Free landing page grader

This tool grades the page that has to convince someone who just arrived: promise, focus, proof, form and speed. If what you want to grade is the payment flow of someone who already decided to buy, the right page is the checkout friction score calculator, which runs a different model with 11 different items. The two scores are independent on purpose: you can have a landing page at 90 and a checkout at 45, and that kind of finding is exactly what stops you from optimising the wrong thing.

How to use it

1. Open your landing page on a phone, in a private window, coming from the ad or email that actually sends traffic. Grading it on your desktop, already logged in, hides half the problems.
2. Count the different actions the page asks for. The same primary button repeated three times counts as one; newsletter signup, chat and an ebook download count as three.
3. Count the form fields. If there is no form, leave zero. A prefilled field the visitor can edit counts; a hidden field does not.
4. Measure the time until the page is usable on mobile. Any public speed test does the job, and day to day the stopwatch on your own phone over mobile data already gives you the order of magnitude.
5. Tick the best practices the page already follows. When in doubt, leave it unticked: an optimistic diagnosis never fixed a page. Then fill in visitors, current rate and value per conversion to turn the gap into money.

How it works: the model behind the score

The score is the sum of 13 weighted items that close at exactly 100 points. Four are numeric and use a linear ramp, nine are yes or no and pay the full weight or zero. No black box: if you disagree with a weight, you know which one to change and by how much.

The weights: promise above the fold 12, match with the click source 9, form fields 10, mobile speed 10, button visible without scrolling 8, one single action 8, proof 8, usable mobile layout 8, objections handled 6, no escape links 6, product shown in use 5, button copy 5, risk reducer 5. The order is not a matter of taste: a confusing promise and a mismatch with the ad show up as the most frequent causes of pages that do not convert, and they kill before any argument is made. Proof, objections and risk reducers weigh less because they only work once the visitor understands what you sell.

The four numeric items use a ramp instead of a threshold on purpose. Going from 9 fields to 7 deserves to move the score, otherwise the tool only rewards people who finish the job and gives nobody a reason to start.

Worked example (reproduces the default output)

With the values that come prefilled, describing a very common lead capture page: 3 actions asked, 7 fields, 3.5 seconds to load on mobile, 2 proof elements, a clear promise at the top, a button visible without scrolling, a decent mobile layout and a product visual, but no match with the ad, no objections section, a full menu at the top, a button that says "Submit" and no risk reducer.

• Actions: 8 × (5 − 3) ÷ 4 = 8 × 0.5 = 4.0 out of 8.
• Fields: 10 × (11 − 7) ÷ 8 = 10 × 0.5 = 5.0 out of 10.
• Speed: 10 × (5 − 3.5) ÷ 3 = 10 × 0.5 = 5.0 out of 10.
• Proof: 8 × 2 ÷ 4 = 8 × 0.5 = 4.0 out of 8.
• Full marks: promise 12, visible button 8, mobile 8, visual 5. Zeroed: match 0 of 9, objections 0 of 6, escape links 0 of 6, button copy 0 of 5, risk reducer 0 of 5.
• Score: 12 + 0 + 8 + 4 + 5 + 5 + 4 + 8 + 0 + 0 + 5 + 0 + 0 = 51 out of 100, band D, the page works against you.
• Potential lift: 30 × 49 ÷ 100 = +14.7% conversion.
• With 8,000 visitors per month at 2.4%: 192 conversions per month. 192 × 14.7% = 28.2 extra conversions per month; 28.2 × $ 320 = $ 9,032 per month, or $ 108,380 per year.

The fix queue comes out like this: match the headline to the ad (+9), handle objections (+6), remove the escape links (+6), shorten the form (+5), speed the page up (+5), rewrite the button copy (+5), add a risk reducer (+5), cut the number of actions (+4), add proof (+4). Those nine add up to exactly the 49 points missing. That is what the tool shows when the page opens, and it is the same reasoning you repeat with your own numbers.

How to read it, and where the score misleads

The score is not a verdict on your business, it is a map of what the model can see. It ignores things that sometimes matter more than everything listed here: a weak offer, a price out of line with the market, the wrong audience in the ad, a product nobody wants. A page scoring 95 selling something the market does not want still will not convert, and that is not a landing page problem.

