Each of these engagements started with a clear brief and ended somewhere else. The write-ups are about how the real question surfaced, and what it changed.
EdReports / Education nonprofit
A new category of Pre-K evaluation reports had to satisfy several different definitions of “enough evidence” at once, without diluting the rigor behind the review.
Read the EdReports case →What didn't fit
Reviewers, districts, and publishers each read the reports differently, and each of them was right.
The proof
A layered path through the evidence, and one of the smoother launches the organization had run.
Oaken / Agriculture
A scoped design-system job surfaced a harder question: what should this product actually remember when the owner, the lease, or the people change?
Read the Oaken case →What didn't fit
Oaken's best customers were checking every payment in a spreadsheet.
The proof
Institutional memory that was supposed to walk out the door for two months, and instead just kept working.
Third-party logistics / Operations
A largely undocumented, ninety-step invoicing process traced back to an onboarding form nobody actually completed.
Read the Ninety Steps case →What didn't fit
The onboarding form behind every invoice only ever got a sixth of its questions answered.
What changed
The invisible process got mapped, and the fix was placed upstream where the missing information was actually created.
Energy / Industrial manufacturing
A recovered relaunch left an unscoped thread behind: buyers weren't searching the way the company was organised. Fieldwork turned a taxonomy fix into the argument for the year's roadmap.
Read the portfolio mapping case →What didn't fit
The navigation was built around internal product names. Buyers didn't search that way.
What changed
A journey map printed at wall size, with deliberate gaps in it, got four divisions to write one roadmap they all owned.
Mercury / Life-science logistics
A new platform for a complicated logistics service was drifting toward a design brief. Customers had already built the system they needed out of spreadsheets, email, and their own ingenuity.
Read the Mercury case →What didn't fit
Customers trusted the squads. What wore them down was not knowing where their shipments stood.
What changed
A home screen that did three things: track, pickup, ship. The rest was sorted into software, automation, or the squads.
Send me the mess as you see it. Don't diagnose, just a high level overview and I'll respond with what catches my attention.