There is too much software. There has always been too much software, but in 2026 the situation has reached the point where a serious person can run out of evenings before they run out of tools to try. The cure is not to give every new app a fair, thorough trial. The cure is to get fast at saying no.
What follows is the checklist I use, refined over a few hundred hours of reviewing submissions for the dotstore catalog. It takes about five minutes per app. It produces, in my experience, the same decision I would have reached after a full afternoon of testing, in roughly nineteen times out of twenty. The other twenty times it is wrong, but those are mostly cases where the app is more interesting than it looks, not the other way round — so the cost of being wrong is small.
Minute one: the landing page
Read the headline. Read the subhead. If you cannot describe, in one sentence, what the product does for you, the marketing has failed and the product is likely to be confused about itself too. The clearest products always have the clearest landing pages. The reverse is also reliable — a vague, AI-adjacent headline ('Reimagine your workflow with intelligence') is a tell.
Note the social proof. Customer logos that are real companies you have heard of. A user count that is honest about whether it means 'free signups' or 'paying customers.' Press quotes from publications that still exist. Absence of all three is not damning, but presence is reassuring.
Check the pricing page. Skim it. The shape of the pricing tells you what the company thinks of you. A clear free tier with a sensible paid tier is a healthy sign. A free trial with no permanent free option means the company has decided it is enterprise-shaped, which is fine, but the product is unlikely to feel built for solo or small-team use. A 'contact sales' wall on every meaningful tier is a different product than the marketing implies it is.
Minute two: signup and onboarding
Time how long it takes from clicking 'get started' to seeing the actual product. Anything under thirty seconds is good. Anything over two minutes is a sign the company has not respected your time, and respect for time tends to be a habit that runs through the entire product.
Watch for these specific anti-patterns:
- A six-step setup wizard before you have seen the product. The company is using onboarding as a sales-qualified-lead form.
- A credit card requirement for a free trial. This is allowed but should make you suspicious — most healthy free trials do not need one.
- An aggressive 'invite your team' prompt before you understand what the product does. Viral growth optimization at the expense of first-experience quality.
- A demo data set that is suspiciously perfect, hiding the difficulty of getting your own real data in.
Minute three: the empty state
Close the demo data. Get the app into its blank, no-content state. This is the most honest view of any tool.
A good empty state explains, gently, what to do next. It offers a single obvious action, plus a couple of optional ones. It is not a wall of feature tiles. It is not a video tutorial that auto-plays. It is the product itself, lightly hinting at how to start.
The empty state is also where you will find the design tax. If the empty state is uglier than the marketing screenshots, you are looking at a company that spends more on landing-page design than product design. This is a category of company, and it has a consistent failure mode: the product never quite catches up to the brand.
Minute four: do one real thing
Pick the single most important thing the product is supposed to do. Do it. Not a demo version of it — the real thing, with your own input. If it is a writing app, write a real paragraph. If it is a project tool, set up a real project. If it is an AI tool, ask a real question.
What you are checking for is whether the basic loop is good. Most products are evaluated on the breadth of their features. This is mostly a mistake. The breadth of features tells you what the company has prioritized. The quality of the core loop tells you whether you will enjoy using the product on a Tuesday afternoon in three months.
Minute five: the exit
Find the export. Find the account-deletion option. Find the data ownership statement. These should all be easy to find. A product that hides them is a product that has thought about retention as a moat rather than a result.
If the export is a CSV download with all your data and a sensible schema, the company has earned a small amount of trust. If the export is a 'request your data via email' workflow, the company has not. If there is no export at all, walk away — your future self will thank you.
Bonus points for an account-deletion option that actually deletes the account, in the product itself, without a support ticket. This is a small bar that most companies still fail to clear in 2026.
Things that take longer than five minutes
The checklist above is for triage. It tells you whether an app deserves more of your time. It does not tell you whether the app is right for you. For that, you need to actually use it for a week — ideally on the project that triggered the search in the first place.
Things you cannot evaluate in five minutes, no matter how good your method:
- Performance under your real workload. Tools that demo well on twenty items sometimes collapse on two thousand.
- Quality of support. The first time you hit an edge case is the test of the support team, not the marketing page.
- How the product evolves. Some teams ship steadily. Some go quiet for eighteen months and come back with a redesign that breaks everything.
- Cultural fit with your team. A tool that is technically excellent and culturally wrong is worse than a mediocre tool that everyone is happy to use.
These take weeks to assess, sometimes months. They are also the things you only need to assess for the small set of apps that survived the first five minutes.
A practical scorecard
If you want a single rubric, here is the one I keep stuck to the side of my monitor:
- 1Can I describe what it does in one sentence? (Yes / sort of / no)
- 2Did the signup respect my time? (Under 30s / under 2min / no)
- 3Is the empty state confident? (Yes / busy / hostile)
- 4Did the core action feel good? (Genuinely good / fine / awkward)
- 5Can I leave easily? (Yes / friction / no)
Three or more 'yes' answers is enough to spend a week with the product. Two or fewer is probably enough information to move on. One is a clear no, and you have just saved yourself an afternoon.
We use a longer version of this checklist for every submission to the catalog. The five-minute version is the one I actually use when a friend asks 'have you seen this new app?' on a Saturday. It is not perfect. It is fast, and being fast is sometimes more useful than being perfect, especially when there are forty new apps a week.
Tags