Website Conversion Rate Optimisation Checklist

Ordered by impact for a service business website. eCommerce funnels have a different order.

0. Delivery — do this first

• [ ] Submit every form yourself, from a real address
• [ ] Confirm the email arrives
• [ ] Check the spam folder
• [ ] Confirm the sending domain has SPF and DKIM records
• [ ] Confirm submissions are stored somewhere durable, not just emailed
• [ ] Confirm a delivery failure would be visible to someone

If the form is not delivering, nothing below this line matters. Test it monthly.

1. Clarity

• [ ] What you do is clear within the first screen
• [ ] Who it is for is stated explicitly
• [ ] The headline says something specific, not a category
• [ ] One primary action per page
• [ ] CTA wording matches how ready the reader is
• [ ] No competing calls to action

2. Friction

• [ ] Form has the minimum viable fields
• [ ] Every field justified by how it changes your handling
• [ ] Phone number optional, not required
• [ ] Inline validation with errors that say what to do
• [ ] Errors surface visibly and do not disappear
• [ ] An email address offered as an alternative to the form
• [ ] Works on a phone, tested on a real one

3. Information

• [ ] Price or price band published
• [ ] What is included, specifically
• [ ] Process and rough timeline
• [ ] What happens after they enquire, and when
• [ ] FAQ answering what people actually ask

4. Trust

• [ ] Proof is real and verifiable
• [ ] No invented testimonials, ratings or statistics
• [ ] No star ratings without reviews visible on the page
• [ ] Real contact details, including a location
• [ ] Claims are specific enough to be checkable
• [ ] Nothing you could not defend if asked

That fourth item is not moralising. A visitor who suspects one claim discounts every other claim on the page, which makes fabricated proof actively negative rather than neutral.

5. Speed

• [ ] Loads acceptably on a mid-range phone
• [ ] Core Web Vitals passing on field data
• [ ] Nothing jumping around while loading

6. Measurement

• [ ] A conversion event defined in analytics, not just page views
• [ ] Form submissions tracked as that event
• [ ] Form errors tracked too, so silent breakage is visible
• [ ] Traffic source attached to conversions
• [ ] A baseline recorded before you change anything

Without a defined conversion event, every later decision is argued from opinion. This site ran three months with no analytics at all, which made every question about performance unanswerable.

7. Testing

• [ ] Change one thing at a time
• [ ] Run long enough to mean something
• [ ] Do not conclude from twelve conversions
• [ ] Keep a record of what was changed and when

Most small business sites do not have the traffic for meaningful A/B testing. That is fine — sequential changes with a recorded baseline are a legitimate method at low volume, as long as you are honest that the evidence is weaker.

What to skip

• Exit-intent popups, unless you genuinely have something worth interrupting for
• Countdown timers on things that are not actually time-limited
• Live chat with nobody answering it
• Anything that makes the page harder to read

Why website visitors are not turning into leads covers the diagnosis behind this list.