Should I audit or optimise first?

Audit first. The audit finds the leaks; optimisation fixes them in priority order. Optimising before you audit means polishing the wrong step.

What can’t a conversion audit fix?

A weak offer. It optimises the path to purchase; it can’t validate whether people want what you’re selling. And it’s only as good as your instrumentation.

The Conversion Audit Cheat Sheet

Q: What is a conversion audit?

A systematic walk through every step between first click and completed action, measuring where users drop off so you fix the biggest leak first.

Q: What are the most common conversion killers?

Friction: extra steps that add no value, forced account creation before users see value, slow page loads, and asking for information you don’t need yet.

At eBay we tracked confirmed registered users (CRU) as our conversion. Then we switched to activated confirmed registered users: ACRU. Registered, confirmed, and having bid on, bought or listed an item. It sounds like a small definitional change, but it gave us the right conversion and fixed our affiliate incentives. We stopped paying for signups that never did anything.

That is the first job of a conversion audit: before you fix the flow, make sure you are measuring the right conversion in the first place. After that it is a systematic walk through every step between first click and completed action, measuring where users drop off and fixing the biggest leak first. As I say in the book: mediocre marketing with a great conversion flow is way better than brilliant marketing with a broken one.

🧪

The 6-Step Conversion Audit

Map your full conversion flow. I draw every step from first click to completed action, including confirmation pages, email verifications, and loading states. If it is not on the map, it is not in the audit.
Instrument every transition. Add tracking to each step. You can’t optimise what you can’t see.
Find the largest drop-off. Your biggest leak is usually your highest ROI fix.
Hypothesise friction causes. Too many steps, confusing UI, slow load, or asking too much too early.
A/B test targeted fixes. Isolate tests; multivariate is fine, cross-contamination is not.
Iterate relentlessly. Conversion work is never “done”: every winner eventually degrades.

The audit finds the leaks. What to change, and how to test it, lives on the conversion optimisation sheet.

🔍

Not to Be Confused With

🔍 Audit vs optimisation

The audit finds the leaks; optimisation fixes them in priority order. Don’t optimise before you audit: you’ll polish the wrong step.

📊 Conversion rate vs conversion volume

A higher rate on a shrinking base is a loss disguised as a win. Watch the absolute numbers alongside the percentages.

🧹 Friction removal vs persuasion

Most audit wins come from removing obstacles (fewer fields, faster loads, clearer CTAs), not from cleverer persuasion. Delete before you decorate.

🚫

Conversion Killers (Avoid These)

❌ Extra steps that add no value.

❌ Requiring account creation before users see value.

❌ Slow page loads, especially on mobile.

❌ Asking for information you don’t need yet.

❌ Poor mobile UX that breaks flow.

❌ No clear call to action.

The “Create an Ad” lesson

In France the button read “Create an Ad” rather than “Advertise”, and converted ~40% better. We rolled that wording out globally and kept the uplift. Test your button copy. The basics matter.

🧭

Know Its Limits

My position: define the conversion before you optimise it. An audit finds friction along the path, but it cannot tell you whether the destination is worth reaching. A perfectly smooth flow for a product nobody wants still converts at roughly zero.

And you can’t audit what you haven’t instrumented. At Facebook we once found server-side and client-side counts disagreeing by over 10% because a pre-fetching bug was logging “active” users who never opened the app. An audit built on broken tracking will confidently point at the wrong leak.

The audit is a habit, not a project.

THE CONVERSION AUDIT CHEAT SHEET