Getting more people to the page is a solved problem with fifty thousand people selling it. What happens after they arrive is psychology, and that is a different profession. I am an I/O psychologist who also writes the code, so the findings come with measurements and the fixes come with a diff.
Every option you add is a cost paid by the visitor. Beyond a certain number people stop choosing and leave, and they never tell you why.
Each field has a price. Most forms ask for things nobody needs, and the phone number field alone can be the most expensive box on the site.
There are four separate moments of doubt in a purchase and reassurance placed after the moment has passed does nothing at all.
Five things get worse on a phone and only one of them is screen size. Most checkout abandonment is decided here.
Not page weight — the gap between the tap and the first thing the visitor can see. It is felt as doubt about whether the site works.
A silent failure after payment costs you the repeat purchase and produces a support ticket instead. It is also the easiest to miss.
Measure first, decide what it costs, then change one thing at a time.
An audit that begins with an opinion finds evidence for that opinion. So the first pass is the fourteen checks, which produce numbers rather than impressions, plus three tests that need no software and repeatedly find that the most prominent thing on a page is not the thing the page is for: the five-second test, the squint test, and reading the link text on its own.
Twenty problems is not a plan. Findings get ordered by the cost they impose on the person trying to buy, not by how easy they are to fix, and that ordering is usually different from the one a development team would choose. Three things at the top of that list is a plan.
Changing six things and watching the number move tells you nothing about which of the six did it, and one of the six may well have made things worse while the others carried it. If your traffic is large enough to test properly, test. If it is not — and most sites are not — then say so honestly and rely on measured mechanics instead of pretending a sample of forty conversions means something.
Most CRO advice arrives as a list of tactics: move the button, change the colour, add urgency. Some of it works somewhere. None of it tells you why your page in particular is losing the people it is losing, and a tactic borrowed from a site with different visitors and a different offer is a guess wearing a case study.
The alternative is unglamorous and it works: measure first, decide what it costs, then change one thing at a time.
Anybody can produce a list of twenty things wrong with a website. Knowing which three are actually costing you money — and being able to say why in a sentence a non-technical person accepts — is the part that is hard and the part that changes what happens on Monday.
This is where a lot of CRO goes wrong. A test that has not reached significance is not a result, and stopping a test the moment it looks good is how a business convinces itself of something untrue for a year.
| A generic CRO retainer | This | |
|---|---|---|
| Basis | Hourly, ongoing | Fixed scope, fixed price |
| Findings | A list of tactics | Measured, ordered by cost to the reader |
| Who wrote it | An account manager | The person who does the work |
| Implementation | Handed to your developer | I can write it, or hand it over cleanly |
| Dark patterns | Often, quietly | Never, and the reasons are in writing |
| What you keep | A slide deck | A report written to be forwarded |
If you have very little traffic, conversion work is premature: you cannot learn anything from twelve visitors and the money is better spent finding out whether anybody wants the thing at all. If your traffic is large but badly targeted, no amount of page work will fix an audience mismatch, and the audit will say so rather than take the fee quietly.
I would rather tell you that in the first conversation than three months in.
47 full brand psychology audits are published on this site — Netflix, Nike, Tesla, Shein, Careem, HBL, Khaadi, Jazz and more. They are free to read, they are not summaries, and they are the closest thing to watching me work before you decide whether to hire me.
Before you buy anything: run the fourteen checks on your own page and see the numbers for yourself.
Fixed price, fixed scope. The market rate for CRO consulting is $100–300 an hour; these are priced as projects so you know the number before you start.
The written diagnosis
I write the fixes too
For sites with enough traffic to test
Agencies, benchmarks, A/B tests, scarcity and start dates.
An agency is largely selling you traffic, and traffic is a crowded, well-supplied market. This is about what happens after somebody arrives, which is a psychology question. You also talk to the person doing the work rather than to an account manager relaying it.
Anybody who answers that before seeing your site is guessing, and the “average conversion rate” figures published online are averages across wildly different businesses. Your useful comparison is your own page last month, not somebody else's industry benchmark.
When your traffic supports them, yes. When it does not — which is most small and mid-sized sites — I will say so rather than run a test that cannot reach significance and then read meaning into noise.
Yes, partly. The free audit tool runs fourteen measurable checks on any page and gives you the same numbers I would start from. It cannot do the journey, the forms, the ordering, or the writing up.
No. Fake scarcity is now actively enforced against in several markets, and the version that is not fake — telling the truth about genuine limits — works better anyway and does not expire.
Usually within a week. The audit itself takes ten working days for a site and three for a single page.
If the visitors are arriving and the sales are flat, the problem is on the page. Let us find out where.
Your cart is empty
Browse Shop