User:Rw/Proposal:Welcome Working Group archive

From OpenStreetMap Wiki
Jump to navigation Jump to search


The Welcome Working Group (interim name, please improve. see below) is formed with the intent of getting to know mappers better, to understand the motivations and de-motivating factors that effect mappers, and to allow more mappers to feel more welcome.

  • To study and understand mappers and mapping behaviour.
  • To research, discover and disseminate best practices relating to OpenStreetMap data contributors.
  • To improve OpenStreetMap through better understanding of data contributors.


The Welcome Working Group will provide to the community, information on mapper motivation. For example, if supported by experimental data WWG will report, "New mappers are more likely to continue mapping once per week, if greeted by a short message from a local mapper within a month of their first edit." And the numbersfrom the experiemnt to back up that claim, with any interesting exceptions noted.


A proposed Working Group should be supported by at least three OSMF members and at least one board member. Add yourself to the following lists if you support the formation of this Working Group and will participate in it. Please indicate with your name in which staff positions / tasks you intend to participate.

Board members supporting proposal (at least four please.)

OSMF members supporting proposal (at least ten please.)

OSM contributors at large who support this proposal (Speak up!)


Observers of the OpenStreetMap community have noted that many OSM accounts are created, that never contribute data to the project. A smaller, and significant number of accounts contribute data to OpenStreetMap only once. It is generally accepted that these "never-mapped" and "one-and-done" accounts are a type of lost opportunity for OpenStreetMap, based on two simple and related assumptions:

  • "having more mappers is better for OSM"
  • "a mapper who returns to map again is better for OSM than a mapper who maps once"

The work of Haklay suggests that "more mappers in a particular area is better for data quality". (link to Muki's presentation at Amsterdam, please)

The Welcome Working Group will confirm or disprove hypotheses relating to mapper activities through research and measurement. Beginning with:

A first hypothesis "Early contact increases mapper engagement."

In an effort to reduce the number of "one-and-done" mappers who contribute once, then never return, the Welcome Working Group will contact new mappers shortly after their first data contribution.

The goal is to determine the best practices for communicating with new mappers.

Middle distance and future goals

There are big questions on the horizon. We have to start smaller than this.

  • How good is a particular edit? Is one address better than one hospital?
  • How can we reliably and automatically distinguish between editing and vandalism?
  • What is the best edit? 1000 nodes from aerial imagery? 100 nodes from GPX? One business from dead-reckoning?
  • How do mappers with local interests compare with mappers with global interests? What is the ideal combination of these for the best data?
  • How helpful are QA renderings? What are the best practices?
  • How helpful are special interest renderings? What are the best practices?


It is expected that the Welcome Working Group will continue until mapper motivation is fully understood. :-) Initial experiments should approach simpler questions and answers. Future work may build upon earlier work.


To be determined.

A method is envisioned where by a suite of welcoming messages are crafted, in several languages. Welcoming messages are sent, in the preferred language, to a portion of new accounts. Contributions by those accounts, over time, will be compared with control group accounts which are not contacted.

The suite of messages may include:

  • Longer vs. shorter messages
  • messages attributed to "Welcome Group" vs. from a specific mapper
  • messages that are more or less specific. ("I see that you started mapping by adding a restaurant on Pine Street." vs. "Congratulations on making your first edit." )
  • messages that offer guidance. ("You've added a restaurant. Did you know that you can include an address as well?")
  • messages that suggest "social tasks". ("You might enjoy meeting nearby mappers. See who maps near you by setting a home location on your profile page")

The study may research mapper responses based on:

  • time since their first contribution. Are greeting emails more- or less-effective on mappers who are already idle for 3 months?
  • Classification of first object mapped. Are mappers of highways more- or less-likely to return than mappers of restaurants?

Interim name: Welcome Working Group

"Welcome Working Group" might not be the best name for this group. Suggest something better. How about some acronym that implies "welcome" and expands to decribe best practices, mappers, community and science.


Financial budget - Zero. No expenditures are anticipated in the first year of operations.

Volunteer budget - Priceless. All activities are expected to be performed by volunteers.


This working group is likely to have a core group responsible for tasks during the planning stages for each experiment. Those tasks will include, at least the following tasks. Volunteers with specific expertise are warmly welcomed:

  • writing code. Specific software tools to ease recording of contact variables (long or short email? After three days, or three months?) and measuring / recording responses.
  • message translations. It is important that messages to mappers consider language and cultural concerns.
  • scientific merit. Is the experiment valid, or full of unintended bias?
  • chair. The group will require a chair person to attend and keep meetings or projects moving. Chair may also be required to report to MT meetings.
  • board overseer. A board member is expected to participate.


A few hours per week. It is expected that Working Group members will attend a weekly IRC meeting and or telephone conference during the initial stages of creating an experiment. As much as possible, activities will be coordinated by group email. Once an experiment has begun, it is likely that ongoing maintenance will drop to about one hour per week.

No formal process

At this time there is no formal process to propose a new working group, a proposed process is found here. This proposal attempts to follow the informal guidelines as a way to improve this proposal and that one. :-)