About PSI · v0.3.1 · Authored by Andrew Eriksen
A score for whether value will actually be realized.
Most major buying decisions conflate two things that should be measured separately. "Does this solution create value?" and "Will this organization actually capture that value?" are different questions. PSI measures the second.
The math, in one line
- G
- Annual net value to the buyer: revenue uplift plus productivity capture plus cost take-out, minus what the buyer pays per year. Counted in units of $10,000, so $250K is 25.
- Gₙ
- Log-normalized Gain: 10·log₁₀(1 + G). Each 10x in Gain adds 10 points, so a large deal is compressed but still scores higher than a small one.
- P
- Change cost: the average of technical complexity, organizational disruption, risk exposure, and change-management burden, each rated 1 to 10.
- E
- Adoption signal: time to value, training load, projected adoption, support load, and champion strength, each −1 to +1, summed to −5 to +5.
- Note
- When Gₙ is below P the deal is underwater, and the Experience term flips to (1 − E/10). That way a strong champion always improves the score and a weak one always lowers it. Time decay works the same way: a stalled deal never looks better for being stalled.
Worked example: $300K net value is G = 30, so Gₙ = 10·log₁₀(31) = 14.9. With P = 5 and E = +2, PSI = (14.9 − 5) / 5 × 1.2 = 2.38, which is Strong.
The zones
A PSI reading lands in one of five zones. Coaching follows from the zone, not just the score.
Why this exists
CRMs measure pipeline. Forecast models measure probability. ROI calculators measure value-on-paper. None of those measure the thing that actually determines whether a decision delivers: does the buying organization have the capacity to realize the value once they've paid for it?
That's the gap PSI fills. It scores deals at stage transitions, surfaces dispersion when the deal's sub-signals disagree with each other, and gives reps coaching tied to the specific thing in front of them, not a generic pipeline-health score.
How to use it
Two surfaces, one scoring engine.
Sliders, sensitivity heatmap, portfolio scatter, decomposition. Use this to develop intuition for the math, pressure-test deal shapes, and show prospects what good looks like.
For sales reps scoring real opportunities. 23 questions across 7 short steps (28 for historical deals), behaviorally anchored. Picks between in-motion (active opportunity, score informs forecast) and historical (past deal, outcome captured for calibration).
Buyer-facing version. 25 questions framed for the customer's perspective, about 10 to 15 minutes. Initially used by AEs walking customers through on a call; eventually a self-serve lead-magnet at scale.
Using the score today
PSI is a theory in the proving process. It is usable today as a structured conversation and a consistency check, not as a verdict. Here is how each role should read it.
For sellers
- Score at every stage change. A score older than one stage is stale.
- Read the zone first, then the band. A Strong score with a band wider than 1.5 is not ready for commit.
- Work the top lever on the result page. It shows which single step (more proven value, one less point of change cost, or one better adoption answer) moves the score most.
- Every "I don't know" widens the band. Close those gaps before you forecast.
- If your gut zone and the PSI zone disagree, write down why. That disagreement is the most useful calibration data we collect.
For implementation and technical staff
- Own the Pain and Experience answers. You know the integration, training load, and support reality better than the seller.
- Score independently before comparing with the seller. Premature agreement hides the signal.
- Technical Pain of 8 or more, or Pain items more than 4 points apart, is the scope conversation to have before contract. The average hides the spike.
- Training hours, 90-day adoption, and ticket change are what you will be measured on after go-live. Answer them as the person who has to deliver them.
- After go-live, record the outcome. Scores without outcomes cannot prove or disprove the model.
Evidence for and against
The core idea, that creating value and capturing it are different questions, is well supported in implementation science and health IT research. The exact formula, weights, and zone cut points are not derived from any outside source; they are hypotheses that only outcome data can confirm. Several sources also challenge parts of the model, including averaging 1 to 10 ratings and trusting seller-entered estimates.
Read the evidence review →The three-movement vision
PSI starts as a measurement instrument and grows into a category.
Internal instrument
PSI scores every deal in active pipeline. Runs at stage transitions, trends weekly, blocks low-PSI deals from entering commit-forecast. Shortest horizon, highest-certainty value: sales discipline as an asset.
Public-facing tool with a benchmarking corpus
The visible output is the score; the durable output is the dataset. Every submission builds a proprietary corpus that, at scale, produces percentile rankings no one else can: ‘Your RCM modernization PSI of 0.6 sits at the 23rd percentile for regional health systems.’
Methodology, book, certification
What NPS is to customer loyalty, PSI aims to be for any major operational decision (enterprise tech, credentialing, infrastructure, MSP services, voice AI): anything where the buyer pays once and adoption decides the outcome. That status has to be earned with outcome data; NPS itself has been challenged on predictive validity. An annual State of Change Readiness report. A PSI-Certified Advisor program. Long horizon; called at month 18 based on signal.
What this is not
PSI is a structured decision input. It is not investment advice, not a substitute for diligence, and not a replacement for human judgment. The score is a tool: useful when treated as one, dangerous when treated as truth.
Have your CFO or outside counsel validate financial assumptions before acting. The model is in active calibration; scores may be retroactively updated as the rubric tunes against real outcome data.
Author
PSI is the original work of Andrew Eriksen. It was first deployed at Continuant Healthcare Solutions, where it serves as the internal commit-forecast instrument. Authorship and intellectual property are personal, not institutional; PSI is portable to any organization that adopts the methodology.