Your conversion rate is 2.5%. Great. What does that actually mean?
It means nothing. It’s an average of a 6% rate from branded search, a 3.8% from desktop returning visitors, a 0.4% from cold mobile social traffic, and everything in between. Averaged into one number that tells you precisely nothing about what’s working and what isn’t.
I’ve audited hundreds of accounts. The single biggest unlock — before any A/B test, before any redesign, before any new landing page — is segmentation.
The segments that matter
Source × Device × Intent. That’s the matrix. A mobile user from a Facebook ad has completely different behaviour, expectations, and conversion likelihood than a desktop user searching your brand name. Treating them as the same audience with the same conversion rate is why your “optimisations” aren’t moving the needle.
In our Knowledge Bank, we have 2,000+ experiments. The ones that consistently produce the biggest uplifts aren’t site-wide redesigns. They’re targeted interventions on specific segments — mobile checkout for paid traffic, product page layout for organic desktop, form simplification for returning visitors.
What to do about it
Open GA4. Create segments for your top 5 traffic source × device combinations. Look at conversion rate, bounce rate, and revenue per session for each. You’ll find that one or two segments are dramatically underperforming — and that’s where the money is.
The fix is almost never a site-wide change. It’s a targeted test on the experience that specific segment sees. We’ve seen this approach deliver 2-3x the uplift of untargeted testing, consistently, across 20+ client programmes.
The takeaway
Stop reporting one conversion rate. Start reporting five. The gaps between them are where your growth is hiding.