We help small teams pick software they won’t regret
Our mission and vision
Our mission
Help small teams pick software they won’t regret. We show the real price, the limits and the catch, each checked on the vendor’s own page and dated, so you can decide faster and pay less.
Our vision
No small business should overpay for software or lose months to the wrong tool. The true cost and the trade-offs should be plain before anyone buys.
What this site is
ToolsProof compares business software for small teams. We read the vendor's own pages by hand, write down what we find, and date every fact. We do not scrape review sites and we do not treat blog roundups as data.
Where we have not checked something, the page says Not verified. We would rather leave a gap than fill it with a guess.
How we earn, in one line
Some links to vendors pay us a commission. Our score never reads them. How we make money explains it in full.
What we will never do lists the tactics we ban and how we disclose paid links.
Contact us
ToolsProof is run by John.
Email John@toolsproof.com. A person reads every message.
- Found a wrong fact? Send the page and the vendor link. We check it and update the date.
- Vendors and affiliate teams: ask us anything about how we rank or label links. Payment can't change a score.
- Your data: see “Your rights” on the privacy page.
How we score tools
- Features40Key features for the category that the vendor confirms.
- Price30The cheapest verified entry price in the category scores highest.
- Free plan10A free plan, confirmed on the vendor’s page.
- Works with apps10How many apps it connects to, against the category leader.
- Up to date10How recently we checked. Older facts earn less.
This page is shown at /about#method. It must match the code in src/lib/core. If one
changes, change the other in the same commit.
The short version
Each tool gets a score from 0 to 100. The score is a sum of five parts. Every part comes from facts we read on the vendor’s own site. We date each fact and link to its source.
What we earn from a vendor is not a part of the score. The scoring code does not read our affiliate links at all, and our automated tests check this.
The five parts
| Part | Points | How a tool earns them |
|---|---|---|
| Features | 40 | Share of the key features for its category that we verified it has. |
| Price | 30 | The cheapest tool we compare gets all 30. The most costly gets 0. |
| Free plan | 10 | All 10 if we verified a free plan. If not, 0. |
| Integrations | 10 | Its count of listed integrations, set against the most in the group. |
| Freshness | 10 | All 10 if we checked every fact today. It drops to 0 at 90 days old. |
The points live in one place in the code: SCORE_WEIGHTS in src/lib/core/scoring.ts.
Features
Each category has its own list of key features. Every tool in a category is checked on the same list. Right now each feature on a list counts the same. If we ever make one feature count more, we will say which one and why on this page.
A feature we could not confirm earns nothing. We do not guess.
Price
We take the lowest price we verified for one user for one year. We look at every plan, paid by the month or by the year, and use the cheapest.
Tools in between the cheapest and the most costly sit on a straight line. A tool that costs half way between the two gets half the points. If all tools cost the same, all get full points. A tool with no verified price gets 0 price points and shows “Not verified”.
Freshness
We find the oldest fact we hold for the tool. The older it is, the fewer points. A fact checked 45 days ago earns half. Past 90 days it earns nothing, and our build fails until we check it again.
How we rank
You can sort a list three ways.
- Best fit (the default): highest score first. If two tools tie, the cheaper one goes first. If they still tie, we sort by name.
- Price: cheapest first. Tools with no verified price go last. Ties go to the higher score, then to the name.
- A to Z: by name.
The order never depends on what a vendor pays us.
Ranked for what frustrates you
On the home page you can pick one frustration. We then move weight onto the part of the score that answers it. The facts and the five parts stay exactly the same; only how much each part counts changes, and the page says the list is ranked for you, with the standard ranking one tap away.
| Frustration | What changes |
|---|---|
| Costs keep creeping up | Price counts 70 instead of 30 |
| Too hard to get started | Free plan counts 50 instead of 10 |
| It doesn’t work with my other apps | Integrations count 50 instead of 10 |
| I can’t trust the info I find | Up to date counts 50 instead of 10 |
| Too many features I never use | Features count 10 instead of 40 |
The total is then scaled back to 100. A frustration is only offered when at least two tools in the category have verified data on that part; for example, no cost ranking until prices are verified. Commission never plays a part.
Our pick
Our pick in a comparison is the tool that ranks first for best fit. With each pick we say who it is for, who it is not for, and one real trade-off. If we have not written those three things for the top tool, we show no pick at all.
Cost over time
The cost tool adds up what a plan costs for your team for 1 year and for 5 years.
- You choose the team size, monthly or yearly billing, and any add-ons. We have not verified any vendor’s add-on prices, so you type your own. Each add-on is added to every tool the same way, priced per month with no yearly discount.
- A per-user price is multiplied by your team size. A flat price is not.
- The 5-year cost is the 1-year cost times five. It assumes the price stays the same. Prices often rise, so treat it as a floor.
- “Save X% with yearly billing” is worked out from the plan’s own two prices. We do not copy the vendor’s claim.
- If we lack a price the sum needs, we leave that plan out and say why. We never fill the gap with a guess.
- When a tool has several plans, the calculator shows the cheapest plan we can price and says how many plans it could not compare. Tools are listed cheapest first by their five-year cost. Commission is never read.
How a fact gets in
AI helps us, but it never decides.
- A person reads the vendor’s page and saves its text.
- An AI model suggests facts from that text. It must quote the exact words. If the quote is not on the page, or leaves out a number it claims, we throw the fact away.
- A person checks each suggested fact against the page and accepts or rejects it. Only accepted facts are saved, with the page link and the date.
- A tool shows on our pages only when every fact it lists has a source and a date.
- Plain code, not AI, scores and ranks tools. Facts older than 60 days go on a list to check again. After 90 days, the site will not build until they are checked.
What the score cannot tell you
- A score is only fair inside one category, and only among the tools on that page. Price and integration points depend on which tools are compared.
- Price is judged for one user. A flat-price tool can be a better deal for a big team than its price points suggest. Use the cost tool for your own team size.
- More integrations is not the same as the one integration you need.
- We do not score ease of use, support, or speed. We have no fair way to measure them.