Campaign Goals
Marketers could track only two goals per campaign, and the goals they could set were too broad to answer the question they actually had. I led the research that redefined the problem — and falsified my own assumption on the way.
Goals were how marketers at Swrve attributed business value to marketing effort. If you sent a push notification, goals were the mechanism that told you whether it moved sales of the thing you were pushing.
Three complaints had built up around them: goals were too broad to be useful, you could only track two per campaign, and the reporting on them fell short of what customers expected. Engineering had already built a bespoke version for one large customer — it worked, and it could never be rolled out to anyone else.
What I did
I started with the stakeholders — PM, engineering lead, CSM leads — to get the business, technical and customer views on the table, and wrote it all into a design doc the team could argue with.
Then I went looking for evidence, because none of the three complaints had any. I read back through support tickets, queried the campaign database directly with SQL, ran a survey that doubled as an interview screener, and followed up with customers in depth.
What I took from it
Four years on a product gives you strong instincts and no way to tell which of them are wrong.
I went in confident that custom goals were a niche feature used by a handful of sophisticated customers. The database said the majority of campaigns were already using them, and that some customers had accumulated goals by the dozen — one of them well into three figures. Nobody accumulates that many goals because a feature is working. They do it to route around the absence of one.
That single query reframed the project from a refinement into a structural problem, and the definition it produced was what I handed to product.