How I Score Tools at Clear-Cut Reviews

Last updated: October 10, 2026

Every tool I review gets a score out of 5, built the same way every time. This page explains exactly how that number is put together, so you can trust it, or argue with it. There’s no secret sauce and nothing bent to suit a conclusion. Here’s the whole method, rounding rule included.

The four things I score

Every tool is rated on the same four criteria, each out of 5:

Each of these is scored in half-point steps, from 1 to 5. No quarter-points like 4.3 or 4.8, because that’s false precision: I can’t honestly tell the difference between a 4.3 and a 4.4 tool.

What the scores mean

Each score maps to a verdict badge you’ll see on the review:

I rarely publish anything below 3. If a tool is that flawed, it usually doesn’t make a roundup in the first place. But when one does, I still say so.

How the number is calculated

The overall score is the plain average of the four:

Overall = (Ease of use + Value + Features + Support) ÷ 4

I then round to the nearest half-point. For example, a tool scoring Ease of use 4, Value 5, Features 5, and Support 3.5 averages to 4.375, which rounds to 4.5 / 5. The overall is never hand-picked to fit a conclusion; it’s whatever the math produces from the four sub-scores you can see on every review.

When an average lands exactly halfway, like 4.25 sitting dead between 4 and 4.5, I round up. It’s a fixed tie-break, written here in advance rather than decided case by case, and it applies to every tool in both directions: a tool averaging 2.25 rounds up to 2.5 and still carries a Disappointing badge. The sub-scores are always published, so you can do the math yourself and disagree with the rounding if you want to.

How I order tools that score the same

In a ranked roundup, tools regularly end up with the same published score. Somebody still has to be listed first, so here’s the rule, again written in advance rather than decided tool by tool.

When published scores tie, the unrounded average breaks it. In the free Canva alternatives roundup, Adobe Express and VistaCreate both publish 4.5 out of 5, and Picsart and Microsoft Designer both publish 4. Their four sub-scores add up to 17.5 and 17, then 16.5 and 15, so they are listed in that order.

This isn’t sneaking in precision I just said I don’t have. I won’t publish a tool as a “4.1”, because I can’t honestly tell a 4.1 tool from a 4.2 one. But adding up four scores I do stand behind is arithmetic, not a finer judgment, and it would be strange to throw that away and pick an order on gut feel instead.

Every sub-score sits on the page next to every tool. So you can add them up and check my ordering yourself, and tell me I got it wrong. That’s the point of putting the rule here: the alternative is an order you’d have no way to verify.

When the unrounded averages tie as well, I say so, and then break it on a stated criterion: the one closest to the question that post is asking. In a free-plan roundup, that is Value, because the question is how much you can get done without paying. Naming the criterion is the part that matters. It’s the difference between a reasoned order and a coin toss you can’t see.

How I test

I sign up and use the tools myself, with my own files rather than the sample assets a company hands out to make itself look good. Some of that is real work. Some of it is a design built specifically to find where a limit sits, which is the part a real project rarely does on purpose. For free-plan roundups, that means actually hitting the limits: exporting files, checking for watermarks, and seeing where the paywall really sits. Where I couldn’t test something personally, I say so plainly and flag it.

The order is always the same. Where a tool has a free plan, I sign up on that first and review it in full, hitting every paywall along the way, before I pay for anything. Only after that do I upgrade and test the paid plan on its own, so a review can tell you where the free plan stops and what the money actually buys. Sometimes a vendor gives me the paid access for a test instead, a free month for example; when that happens the review says so in its disclosure, the vendor doesn’t see the review before you do, and the price, checkout and cancellation are still checked on the public pages or named as untested.

Every score is built from itemized deductions

A score here is never an impression. Each tool starts at 5 in all four categories and loses points only for specific, tested failures, each one written down with what it costs you and how I found it. That is why scores move in half-points and why I can tell you exactly what a missing half-point was for.

The part that matters more: every deduction is checked against every tool in the comparison. If one tool loses a half-point for hiding its help page, I go and look for the help page on all of them. Without that, a tool gets marked down simply because I happened to look harder at it, and the ranking becomes a record of where my attention went rather than what the tools do.

I also don’t deduct for a tool lacking something you don’t need. A capability one tool has and the others don’t earns that tool a point in its favor; it doesn’t earn every other tool a penalty. Something counts against a tool only when its absence makes a job you’d actually do harder or impossible.

Support is tested, not read off a pricing page

On every plan a tool offers, I check four things: whether the help page can be reached from inside the product and how long that takes, whether the assistant will say in writing that a human is reachable, how far the escalation path goes before you’d have to submit anything, and whether there’s a public community that actually answers questions. A missing community costs a tool half a point only when there’s also no way to reach a person, by chat, email, phone or a form someone answers.

