An RFP for a value realization platform fails in one of two ways: it asks for features and gets prose, or it runs for two quarters and ends with nobody able to say why the leading vendor is leading. This kit is written to avoid both. Every requirement is answered with one of four codes and a piece of evidence, the matrix scores itself, two gating rules remove vendors before any total is read, and the proof of value is a paid 90-day launch that ends on a date you set before it starts.
Everything in it describes practices and tests, not products. Propelon answers the same matrix with the same codes and labels its own roadmap items as such. The kit is free to copy and adapt with attribution.
What is in the kit
| File | Format | What it holds |
|---|---|---|
| Evaluation guide | Word, 11 pages | The buyer’s guide: why this is not a tool evaluation, eight baseline measures, ten vendor questions, 26 criteria with first-90-day tests, the alternatives, evaluation design, the economics, red flags and sources. |
| Evaluation scorecard | Excel | The 26 criteria with default weights, the 0-5 scale, three vendor columns, the must-have gate, totals by group and a baseline tab. |
| RFP template | Word, 14 sections | Introduction, instructions to bidders, timeline, objectives and success criteria, a current-environment questionnaire, scope, requirements, the evaluation method, the proof-of-value stage, implementation and support, commercial terms, security and legal terms, submission format, and a glossary. |
| Requirements matrix | Excel, 104 requirements in 14 sections | Each requirement carries a priority (Must, Should, Could), a default weight you can change, a response code the vendor must choose, an evidence column and an automatic score. A summary tab totals by section and applies the gating rules. |
| Proof-of-value test plan | Word, 22 tests over 90 days | Each test names the requirements it covers, a pass condition anyone can observe, the evidence to keep and the owner, from readiness in weeks 0-2 to the Day-90 executive review. |
| Reference checks and scoring guide | Word | Twelve reference questions to ask with Finance on the line, a demonstration script built around your own data, the rules for weighting, roadmap answers and ties, and a one-page decision record. |
Response codes and gating rules
Vendors answer every requirement with one code and cite the evidence that supports it. The codes carry points; the matrix multiplies them by the weight you set.
| Code | Points | Evidence that supports it |
|---|---|---|
| Fully supported today | 1.00 | Demonstrated on your data, or documented customer use |
| Partial or configurable | 0.50 | Demonstrated with configuration or a documented workaround |
| Roadmap with a committed date | 0.25 | Date in the order form; retested in the proof of value |
| Not supported | 0.00 |
Two rules apply before any total is read. A Must requirement answered Not supported eliminates the vendor. A Must answered Roadmap needs the committed date in the order form and is retested in the proof of value; two or more such items eliminate the vendor.
The five-step method
- Capture your baseline first: initiatives in flight, the share with an attested baseline, Finance-signed value in the last year, idea-to-production time, spend per pilot, adoption against the target population, monthly AI operating cost. The evaluation guide shows how to collect each in a week.
- Weight the matrix to your biggest problem, keep the total within 5% of the default, and fix the weights before vendors respond.
- Issue the RFP with a decision date on the calendar.
- Score on demonstrated evidence, not prose: a capability shown on your data counts; a roadmap slide is a roadmap item.
- Run the 90-day proof of value with the leading vendor, paid and credited against year one, and hold the Day-90 review on the date you set.
Read the evaluation guide, which the first step depends on, before you issue anything.
What the matrix covers
Fourteen sections, 104 requirements. The weight sits in sections 4 to 7, where a number either becomes verifiable or does not.
- Company and references (6)
- Intake and portfolio (8)
- Gates and decision governance (8)
- Baselines and the value model (9)
- Finance attestation and value states (8)
- Adoption and engagement telemetry (8)
- AI capacity and cost governance (7)
- Reporting and executive views (7)
- Learning loop and institutional memory (5)
- Integrations and headless operation (8)
- Architecture, deployment and data (7)
- Security, privacy and AI governance (8)
- Measurement and proof of value (6)
- Commercials, implementation and support (9)
Vendor neutrality
The requirements describe practices and tests, not products. Propelon answers the same matrix with the same codes and labels its own roadmap items as such. Propelon has one Fortune 5 healthcare reference with telemetry-verified adoption at scale and design partners in their first cycles; it does not yet have a public Finance-signed value reference, and the kit does not claim one.
If you would rather not run an RFP at all, the same work happens in the Value Scope Workshop before Day 1 of a 90-day launch: your portfolio, your baselines, your budgets and your gates, in your own numbers, with Finance in the room.
Get the kit
One zipped package: the evaluation guide and scorecard, the RFP template, the 104-requirement matrix, the 22-test proof-of-value plan and the reference questions. Version 1.0, October 2026.
The kit arrives by email with a personal link that stays valid for 72 hours. We use what you tell us to send it and to follow up once, by a person, if what you are evaluating is something we can help with. No newsletter, no drip sequence.
Prefer email? Write to sales@propelon.ai with your organization and we will send it by hand.