How We Review SaaS Tools
Every review on this site follows the same process: we find the tool, we pay for it ourselves, we use it for 30 days, and we score it against five weighted criteria. No exceptions, no shortcuts.
How we find tools
Tools come to us two ways: we discover them independently while researching a category, or founders submit them through our submission form. Both paths lead to the same review process — there is no fast lane for submitted tools.
We do not accept vendor introductions, press briefings, or "partnership" arrangements as a precondition for coverage. If a company contacts us to ask when their review will be published, the answer is when we get to it. Being loud about your product has no effect on our queue order.
Purchase requirement
Every tool reviewed on SaaS Discover is purchased with real money at the standard retail price available to any customer. We do not accept complimentary licenses, extended trials arranged by the vendor, or access granted for review purposes. We also do not disclose to vendors that we are reviewing them before publishing.
This matters because comped access changes what you see. Vendor-arranged access frequently comes with assigned account managers, escalated support SLAs, and hands-on onboarding that no paying customer at your tier would receive. We want to know what the tool is like for the person who found it on a Tuesday afternoon and entered a credit card.
The 30-day trial
Thirty days is the minimum. Some tools take longer; complex infrastructure products may run to 60 days before we have enough signal on edge cases. Here is what we specifically test during that window:
- Onboarding (days 1–3): Can a new user reach core value without talking to a human? We track time-to-first-useful-action, not time-to-signup. We also deliberately skip the guided tour on at least one test account to see what the cold path looks like.
- Support response (ongoing): We submit at least two support tickets — one simple question, one edge-case problem. We measure first response time and resolution quality, not just whether someone replied.
- Billing transparency (day 1 and at upgrade): We read the full pricing page before purchasing and again when we hit a limit. We note every instance where the pricing page does not match the in-app upgrade flow.
- Actual usage (days 4–28): We use the product for the workflow it claims to support. If it's a project management tool, we run a real project through it. If it's an analytics platform, we connect real data sources. Synthetic use cases produce synthetic verdicts.
- Cancellation (day 30): We cancel the subscription. Every time. We track how many clicks it takes, whether there is a required phone call, and whether the cancellation date aligns with what the billing page stated at signup.
Scoring rubric
Every tool is scored on five criteria. The weights reflect what we have found matters most to users who have been burned before.
Is the full cost knowable before you enter payment details? Are usage limits stated in absolute numbers, not vague tiers? Does the upgrade path make sense, or does it require a sales call to get a number?
Does the product do what it claims, reliably, at the plan level you purchased? We test advertised features, not demo features. A feature that requires an enterprise contract gets noted.
Time-to-value without human hand-holding. Documentation quality. Error message quality — does the product tell you what went wrong or just that something did?
First response time, resolution rate, and whether the support team has product knowledge or just a script. We weight async support (email, chat) because that's what most users at standard tiers actually get.
How hard is it to leave? Can you export your data in a usable format? Does cancellation take effect immediately or bury you in a 30-day notice cliff? This criterion is pass/fail below a threshold — tools with dark-pattern cancellation cannot score above 3/10 overall.
What gets rejected
Some tools fail before we finish the 30 days. The following are automatic disqualifiers — we will note them in a brief entry but will not produce a full review:
- Pricing requires a sales call to obtain a number. If you cannot put a price on a public page, you are optimizing for lock-in, not for customers.
- Cancellation requires a phone call, a retention chat, or a written request to a human. Cancel buttons exist. Use them.
- No product updates in the past 90 days with no public explanation. Stale tools without a stated reason are maintenance-mode tools, and we do not recommend maintenance-mode tools.
- Advertised features that are gated behind a tier not disclosed on the main pricing page. If the feature is in the hero banner but requires an enterprise contract, that is a misleading claim and we treat it as one.
- Data export that produces a CSV with internal IDs instead of human-readable fields. Your data should be portable. Non-portable data is a hostage situation.
How we handle updates
Every published review carries a "last checked" date. We re-check reviews on a quarterly cycle — pricing pages, feature availability, support SLA claims, and changelog activity. If something material has changed, we update the review and note what changed and when.
Ratings can go up or down. A tool that ships a genuinely better onboarding flow in Q2 will see that reflected in Q3. A tool that quietly changes its cancellation policy will also see that reflected. We do not grandfather old ratings because a product has been on the site a long time.
Delisting happens when a tool is discontinued, acquired and substantively changed, or fails our quarterly re-check on a disqualifier criterion. Delisted reviews stay up as historical records with a prominent notice at the top.
What we don't do
- No sponsored content. A company cannot pay to appear on this site in any form, including "featured" placements, "sponsored review" labels, or category sponsorships.
- No affiliate links. We do not earn a commission if you sign up for a tool we reviewed. There is no financial incentive tied to your clicking through.
- No paid "best of" lists. The ranking order in any comparison or list on this site reflects our scoring, not a vendor's marketing budget.
- No embargo agreements. We do not agree to hold a review until a vendor's launch date or coordinate publication with their PR calendar.
Our reviewers
Three people produce the majority of reviews on this site. Each brings a different professional background, which is the point — a tool that a developer finds obviously well-designed may be completely opaque to the business operator who has to configure it without engineering help.