Why testing often does not apply
A split test needs enough conversions in each variant to distinguish a real difference from chance. For a site with a modest number of enquiries each month, detecting a small improvement can take many months, by which time the market, the offer and the page have all changed.
Teams in this position often run tests anyway, stop them early and act on results that are indistinguishable from noise. That is worse than not testing, because it produces confident decisions with no basis.
Replace testing volume with research depth
When you cannot let the numbers decide, gather stronger evidence before making a change.
- Watch session recordings on the pages that matter most.
- Read form analytics to see which fields cause abandonment.
- Interview recent customers about what nearly stopped them.
- Ask sales which questions they answer on every first call.
Fix first, then redesign
Some problems do not need a hypothesis. A form that fails on a phone, a page that takes too long to load, a call to action below a wall of text, a price that cannot be found: fix them. No one needs a test to confirm that a broken step loses customers.
After the evident problems are gone, make larger changes instead of small ones. A different headline colour will never be measurable at low volume. A page rebuilt around the objections you heard in interviews might be.
Measure honestly
Compare a full period before the change with a full period after it, long enough to cover your normal weekly and monthly patterns. Note anything else that changed at the same time, such as a campaign, a season or a price.
Be modest about the conclusion. Before-and-after comparison shows direction, not a precise uplift. Look for changes large enough to be obvious, and follow them through to lead quality so that a higher conversion rate is not simply a lower bar.
When testing becomes worthwhile
As traffic and conversions grow, formal experimentation starts to pay. Until then, the discipline is the same one that makes testing useful later: a written reason for every change, and a record of what happened next.