That method exists because the old one published two wrong scores in a single evening. Canva’s pricing page listed live chat on Pro, so Pro scored full marks for support. A paying subscriber then couldn’t find chat, a ticket form or a contact link anywhere. Adobe, which I’d marked down for having no human to escalate to, turned out to offer a Premium advisor on request. Both claims came from reading plan tables instead of clicking, and a plan table is a claim rather than evidence.

Public ratings on G2, Capterra and Trustpilot still have a job, but it’s a cross-check rather than the source. Review platforms skew negative, because people rarely post five stars about a billing cycle that went fine, so I read what the complaints are actually about before letting them move anything. A pile of one-star reviews about a refund policy is a pricing frustration, not bad support. Where aggregates matter to a score I say so in the review itself and date them, since they shift constantly.

What a score covers

Every score on this site is for a plan one person buys for themselves.

Tools with a free plan and a paid plan

A free plan and a paid plan are two different products, and they don’t deserve one blended number. Averaging them would hide the thing you’re actually trying to find out.

So when a tool has both, I score both. Each tier gets its own Ease of use, Value, Features, and Support, judged on what that tier actually gives you rather than copied across from the other one.

What you see on the page is one score, for the plan the review is about, and a plain sentence whenever the other plan lands somewhere different. Something like: this covers the paid plans; the free plan scores 3.5, because it won’t let you download your own work.

I don’t print two full score tables. One review should give you one answer, and a second set of numbers next to the first mostly raises the question of which one is real. The sentence tells you what you need without the clutter.

Tiers diverge more often than they match. Support is a common one: a tool might give free users a contact form and paid users live chat, which is a real gap in what you get when something breaks. So is a free plan that won’t let you export your own work, or one that stamps a watermark on it. Those are exactly the differences a single averaged score would bury, which is why I score them apart even when only one number ends up on the page.

Where a review scores only one tier, it’s because the review is about that tier. A post titled “Canva Pro review” is answering whether the upgrade is worth paying for, not rating the free plan alongside it. Either way, every feature in every review is labeled with the plan it’s on, so you always know which side of the paywall something sits.

What the colored labels mean

Every review marks which plan a score or a feature belongs to with a small colored label, so you can see which side of the paywall you’re on without reading the small print. There are four, and each one always means the same thing.

Free. The free plan. Anything with a green label works without paying, and a tool reviewed on its free plan carries it on its score card. Next to a feature it can also say “On Free too”, “Free, with limits” when the free plan caps it, or “Every plan” when something applies to every plan alike.

Paid. The tool’s main paid plan, the one most people upgrade to. In a review it carries the tool’s own name for that plan, like PRO on Canva or PREMIUM on Adobe Express, so it matches what you’ll see when you go to buy.

A higher paid plan. A step above the main paid plan, like a Plus or Ultra plan.

The top plan. The highest plan a tool sells to one person.

A feature that’s partly free gets two labels, one for the part you get free and one for the part you pay for. A score card carries one label for each plan its score covers: Adobe Express shows FREE and PREMIUM, because both of its plans land on the same score.

Does a paid plan get judged more harshly? No, and not more gently either. Every plan is scored on what it actually gives you. The one difference in how a plan is judged is Value, because Value is about what you get for the money. For a paid plan, that means what the upgrade adds on top of anything that was already free, weighed against what rival tools charge. That’s why, wherever a tool has a free plan, I test it first, even when the review is about the paid one.

Team and business plans are not scored

Most of these tools also sell team and business plans: an admin console, invited members with roles, shared brand controls, approval workflows, and a different price per seat. Canva Business and Adobe Express for teams are the two you’ll meet most often.

Those are different products, and I don’t test or score them. A team plan changes what support you get, what storage you get, and often what the tool costs per person, so scoring it against an individual plan would tell you nothing useful.

If you’re buying for more than yourself, treat my scores as the floor and check the team plan separately. I say so in each review rather than leaving you to work it out.

What earns a place in a roundup

A “best free” roundup has one bar: could a freelancer finish and publish real work on that free plan? A tool that looks good and can’t deliver a usable file doesn’t make the list, however well known it is.

That’s why my roundups are usually shorter than other people’s. I’d rather name what didn’t qualify and why than pad a list to a round number. The exclusions are often the most useful part of the page.

The smaller rules

Prices

All prices are in US dollars, taken from each vendor’s own pricing page on a US connection, and never converted. Prices vary by region, so what you see may differ from what I list. I date every price check and re-verify on a schedule, because pricing pages change more often than features do.

How-to guides

My step-by-step how-to guides don’t carry this four-part score, because there’s no Ease of use or Value to weigh when I’m just showing you how to do one task. Instead, those guides include a simple difficulty rating from 1 (very easy) to 5 (very hard), based on my own run-through.

Why you can trust the number

Once affiliate links go live on Clear-Cut Reviews, a commission still won’t change a score. A tool with a higher commission doesn’t get a higher number, and I’ll happily rank a tool that pays me nothing above one that pays me well if that’s where the testing lands. The scoring method on this page is applied the same way to every tool, every time. For how affiliate links will work here, see the Affiliate Disclosure page.