How to Start Without Rebuilding Everything

The right first step is not a rebuild. It is not a new platform, a new AI tool, a new CDP or a six-month change programme. The right first step is a focused audit or pilot around one category, one product group or one commercial problem.

Why this matters now

Retailers already have enough moving parts. Trying to fix product data, content, feeds, tracking, customer insight, reporting and operating ownership across the whole business at once will usually create more noise than progress.

A focused pilot gives the team a safer way to learn. It shows what data is missing, which systems are difficult, which teams need to be involved and where the commercial value might sit.

What is actually changing

The shift is from broad readiness talk to controlled action. A pilot category can test product truth, feed enhancement, content intelligence, customer insight, signal quality and reporting rhythm without asking the whole business to change at once.

The best pilot is not always the biggest category. It may be a category with high returns, high media spend, weak feed visibility, strong margin potential, complex buying decisions or strategic importance.

What is often misunderstood

The misunderstanding is that starting small means thinking small. It does not. It means creating evidence before scaling. A good pilot can reveal a repeatable operating model for the wider business.

Another misunderstanding is that the pilot should be owned by one person quietly. It needs cross-functional input, but a tight scope.

What retailers should review

• Which category has a clear commercial problem or opportunity?
• What product, customer, order, feed and tracking data is available?
• Which PDPs or products should be reviewed first?
• Which teams need to act on the findings?
• How will success be measured after 30, 60 and 90 days?

What good looks like

Good looks like a pilot with a defined scope, named owners, a short action list and a review rhythm. The team knows what it is testing and what would justify scaling the work.

Outputs are practical: feed rules, PDP requirements, content briefs, tracking fixes, customer insight, reporting changes and agency instructions.

What not to overdo

Do not select a pilot so broad that it becomes a programme. Do not choose a category where no one will act on the findings. Do not measure only traffic or revenue if the problem is customer quality, margin or returns.

The pilot should be narrow enough to finish and useful enough to matter.

Practical next step

Choose one category and define the question. For example: why is paid traffic not converting profitably, why are returns high, why is feed visibility weak, or why are customers not repeating? Then build the audit around that question.