Writing samples
Judge the writing. It is the thing you can actually check.
Two complete documents, published in full rather than sent on request, and the score they were given against our own rubric before anyone else could give them one.
Three documents
One letter, the proposal that follows it, and the marks both were given
Sixth Street Clinic to The Colorado Health Foundation
An overnight peer recovery specialist for post-overdose outreach in Pueblo, Colorado. It names one condition and one gap, gives the organization's own record before it asks for anything, and closes by telling the funder that the program does not become self-funding.
Read the letter Full proposalThe proposal that follows that letter
Same applicant, same funder, same $74,985. Logic model, evaluation design, a line budget with justification per line that reconciles to the dollar, a risks section, and a sustainability section that declines to claim the program pays for itself.
Read the proposalHow both documents score against our own rubric, including where they fall short
A self-grade is worth less than an independent one. It is published anyway, because a sample delivered without a score invites you to assume the writer thinks it is perfect. Two dimensions score 3 out of 5, and the reasons are the most useful part of the page.
Read the marks and the reasonsThe limits
The limits, stated here rather than found later
- Neither was paid for and neither has been sent to a funder.
- The applicant is invented. The clinic, its staff, its partner hospital and its evaluator do not exist. Every figure attributed to them is invented, and internally consistent, so the arithmetic can be checked. The two documents reconcile to the dollar against each other.
- The community statistics are illustrative. They show how a statistic gets sourced inside a sentence. In client work every one is traced to a live, dated, named source before it ships.
- Funder fit is not verified to the dollar. The Foundation's published priorities are quoted and checkable. Its grant range, deadline and intake process are not asserted, because that would mean asserting figures we cannot cite. That caps funder fit at 3 out of 5, and the self-grade says so.
- Neither is submission-ready. A live proposal is written to the funder's own application questions, in their order, inside their character limits. That work has not been done here.
Why they read this way
Both were rebuilt after reading how proposals are actually scored
Both documents were rebuilt against the published criteria in 17 real FY2026 federal notices of funding opportunity, roughly 158,000 words of rubric. The effect is visible in the writing. Need is given 9 percent of the words on purpose, because need carries a median 11.25 percent of the points while approach carries 36.4 percent. There are exactly three objectives, because restraint is scored. There is a real risks section, because eight of the fourteen scored competitions pay for one.
Also available
Earlier samples, in other cause areas
Three earlier documents exist in housing and family services and in rural food security, including a second full proposal. They predate the rubric research above and are weaker for it, which is why they are not the front door here. If your work sits closer to those cause areas, ask and they will be sent.