Wikimedia Deutschland/Data Partnerships Model/The Phases of Data Partnerships

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search
Data Partnerships Model Banner.png

 Main page Executive Summary Background Purpose Partners The Phases of Data Partnerships Data Partnerships Scenarios Decision-Making Conclusion 
The Phases of Data Partnerships

Data partnerships typically go through six phases: initial contact, feasibility and profiling, assessment, implementation, conclusion and debrief, and post-project monitoring and support. However, it is worth noting that the boundaries between these phases are not always distinct, and partnerships do not always follow a neat linear trajectory. Likewise, the speed at which a partner progresses through the phases can vary significantly.

1. Initial contact[edit]

The initial contact phase is bidirectional where both parties engage in tentative enquiries about a potential partnership. This can occur digitally, in person or through an intermediary.

This is typically a reactive phase for Wikimedia Deutschland, wherein we respond to incoming requests and inquiries. However, we can also proactively reach out to partners.

a. Reactive contact[edit]

In the case of a reactive contact, a potential partner finds us. This occurs in a variety of situations: at conferences, via word-of-mouth or referrals, through volunteers in the Wikimedia projects or from other partners. We have found it very useful to offer an email address as a point of contact ( and as a place to forward requests which end up in our respective inboxes.

Although feasibility and profiling and assessment follow first contact, portions of these later phases can occur in this stage to help us decide how to proceed.

b. Proactive contact[edit]

In the case of a proactive contact, a potential partner may be completely unaware of the potential offered by Linked Open Data and by tools created by the Wikimedia movement. Partners may show up through our own research.

This phase is currently underdeveloped, and we have not specified many of the necessary prerequisites for this case, which, along with possibly shifting from reactively to proactively initiating partnerships, remains a work in progress on our part. We presume proactive contacts would be guided by both our strategy papers and the 2030 movement strategy. This means that future potential partners we want to seek out can come from the field of science, as explicitly called out in the Wikidata and Wikibase strategies or represent marginalized knowledge as central to the movement strategy. Other fields that have come up in past partnerships and should be considered for further exploration include government data and digital humanities.

A process to develop tools for such a situation would necessarily include setting knowledge area priorities, developing a process to find and assess projects within each priority area and researching tools to seek out projects, as well as a way to assess the projects we find and decide which of them to pursue.

2. Feasibility and profiling[edit]

This phase begins in a conversation with the potential partner; various scenarios and options are presented, and we try to categorize an initial scenario and to come to a decision with the potential partner about the partnership.

In this phase we decide whether we want to proceed with the partner based on a series of questions related to impact. We then assess the partnership according to a number of make-or-break requirements.

a. Impact[edit]

Would they be a low-, middle- or high-impact partner? We assess the level of impact using two criteria:

  1. Organisation size and reach: is this an international, national, or federal organization (high), regional association (middle) or local organization or project (low)?
  2. Importance of the knowledge area: is the organization working in a strategically important field to us (e.g. GLAM, science), or might they help us to close knowledge gaps in that they offer expertise and data for underrepresented knowledge?

b. Make-or-break requirements[edit]

Among the things to look for in a partner is the way they describe their data. Do they have some idea of what structured data means (i.e., machine-readable and available in a non-proprietary open format with documented metadata)? Does their data translate easily into the Wikibase data model? Do they mention structured formats they already use? For instance, someone from the library world might mention RDF, Bibframe and/or UNIMARC. And what about metadata?

All of the foregoing gives us the necessary information to decide how well the potential partner understands the area and whether or not they will require technical assistance. Another possible question to ask during this phase is whether they have a technical team supporting them who would take this on, e.g., an IT department or a systems administrator.

If a partner does not meet our requirements, or we theirs, we look for third parties (particularly other Wikimedia chapters) who can assist. If that is not possible, we do not pursue a formal partnership further, but the partner remains in our pool of contacts. This allows us to follow up with the partner in case circumstances change in the future.

3. Assessment[edit]

In this phase we gather all information necessary to judge the suitability of a partnership, for instance the partner’s requirements and those of various teams inside Wikimedia Deutschland, such as engineering teams. This information gathering process may go through several interactive steps (e.g., workshops), and should ultimately lead to an agreed-upon roadmap and plan outlining how the partnerships project will be carried out.

In particular, we try to assess the needs and assets of the partner and their data infrastructure. In some cases, this portion of the analysis can be done by the partner, especially if they evaluate Wikibase and whether it fits their needs — as has been the case with the Deutsche Nationalbibliothek and the Bibliothèque Nationale de France.

The assessment phase reveals the needs, capacities, work packages and timelines on both sides and determines whether there is a match. If there is a match, we undertake contract negotiations and create project milestones, deliverables and their associated success indicators. These are to be reported on regularly; the exact frequency should be specified in the formal agreement.

As at the end of the previous phase (feasibility and profiling), if a partner does not meet our requirements, or we theirs, we look for third parties (particularly other Wikimedia chapters) who can assist. If that is not possible, we do not pursue a formal partnership further, but the partner remains in our pool of contacts. This allows us to follow up with the partner in case circumstances change in the future.

4. Implementation[edit]

This phase includes all activities on the roadmap such as capacity building, training, software development, stakeholder meetings, user surveys and workshops. Activities are monitored and evaluated based on the agreed-upon plan.

5. Conclusion and debrief[edit]

In this phase we engage in a final evaluation where we review documentation, deliverables, milestones, success criteria and outcomes to determine how successful the project was and what can be improved in the future. Questions we ask may include:

  • Did we complete all activities outlined in the original agreement?
  • If there were certain activities we failed to complete, why?
  • What worked well in the project team?

We produce a full report with this information and hold two debrief meetings. The first one is internal and includes all Wikimedia Deutschland staff members involved in the partner project’s implementation phase. The second includes both Wikimedia Deutschland and all partner staff involved in the project. Both debrief sessions cover what went well, what did not go well and highlight areas for improvement.

6. Post-project monitoring and support[edit]

We would like to see a growing Linked Open Data Web that is active and sustainable after our involvement ends. In this last phase, we ask questions about this longer-term impact: did the project continue after our participation? How is it faring? Is it growing? To what extent are we still assisting? Is the partner now acting as an ambassador in their field? This last phase begins three to six months after the conclusion-and-debrief phase is complete.

Answers to these questions will help inform future data partnership work and ensure that we learn from the partnerships in which we engage. This phase goes beyond a typical project management task that has a defined end. We remain in touch with our partner and maintain communication. While the project itself is complete, the partner may come to us with a new project in the future — and the circle of data partnerships begins again.