Map the notification catalogue
We list every live push type, its owner, and whether it is transactional, promotional, or educational. Consent prompts and preference centres are noted with recent change dates.
Our method
Four stages from catalogue to first experiment
Every push notification engagement review follows the same path so findings stay comparable across weeks and stakeholders share a clear sequence.
Read engagement exports
Open rates, tap-through, and opt-outs are examined by daypart and segment. Quiet-hour collisions and stacked sends on the same device are flagged before any copy advice.
Workshop the findings
Two working sessions surface constraints — engineering capacity, vendor limits, brand rules — so the experiment backlog reflects what the team can actually run.
Launch the first experiment
You leave with owners, measures, and a calendar for the first two tests. A follow-up call reads early results and decides whether to promote, discard, or adjust.
What you prepare
Anonymised engagement extracts for at least eight weeks, a sample of recent notification copy, and a contact who can explain how deep links are built. We never need production credentials.