Most landing page feedback is decoration. Somebody looks at a page, says the headline could be punchier and the button should be a stronger colour, and everyone nods. Nothing measurable changes, because nothing measurable was examined.
An audit is different. It is a fixed sequence, run in the same order every time, that ends with a list of changes ranked by what they are likely to be worth. Thirty minutes is enough. Set a timer, because the discipline is the point.
Minutes one to five: read the promise back
Open the page and, without scrolling, write down in your own words what is being offered, to whom, and what happens if you act. If you cannot do that in one sentence, neither can a visitor who arrived from an advertisement and is deciding whether to stay.
Then check the promise against the business. Plenty of pages describe the product accurately and still fail, because the thing they promise is not the thing the buyer is trying to achieve. A course page that promises a syllabus is answering a different question from the one a career changer is asking.
Minutes six to ten: check the match with the traffic
A landing page is never good or bad on its own. It is good or bad for the traffic pointed at it.
Take the actual advertisement, the search query or the email that sends people here. Read it, then read the top of the page. The wording should feel continuous. If the ad says “part time digital marketing course in Dhaka” and the page opens with a paragraph about the future of work, you have found the leak, and no button colour will close it.
This is the check that most often produces the largest change, and it is the one most audits skip because it requires leaving the page to do it.
Minutes eleven to sixteen: read it on a phone, on mobile data
Not the responsive preview in your browser. A real phone, on a real connection, ideally not the fast one in your office.
You are checking three things:
- What loads first. If the first thing to appear is a cookie banner and a chat widget, that is what your visitors are paying attention to.
- How far the first action is. Count the scrolls to reach the form or the button. Then count how many of those scrolls were content the visitor needed.
- What breaks. Text that touches the edge, a table you can only read sideways, a sticky bar that eats the submit button. These are common, and they are invisible on a desktop.
Minutes seventeen to twenty: find the one action, then count the competition
A landing page should have one job. Name it out loud.
Then count everything else on the page that can be clicked and leads somewhere other than that job. Navigation links back to the main site. Social icons. A second offer in the footer. Every one of them is an exit that you built and paid for.
You are not counting them to remove all of them. You are counting them so the removal becomes a decision somebody has to defend rather than a default nobody noticed.
Minutes twenty-one to twenty-six: use the form, do not look at it
Fill it in and submit it. Properly. This single step finds more broken conversions than any tool, and almost nobody does it before writing an audit.
While you are there, check the things that quietly cost you completions:
- fields that are required but not marked, so the first the visitor knows is a red error
- a phone field that rejects a valid local format
- error messages that appear above the fold while the error is below it
- a submit button that gives no sign it has been pressed on a slow connection
- no confirmation of what happens next, or when
Then confirm that the submission actually arrived where it is meant to go. A form that reports success and delivers nothing is not a conversion problem, it is a revenue problem, and it can run for months without anyone noticing.
Minutes twenty-seven to thirty: check what happens after the click
The page does not end at the button. Look at the thank you page, the confirmation email and the time to first human reply.
An enquiry that gets answered in an hour and an enquiry that gets answered in three days are two different products, and the page is making a promise about which one you are. If the site says a reply within one working day, somebody has to own that.
Write it up as decisions, not observations
An audit that ends in a list of things you noticed is a document nobody acts on. Turn each finding into a line with four parts: what is wrong, why it costs something, what to change it to, and how you will know if it worked.
The phone field rejects numbers entered with a leading zero, so a share of mobile visitors cannot submit at all. Change the validation to accept local and international formats. Watch form starts against form completions for two weeks.
Three findings written like that are worth more than twenty bullet points, and they are also the version you can put in a portfolio. Anyone can list problems. The evidence that you are employable is the column that says what you would change and how you would check.