Grant Compass

Together with DLH we asked: How might we create a more intuitive and seamless user experience for both reviewers and regional offices that enhances usability and improves workflows?

I was the Senior Interaction Designer on a six-person IDEO team: a project lead, a design research lead, a product designer, a visual communications designer, and our executive director. I ran the field research, designed and prototyped all six concepts myself, and presented them to DLH leadership.

• 17 interviews across reviewers, review leads, field operations managers, program specialists, and regional program managers.
• 3 reviews shadowed in the field, at the Fond du Lac Reservation, Kaysville, Utah, and Duluth, Minnesota.
• 6 role one-pagers mapping who does what across the Office of Head Start and DLH.
• 2 end-to-end journey maps of the FA1 and FA2 review process.

Working with the Head Start Community

The first step was shadowing the Head Start review process at grant recipient sites: the Fond du Lac Reservation, Kaysville, Utah, and Duluth, Minnesota. I spoke with preschool educators, Head Start parents, and reviewers, and got to look over their shoulders as they used the software.

Two journeys, mapped end to end

Focus Area reviews come in two shapes: a two-and-a-half-day virtual review, and a five-day on-site review. I mapped both end to end across every role, surfacing handoffs and gaps no single interview would have caught alone.

Six roles, one review

A single review runs across six roles, three inside the Office of Head Start and three at DLH, and almost nobody in it had a clear picture of the other five. I wrote a one-pager for each: what they are accountable for, what blocks them, and where in the review each frustration actually lands.

Read across all six and the same problem turns up in three separate roles, described independently and without prompting: work disappearing. That is the thread I pulled on later.

Through embedding with this community we arrived at five opportunities for the next version of this software: make communication trackable and traceable; shift from people informing systems to systems informing people; increase transparency in shared workflows; make collection reliable, comprehensive, and collaborative; and surface the urgent and actionable automatically.

The shift we were designing for

Across every role, the same six words kept describing the current system. We used their opposites as the test for every concept that followed.

We generated far more than six ideas, then sorted everything into three tiers: six primary concepts, the ones that would do the most and could be built soonest; six secondary concepts straight out of research but less urgent; and a set of quick fixes that could ship without waiting on a redesign.

How we chose, and where it got uncomfortable

Picking between the six was the hardest part of the project, so we did it in the room. In the final session, DLH leadership spent a fixed budget across the six concepts, putting real weight behind what they wanted most; separately, we plotted the same six on an impact-versus-effort matrix.

The two exercises disagreed, and that disagreement turned out to be the most useful thing we made. Leadership’s biggest bets went to the Personalized Homepage, the most visible and ambitious of the six. The matrix favored the Review Tracker instead: close to the same impact for a fraction of the effort. Multiplayer Mode, despite addressing a frustration nearly every interview raised, scored lowest on impact of all six.

I argued for starting with Multiplayer Mode anyway, the concept our matrix ranked lowest. Three of the six roles had independently told us they’d lost work in the system, and during the shadowings I watched it happen: the system kicked a reviewer out mid-session and took their notes with it. Losing an afternoon of notes doesn’t show up on a matrix, but it’s exactly what makes people stop trusting a tool. We estimated the fix alone would roughly halve the time spent editing a review.

So the room held three different answers: leadership’s budget said Personalized Homepage, the matrix said Review Tracker, I said Multiplayer Mode. We handed DLH all six with the matrix, disagreement included, because the sequencing decision belonged to the team who’d build it, and the open tension was more useful to them than a ranked list that hid it.

The Concepts

What we deliberately did not lead with

Six more concepts came directly out of research but were less urgent, and we said so rather than quietly dropping them: user feedback, release communication and onboarding, a community of shared practices, a grant recipient page, data visualization, and reviewer teambuilding and scheduling. That last one came from reviewers describing unpaid travel days and evenings finishing notes. A real problem, just not one a monitoring-system redesign was going to fix first.