

• The App You Use Twice

  A large share of installs serve a task that happens once. Five signs you are in that situation, and what to do instead.

• When Not to Install

  Six situations where the correct answer is no app at all, and the comparison that never appears in a recommendation.