Treating the keyword export as the strategy.
A list of phrases does not establish which audience matters or whether the product fulfils the implied tasks. Selecting terms solely because a tool shows a large number can create a misleading listing.
Use relevance checks and record why candidates belong. The research should support a decision rather than end with a bigger spreadsheet.
Changing several things without a question.
A broad release can be appropriate, but it answers a broader evaluation question. If copy, order, visuals and traffic all change, do not later credit one caption for the outcome.
Use a hypothesis and a release record. Define the treatment that actually reaches users.
Copying creative without product context.
A competitor’s visual style does not establish why its users choose the app or whether the same promise fits yours. Borrowing the surface can conceal a missing argument.
Use screenshot roles to build evidence around your own supported workflow. Describe public competitor observations without inventing private conversion results.
Reporting a rate without its denominator.
A percentage can look comparable while counting a different event or audience. A changed mix can also alter an aggregate even when segment performance is stable.
Use rate definitions and preserve source context. A confident chart title cannot repair incompatible inputs.
Producing more work than the team can release.
A growing backlog may reflect missing ownership or capacity rather than a shortage of ideas. Agree who implements and verifies before expanding scope.
Use a decision-led roadmap. The corrective habit is to close the loop between evidence, action and review—not to add another checklist nobody can finish.
