There's a version of CRO that looks impressive in a weekly report and quietly destroys a SaaS business. It goes like this: the marketing team runs a series of tests on the homepage, improves signup rate by a meaningful amount, and celebrates. Six months later, revenue hasn't moved. Trial-to-paid hasn't budged. Retention is flat. Nobody can explain why.
The explanation is usually that the optimisation attracted more of the wrong people. Easier signups often mean lower-intent signups. And lower-intent signups don't activate, don't convert, and don't stay.
The funnel you should actually be measuring
- Visit → Trial signup (the only step most teams obsess over)
- Signup → Activation (your product's 'aha moment', often unmeasured)
- Activation → Paid conversion (the commercial step that actually generates revenue)
- Paid → 90-day retention (the signal that tells you whether the above is working)
The highest-leverage optimisation opportunities are almost always at steps 2 and 3. They're also the hardest to improve, which is partly why most teams ignore them in favour of the more tractable problem of getting more people in the door.
Activation rate
Define your activation event first. Not the first login, not the first feature click. The specific action that, in your data, correlates with long-term retention. For a project management tool it might be creating a project and inviting a teammate. For an analytics platform it might be viewing a dashboard populated with real data.
A common pattern: activation rate is 30%, but for users who activate within day one, trial-to-paid is 60%. For users who don't activate until day five, it's 8%. The first 24 hours are doing more work than the rest of the trial combined.
Time to Value
Every minute between signup and first meaningful experience of your product is a minute in which a competitor's trial email is landing in the same inbox. Time to Value (TTV) is the metric that captures this: the median time from signup to activation event. Use median, not mean. Trial TTV data is heavily skewed by power users and a long tail of people who never activate.
Trial-to-paid conversion rate, segmented properly
An aggregate trial-to-paid rate can hide meaningful differences between cohorts. For example, one acquisition channel may be converting far better than the others. Segment by channel, plan, and company size before choosing an optimisation target, then investigate what distinguishes the cohort with the strongest qualified outcomes.
PQL rate
Product Qualified Leads are trial users who've hit the in-product signals that predict conversion. Your PQL rate is the percentage of trial users reaching those thresholds - and indirectly, a measure of whether your landing page is attracting the right audience.
Where to focus
- Landing page → signup: headline clarity, CTA specificity, social proof, pricing transparency
- Signup flow: fewer fields, faster progress, clear indication of what comes next
- Onboarding: guided activation toward your defined aha moment, not a feature tour
- Trial communication: email sequencing timed to activation signals, not calendar days
- Pricing page: plan differentiation, annual framing, objection handling in the FAQ
The mistake most SaaS teams make isn't ignoring CRO. It's applying CRO exclusively to the marketing site while leaving the in-product experience - where high-intent users are already trying to convert themselves - entirely unoptimised.