Pursuit ReplayPricingLearnWorkbenchAboutThe SprintResultsFAQBook a Pursuit Replay →
‹ frwrd.eval

What the board hands you.

Four artifacts, shown in full. Everything on this page is a constructed demonstration — the same made-up bidder as the Replay sample (“Harbor Data Systems”), the same made-up state DOT solicitation, this time with a draft technical volume run through the board twelve days before the deadline. The pursuit is synthetic. The method is the real one.

1 · The compliance screen

The Gatekeeper runs before anything is scored, because real boards screen before they read. Every Section L instruction, graded. synthetic demonstration

Section L instructionStatusFinding
L.4.2 — Volume II shall not exceed 30 pagesFAILED31.5 pages once resumes are counted. Amendment 2 changed the counting rule to include appendices; the draft still assumes the original rule. This alone is grounds for elimination.
L.4.5 — 12-point font, 1-inch marginsmetCompliant throughout, including tables.
L.5.1 — Acknowledge all amendments on SF30AT RISKAmendments 1 and 2 acknowledged; Amendment 3 (posted 4 days ago) is not. The form is drafted but unsigned.
L.6.3 — Staffing plan as Attachment J-4 formatmetFormat matches; two labor categories use titles that differ from J-4's — flagged for consistency, not compliance.
L.7.1 — Submit via the state portal by 14:00 CTAT RISKNo dry-run submission has been made. Portal accounts lapse after 90 days; last login was 6 months ago.
Why this runs first: most preventable losses are screens, not scores. A deficiency in the matrix costs points; a failed instruction costs the whole bid.

2 · The Skimmer’s read trace

Eight minutes, the way a tired evaluator actually reads: headings, topic sentences, callouts, captions. The trace records what was seen, in order, and what never was. synthetic demonstration

MinuteWhat the Skimmer sawResult
0–1Executive summary, read in fullWin theme 1 (state-portal integration record) lands — it's the first sentence.
1–3§2 headings + topic sentences§2.1's heading promises a 'proven approach' with no proof in the first sentence. Skipped to the figure. The figure caption carries no claim.
3–5§3 staffing — table scanned, prose skippedWin theme 3 (named PM with DOT history) is buried in paragraph four. Never seen.
5–7§4 transition plan headingsThe 30/60/90 callout box lands. Win theme 5 (zero-downtime cutover) is findable.
7–8§5 skimmed, appendices skippedWin themes 6 and 7 live in an appendix. An evaluator giving you eight minutes does not read appendices.

Verdict: 3 of 7 win themes findable in one pass.Nothing here is a compliance problem. All of it is points left on the table — and it’s fixable by moving sentences, not rewriting sections.

3 · The evaluation worksheets

Each evaluator writes findings in debrief language, tied to the sentence — or the silence — that caused them. An excerpt across three of the ten worksheets. synthetic demonstration

The terms are FAR 15.001’s: a weakness increases the risk of unsuccessful performance; a significant weakness appreciably increases it; a deficiency is a failure to meet a requirement.

EvaluatorFactorFindingNote
The SMETechnical approachWeakness§2.3 claims a 40% reduction in manual review 'based on prior deployments' — no deployment, customer, or measurement is named. Unsubstantiated as written.
The SMETechnical approachStrength§4's cutover plan names the rollback trigger, the owner, and the threshold. Checkable and specific — evaluators can score this.
The Points AccountantManagementSignificant weaknessM.3.2 awards points for 'demonstrated key-personnel retention.' The draft never addresses retention. Best material on this sits in the appendix, which M does not evaluate.
The Past-Performance HawkPast performanceWeaknessReference 2 is relevant and recent but one-fifth the contract value. The stated test includes scale; the draft argues similarity instead of addressing it.
The Cost RealistCross-volumeWeaknessThe staffing table totals 11,840 hours; the narrative commits to 'day-one full staffing' that the ramp in the table contradicts. One of them is wrong.

4 · The consensus debrief

The board’s combined read, and the fix list ranked by score impact — what actually changes points, in order. synthetic demonstration

#FixWhy it ranks here
1Cut Volume II to 30 pages under Amendment 2's counting ruleElimination risk. Nothing else on this list matters if the volume is thrown out.
2Acknowledge Amendment 3; sign the SF30; dry-run the portalTwo more screens closed in under an hour.
3Answer M.3.2 retention in the management section, not the appendixClears a significant weakness in a factor worth 25 points.
4Name the deployment behind the 40% claim, or cut the numberAn unsubstantiated claim reads worse than a modest one.
5Move win themes 3, 6, 7 into first sentences under their headingsRaises findability from 3 of 7 to 7 of 7 without adding a page.

What the board didn’t say:whether this DOT trusts Harbor, what the incumbent will bid, or whether the price wins. It can’t know those. A mock evaluation is a measured read of a draft under stated conditions, not a prediction of award — the judgment calls stay with people, which is the point.

This packet, on your draft.

Book a call and we’ll run the board on a live volume with you — screen, traces, worksheets, and the ranked fixes, before the deadline instead of after the debrief.

Run one on your draft