Premise App Patterns: Morning larks and creatures of habit

By Tracey Li, Data Scientist

Each day we catch snapshots of people’s lives from all corners of the earth via the photos our contributors submit. But, it’s not just visual images that tell us about their everyday lives — patterns of app usage can tell us so much rich information about how daily routines differ around the world.

As an example, we took a look at which hours of the day contributors from four cities on four continents incorporate their work for Premise into their daily lives. The cities we chose: Ankara, Turkey; Mwanza, Tanzania; Rio de Janeiro, Brazil; and Manila, Philippines.

Here’s what we found:

Activity levels for contributors by city for food-related research tasks.

The highlights:

  • Activity peaks in all cities during the early evening, most likely corresponding to the end of the work day. While this is expected, it’s surprising to see how well-synchronized this behavior is between such diverse places.
  • In contrast, what distinguishes the cities is the activity level earlier in the day. In Mwanza, there is a notable peak at midday which is similar, but less pronounced, in Ankara. Rio has the flattest distribution. Manila is notable for a few early risers collecting data at times when most people are asleep. It’s likely that a combination of working routine and store opening hours influence this behavior.

In the same set of cities, we also looked at the frequency of the number of days that passed in between our contributors’ visits to the same store.

Days between contributors’ consecutive store visits.

By plotting the time intervals between each contributor visiting a specific store, and the frequency of these visits, we can see the extent to which this behavior is dominated by weekly or daily patterns.

  • In Ankara and Manila, people’s routines are dominated by a seven-day rhythm;
  • In Rio, a two- or three-week cadence is more common than weekly;
  • In Mwanza, a daily routine is overwhelmingly preferred.

This behavior is likely influenced by locally available services and infrastructure as well as cultural norms. For example, contributors in Ankara may drive to the supermarket every week versus contributors in Mwanza, who visit their local market daily.

This data is valuable because it enables us to continually refine and tailor the app for our users, and to iterate on our survey designs to yield the highest quality data possible for our customers. Perhaps most important of all, this is another lens through which we can begin to understand the lives of the people who make our work possible.

Show your support

Clapping shows how much you appreciated Premise’s story.