01. Define the problem and the boundaries.
We start with your product, audience, priority market and business objective. Then we agree what the engagement will produce. Stores, languages, deliverables and review responsibilities are made explicit so a focused project does not quietly turn into an undefined programme.
The scope and pricing guide explains the inputs. The initial brief can be short, provided it identifies the actual decision your team needs to make.
02. Establish what we know.
We review the listing and the evidence available for the question. Public material can reveal messaging problems. Account reports may clarify acquisition patterns. A release or campaign timeline helps explain changes that a listing-only review would miss. Missing data stays visible as a limitation.
When diagnosis is the main deliverable, this becomes an ASO audit. We avoid converting assumptions into facts simply because a chart looks persuasive.
03. Research and recommend.
Research connects user intent, relevant alternatives and product strengths. Recommendations explain the problem, proposed action, confidence and dependencies. Metadata, screenshot messaging and experiments should tell the same story while retaining their different roles.
For example, a keyword cluster can inform the headline promise, while the screenshot brief supplies the product evidence behind it.
04. Make implementation possible.
The handoff defines what changes, who prepares it, who approves it and who publishes it. We record the target store and market, the relevant version and any dependencies on product releases. Design production or translation only appears as an included responsibility when explicitly agreed.
Your team should be able to work from the recommendations without guessing what the slide meant. Questions and review rounds follow the engagement scope.
05. Evaluate and decide again.
Before launch, agree the success measure and observation conditions. After implementation, examine the available evidence alongside traffic mix, product changes and other confounders. A result can support a rollout, justify another test, or remain inconclusive. Each outcome should lead to a considered next step.
The conversion service explains this evaluation in more detail. Follow-up analysis is scoped rather than assumed to continue indefinitely.