Aaptiv had a simple product question hiding inside a messy pile of data. What did a member actually do between opening the app and finishing a class?
I wanted to see the whole path.
The analysis started with membership dates: account creation, trial, first payment, and expiration. Then it moved into activity inside the product.
I looked at monthly visits to the home screen, workout list, workout detail page, and trainer page. For each part, I wanted both the total activity and the number of different days somebody came back.
The last group covered classes. How many did a member take? How many did the member finish? How many separate days included a class?
Those questions turned into seventeen named analyses and a lot of CSV files. Several exports existed twice. There was a folder called Final and another called Updated Queries. Final was already not final.
That was fine. I was trying to understand the behavior, not produce a beautiful folder.
Each step meant something different. Opening the home screen was not choosing a workout. Looking at a workout was not starting it. Starting a class was not finishing it. Taking three classes in one day was also different from coming back on three different days.
The differences mattered because Aaptiv was not a product somebody bought once and put on a shelf. The value came from finding a useful class, doing it, and returning for another one.
I was still new enough to the company that I did not want to rely on a general impression of how people used the app. A member could tell us about one experience. A chart could show a large total. Neither answered every question about the sequence.
So the files kept getting more specific. Some versions used different membership start dates. Some removed rows with no activity. Others joined activity back to the membership timeline so we could compare use during different parts of the relationship.
The output was not a grand product strategy. It was a table of contents pointing to queries and files.
That table was useful. It forced me to name the parts of the product journey in plain language. Home. Workout list. Workout detail. Trainer. Class started. Class finished. Another day.
I still had to decide what to build. The analysis did not do that for me. It gave me a cleaner picture of where people were spending time and which questions needed a better answer.
For the moment, that meant another query, another export, and one more file that probably did not belong in the Final folder.