The second trap is self deception while answering. Almost everyone ticks "clear promise" and "usable on mobile" by reflex, because they wrote the page and understand every word. The honest test is showing the first screen to an outsider for five seconds and asking what the company sells and who it is for. If they hesitate, untick it. The cost of lying here is attacking the wrong item for an entire quarter.

The third is confusing potential with a promise. The estimated lift assumes fixing everything delivers everything, which rarely happens: some of the people who do not convert never would have in that session, they were browsing. Use the number to size the effort, and compare your current rate against the conversion rate benchmarks before deciding the page is your bottleneck.

Finally, the model treats the items as independent, and they are not. Shortening the form and rewriting the button touch the same section, so the gains do not add up cleanly. One more reason to fix one item at a time and measure each one.

From the score to the test

A checklist is not evidence. Every item in your queue is a hypothesis about the behaviour of the people who visit the page, and the only way to know whether you got it right is to measure it with an A/B test. The short path is this:

• Take the first item in the queue and turn it into a testable hypothesis in the hypothesis generator, in the if, then, because, measured by format.
• Size the experiment in the sample size calculator before you switch anything on. Campaign landing pages usually get less traffic than people assume, and the maths drives the duration.
• Confirm the result in the significance calculator before declaring a winner, and see what the change is worth in money in the conversion rate impact calculator.
• If the gap between phone and desktop is large, measure it in the mobile vs desktop conversion gap calculator before touching the copy: sometimes the problem is not the words, it is the screen.
• Regrade the page after each fix ships. The gap shrinking is the progress metric for the whole project.

The full method is in the conversion rate optimization guide, and the execution steps are in how to run an A/B test. To write the hypothesis properly, read how to write an A/B test hypothesis.

FAQ

How does a landing page grader work?

It turns a heuristic checklist into a number. You answer 13 questions about your page, each answer is worth a published weight and the sum gives a score from 0 to 100. The value is not the score itself, it is the queue: because every item carries a weight, the points you lost become an honest priority order, which is the real question for anyone with a weak page and little time. No grader replaces an A/B test, it only decides what to test first.

Does the tool crawl my URL automatically?

No, and that is a design choice. A crawler can measure page weight and count fields, but it cannot tell whether the promise is clear to someone who has never heard of you, or whether the headline matches the ad that brought the visitor in. Those two items are worth 21 of the 100 points and only someone who knows the campaign can answer them. Two minutes of your time buys a far better diagnosis. Nothing you type leaves your browser: this page makes no network calls.

What is a good landing page score?

From 85 to 100 the page is sharp, 70 to 84 is good with tweaks, 55 to 69 means you are losing people to sloppiness, 40 to 54 means the page works against you and below 40 it is worth rebuilding before you buy any traffic. Most pages that have never been reviewed land between 45 and 65, almost always for the same reasons: a generic promise at the top, a form that is too long and no specific proof.

Why does matching the ad carry so much weight?

Because it is the most expensive mistake and the cheapest to fix. Someone clicked a specific sentence, landed on a page that says something else and concluded within two seconds that they were in the wrong place. You paid for the click and lost before making a single argument. Fixing it means rewriting one headline. That is why it is worth 9 points and usually tops the fix queue for anyone running paid traffic.

Is the potential conversion lift guaranteed?

No. It scales a conservative 30% ceiling by the size of your points gap, so it is an order of magnitude for deciding where to invest, not a forecast. Two pages with the same score can react very differently to the same fixes, because audience, offer and price are not in the model. The real number for any change only shows up when you measure it with an A/B test on your own page.

Can I grade a landing page with no form?

Yes. If the page has no form, put zero in that field and the item pays the full 10 points, because there is no filling friction there. The other 12 items still apply: a sales page with a buy button, a booking page and a download page all go through the same test of promise, proof, focus and speed.

Keep going

With the score and the queue in hand, the next step is to stop arguing and start measuring. Check where your current rate stands in the conversion rate calculator, grade the payment flow in the checkout friction score calculator and read the method in the conversion rate optimization guide.