Naming events across journey stages without muddling intent
How West Midlands product teams keep stage names, action names, and property keys distinct when mapping application journeys.
When several teams touch the same application, event names tend to accumulate like spare keys in a drawer. Marketing asks for “lead_created”, product prefers “signup_complete”, and engineering ships “user_registered”. Each label can be true in isolation and still break a journey map.
Start by separating stage labels from action labels. Stages describe where someone stands in the journey — awareness, account setup, first meaningful use, renewal. Actions describe what happened at that stage — viewed, submitted, confirmed, abandoned. Mixing the two in a single string is how “first_login_success_v2” appears in exports nobody trusts.
Agree a short glossary before the first instrumentation sprint. Include the application’s own words for accounts, workspaces, and outcomes. If your product calls a purchase a “booking”, the event catalogue should say booking, not order, unless you deliberately translate for finance reports later.
Properties carry the detail that names should not. Channel, plan tier, and entry point belong as properties attached to a stable event, not as suffixes bolted onto the event name. That keeps longitudinal reports readable when marketing experiments change weekly.
In mapping sessions we ask one practical question for every candidate event: which journey stage does this belong to, and which decision would change if this signal arrived tomorrow? If neither answer is clear, the event waits. A thinner catalogue that product and analytics can recite from memory outperforms a crowded list that only exists in a spreadsheet.