Redesign or CRO? Diagnose the decision before choosing the project

A falling conversion rate creates urgency, but urgency is not a diagnosis. Teams often jump to one of two prescriptions: redesign everything, or run a queue of small experiments. Both can be responsible. Both can also waste a quarter when they are applied to the wrong problem.

The choice becomes clearer when you separate the storefront into four layers: proposition, structure, expression and performance. A redesign can repair a broken proposition hierarchy or merchandising structure. Conversion optimisation is strongest when the foundation is sound and a specific journey can be measured.

Choose the smallest intervention that can responsibly resolve the uncertainty—not the largest project the budget can carry.

First, name the evidence you actually have

Begin with channel mix, device mix, landing-page intent, product engagement, search behaviour, add-to-cart movement and checkout completion. Compare periods carefully: a campaign that brings colder traffic can depress conversion without making the storefront worse. A shift toward mobile can expose interaction debt that desktop averages hid.

• Segment acquisition quality before judging the page.
• Inspect the funnel by device, market and product family.
• Read search terms, support questions and return reasons as customer language.
• Watch representative sessions, but validate anecdotes against aggregate behaviour.

Signals that point toward redesign

Redesign becomes appropriate when customers cannot form a reliable mental model of the offer. Common signals include overlapping categories, inconsistent product information, a visual system that cannot express hierarchy, templates that fight the merchandising model, or an operating team that needs developers for ordinary campaign changes.

Signals that point toward CRO

Optimisation is the better first move when the brand and catalogue are understandable, the technical baseline is healthy, and the uncertainty is local: a product-page proof gap, poor filter use, unclear delivery information, weak bundle discovery or a checkout handoff that surprises shoppers.

A responsible sequence

• Instrument the journey and establish a trustworthy baseline.
• Run a short diagnostic across proposition, structure, expression and performance.
• Fix severe accessibility, speed and tracking defects before interpreting experiments.
• Prioritise one customer decision at a time.
• Escalate to redesign only when the evidence shows the system itself is the constraint.

The outcome may be a focused CRO programme, a redesign with a clear measurement plan, or a hybrid in which critical templates are rebuilt while stable parts remain. The right answer is not ideological. It is the one that turns uncertainty into a testable commercial decision.