Execute

Intended audience: DataKind Volunteers

Introduction to the Execute Stage

This stage is when you actually do the project! It begins after the project launches and ends with a project ready to be delivered. Note that the Execute Stage is the least standardized across all projects - and with good reason, as each project varies in terms of what the team is actually doing. One process does not fit all here so this stage does not consist of discrete steps, but rather recommended best practices for how to operate while actually carrying out the project work.

In most cases, execution includes three sub-stages, each of which ends at a Checkpoint:

Explore: Get to know and understand the data, set it up for analysis
MVP: Build a minimum viable product (MVP) of the output for initial stakeholder review
Execute: Conduct quality checks on the MVP, adjust it based on feedback, and create instructional materials

Team roles in the Execute Stage

Project Manager: The Project Manager leads team working sessions, documents key decisions made, and provides updates to all stakeholders. Note that this stage is the least standardized across all projects - you might need to add your own additional checkpoints, milestones, and deliverables based on the project requirements and type of engagement. Work closely with the Data Ambassador, Chapter Leader(s), and/or DataKind staff support to create the best project plan for execution based on your specific requirements. Project Managers are also responsible for great project management, motivating volunteers to keep working, sharing learnings, ethical review, and keeping the team on task and on time.
Data Ambassador: The Data Ambassador is responsible for overseeing the technical elements of executing the project while working alongside the rest of the project volunteers.During execution, the Data Ambassador is responsible for mitigating bias, reviewing risks, conducting project specific technical quality checks, and conducting code review. They determine the tasks best suited for each volunteer, while allowing for experimentation and flexibility if someone feels strongly about what they would like to or would not like to work on, and unblock volunteers as needed. The Data Ambassador performs regular check-ins with the Project Champion and leads Checkpoints and presentations for the partner organization and DataKind on progress.
Chapter Leader or DataKind staff member: Evaluate project progress and ensure it continues to move forward, while it is still of the highest quality and addresses all possible ethical concerns. They help the project team decide whether to pivot, adjust deliverables, or change the timeline when feasibility issues come up. They manage the relationship with the partner organization, keep volunteers motivated, and deal with any volunteer challenges that might come up. Lastly, they decide when it is time to move from the Execute Stage to the Share Stage, to encourage the team to finalize materials and share the results of the project.

Maintaining creativity throughout the execution stage

Throughout the execution stage, be sure that in addition to your technical expertise you are also flexing your critical thinking skills, entrepreneurial spirit, and innovative side. Across the DataKind network of projects, some of the most innovative solutions have come from highly talented volunteers like you! As you are working on the project, regularly think about:

Improving your understanding of the communities impacted and end users of the project: At DataKind, we design with, not for - therefore it is important to carefully listen to your partner organization, as they are the subject matter experts. However, be mindful to not overburden the Project Champion or other staff at the partner organization (e.g., Would it be helpful to read more about the issue area you’re addressing? Has the partner organization collected any survey data from the communities impacted that you could review the results of before building the tool to better understand the humans you impacting?)
Different data science techniques, algorithms, or resources that you could experiment with for this project: We love to innovate and learn alongside our commitment to drive towards a usable, practical solution in our projects. If you want to experiment with a new method that is not in the project plan, that’s amazing! But don’t do so at the expense of ensuring you have time to provide the most useful solution, even if it is the simplest, to the partner organization.
How your project team could best work together: There are several aspects to keep in mind as you and your team manage your work for the project (e.g., time zone, availability, etc.). For example, would your team work with synchronized online coworking time or completely asynchronously? Would your team benefit from an asana or trello board? Feel free to experiment as you are working on the project on the best structure for team collaboration, especially if you have ideas for work styles that are more inclusive of diverse voices and working styles.
Transferability of the output your team produces: Replicability is the adaption and/or generalization of your solution in similar contexts. Scalability, on the other hand, refers to solutions that can be expanded on a larger scale. Consider how this project could be expanded and scale into a tool that would be useful across an entire sector. What would it take for the DataKind team to drive for a larger impact?

Components of the Execute Stage

Note that we call these “components” and not “steps,” as they should be considered throughout execution, rather than as an ordered task list.

Establish cadence for internal team check-ins and meetings with the project champion
Incorporate creativity and experimentation while executing the project
Manage projects in alignment with DataKind’s project management practices
Create quality, reusable, and well-documented code as you go, following coding and data management best practices and maintaining GitHub hygiene
Regularly share lessons learned and challenges overcome with the DataKind Community
Explore Checkpoint: Project team checks in with advisory team and with the partner organization to affirm the project plan and scoping, make any adjustments, or decide if the project needs a pivot
Evaluate potential biases in your approach, analyses, and data; discuss mitigation strategies; and document results
Conduct project-specific technical quality checks
MVP Checkpoint: Project team checks in with advisory team and with the partner organization to review the MVP, make a plan for refinement, and decide whether a pivot or adjustment to the plan is needed
Create instructional materials for maintenance and scale
Conduct project code and documentation review to ensure the solution is technically correct and documentation meets DataKind standards
MUST DO: Project team checks in with advisory team to review what’s been executed, discuss results of code review, and decide if the project is ready for handoff