Make sure the numbers add up
I check what your tracking counts, where the task starts and ends, and whether a completed enquiry, booking or order is recorded properly.
Conversion researchStop guessing why
Stop guessing why
people drop off.
People start the enquiry, booking or checkout and never finish. I check your numbers are telling the truth, then find out what’s actually stopping them.
At a glance
- Good for
- You know what you want people to do, and your numbers say too many of them don’t.
- Who does it
- Brian, start to finish
- The fixes
- Built by whoever you choose
- Price
- Agreed in writing before we start
The real question
Where do people give up, and which of your theories about why is right?
01 / Use this when
You know what you want people to do, and your numbers say too many of them don’t.
02 / How I do it
First I check the tracking measures what you think it does. Then I test the competing explanations with a review or research.
Enquiries, bookings and shopsFind the point
Find the point
where customers stop.
An enquiry form people abandon. A booking that never reaches confirmation. A shop checkout with baskets left behind. Start with the task you need customers to finish.
Watch the task, not just the totals
We pick one priority task and agree the method around your question, traffic and access to participants. I review the screens or watch relevant people try the task to find what needs attention.
Know what to tackle first
You get the measurement checks, the findings ranked by priority, what each finding is based on and what to check next. You choose who makes the changes.
Not a slide deckWhat lands
What lands
on your desk.
A check of your tracking and baseline
The likely reasons, with how sure I am about each
What to look into first
How to test the fix before you bet on it
How it runsFrom “we think”
From “we think”
to “we saw it”.
Check the numbers
Make sure the events you count describe the task you care about.
Find the friction
Put the data next to what people actually do. A correlation isn’t a reason.
Plan the test
Decide up front what would prove the fix right, or wrong.
Proof, not promisesSee it for
See it for
yourself.
The security-app case shows why the number you watch matters: app opens couldn’t tell happy customers from lost ones.
Opens it oftenmay be worried or confusedRarely opens itmay still be getting the value
App opens couldn’t tell happy customers from lost ones.
Someone can get the value without opening the app. Someone else opens it again and again because they’re worried. So refunds and renewals counted too.
Before we start
What to bring.
The action that matters, whatever analytics you have (warts and all), the paths people take, recent changes, and access details.
Build your brief ↗