Questions for digging into TM data

[smit1678]

@hotosm/major -- Can you list out some specific initial questions to ask from existing data we have? We will then determine where we can answer these questions. This might come from the TM database, website metrics, or other OSM data we can gather. We don't need to define all the questions up front now but some questions to help get you started.

Initial data analysis listed in resources page

cc @ramyaragupathy

[leihla]

Hey @smit1678 and @ramyaragupathy,

We've looked through the initial data that you've compiled on TM EDA file and we have a few follow up questions and other data points that we would like to ask for.

Mappers Data

1. On the "Contribuition-stats" sheet can we add

• number of users who have contributed to more than 1 project
• number of user who have contributed to more than 5 projects
• number of users who have contributed to more than 10 projects

The goal here is to have a more granular understanding of the mapper persona. Even though 97% of mappers are identified as beginners, can we break them down in smaller groups and identify power users?

2. Create "Active Mapper" sheet Can we create another spreadsheet with the Active Mapper data? For now we could define an active mapper as someone who has completed at least 1 task in the last 30 days. If you think this is too short of a timespan we could give it 60 or 90 days.

We would like to see the following data:

• Current number of active mappers
• Evolution of active mapper numbers in the last 6 months (we would like to understand if evolution is organic or follows a specific pattern as related to specific events/mapathons)
• number of tasks mapped per active mapper
• number of projects contributed per active mapper
• average time spent mapping per task

3. Profile data / Engagement From your analysis, only 2581 users have email completed, and only 618 have verified their email. Is there any way to reach out to all mappers by email or create different segments of users of the 154k registered users?

What we want to understand:

• Why are users not converting from registered user to mapper?
• How does having to use multiple platforms impact conversion rates? (OSM registration, TM project and task selection, Editor for mapping, etc)
• Do new mappers understand the complete mapping process? (Choose project, choose task, choose editor, map, save changes, submit mapped area for validation, receive feedback?)
• Do new mappers have enough information to choose a project?
• What makes users choose a project over another?
• What makes users choose a task over another?
• How do new mappers choose a task? What criteria has the biggest impact?
  • Does priority/sense of urgency has an impact on user choices?
  • Does geographic area has an impact on user choices?
  • Does cause / type of phenomenon has an impact on user choices?
• What are the factors that influence their choices? (location, urgency, cause, etc)

I'll create another comment for the project data questions

[ramyaragupathy]

• number of users who have contributed to more than 1 project
• number of user who have contributed to more than 5 projects
• number of users who have contributed to more than 10 projects

Updated numbers under contribution stats - https://docs.google.com/spreadsheets/d/1QIy__tqdED9VZgeGXWkR5HZdIKHLLTP11lRaRmhCvaU/edit#gid=652552763.

As the number of projects increase, the beginners ratio in contribution goes down.

Num of projects	Beginner contribution
1	99%
2-5	95%
5	86%
6-9	79%
10	66%
>10	36%

[pantierra]

Hi @leihla, here some more data:

• Current number of active mappers (active within January 2019): 2121
• Evolution of active mapper numbers in the last 6 months:
  • active in 12/2018: 2553
  • active in 11/2018: 3963
  • active in 10/2018: 3736
  • active in 9/2018: 3188
  • active in 8/2018: 4287
  • active in 7/2018: 2295
  • active in 6/2018: 3015
  • active in 5/2018: 3245
  • active in 4/2018: 2949
  • active in 3/2018: 3815
  • active in 2/2018: 2017
  • active in 1/2018: 1699

And then I have a couple of questions:

• number of tasks mapped per active mapper: During the month the mapper was active. Or during all times?
• number of projects contributed per active mapper: During the month the mapper was active. Or during all times?

And I have bad news: currently there is no way to get the average time spent mapping per task. Only when it has been locked and marked as mapped. Do you want me to get this out? This data is going to be very weakly related on the time a mapper spends on the task.

From your analysis, only 2581 users have email completed, and only 618 have verified their email. Is there any way to reach out to all mappers by email or create different segments of users of the 154k registered users?

Probably not :( We only have their OSM name and we are probably not allowed to go through OSM to contact the users. What do you mean with segments?

@ramyaragupathy I used this script to generate the stats.

[leihla]

@xamanu and @ramyaragupathy thank you so much for this info!

Just to keep everyone on the loop, for these numbers, we used active mapper as anyone that mapped at least one task in the last 30 days, correct? Could we get the same data if we accounted for 60 days?

Regarding your questions:

1. • Number of tasks mapped per active mapper* I would prefer to focus on the active period, but if we can get both that would be nice.
2. • Number of projects contributed per active mapper* can we get both? Otherwise let's focus on the active period
3. Avg time spent mapping per task - that's bad news indeed, but maybe it's even better to understand on the tasks worked on by active mappers, the median time to mark:
   • from blocked to marked as mapped
   • from marked as mapped to in validation
   • from blocked to validated
4. Email we'll address this issue in the design phase, either collecting from current users and fixing for new users

Thank you!