Docs/Intelligence/ROI Projection
Analysis

ROI Projection.

The language every approver speaks: dollars, timeline, payback. ROI Projection models a specific partnership's revenue trajectory, onboarding and ongoing cost, time-to-payback, and sensitivity to assumption changes — across conservative, base, and upside scenarios — so your investment committee can decide on numbers, not vibes.

At a glance
Input: pick an org (typically from a PRISM match or existing Pipeline deal) + optional override of default assumptions.
Output: three-scenario 3-year projection with revenue, cost, payback, sensitivity, and peer benchmark comparison.
Latency: 10-18 seconds via Claude Sonnet + CORTEX benchmark join.
Use case: the number your partnership committee sees before saying yes or no.

What the projection contains

Every ROI Projection produces eight structured sections:

1Three-year revenue model

Year 1, Year 2, Year 3 revenue attributed specifically to this partnership. Shown under three scenarios: conservative (70% of base, accounting for typical slippage), base (the model's central estimate), and upside (140% of base, accounting for accelerated adoption). Each year includes a confidence band.

2Onboarding cost

One-time cost to get to first-revenue. Includes: integration engineering (days × loaded rate), legal review (standard partnership contract + any bespoke structures), compliance review (KYB re-verification cost if needed, internal risk review, any required examination-adjacent preparation), initial marketing collateral.

3Ongoing annual cost

Partnership success manager time, technical support, quarterly review cadence, compliance monitoring contribution, incremental marketing. Scales with revenue per the deal structure.

4Payback period

Number of months until cumulative revenue surpasses cumulative cost. Computed separately per scenario. Anything over 36 months flags for scenario-planning review.

5Sensitivity analysis

Which single assumption most affects the outcome, and what happens if it moves 20% in either direction. Typical most-sensitive assumptions: customer acquisition rate, average deal size per end customer, integration timeline (longer integration = later revenue = worse NPV).

6Peer benchmark comparison

How this projection compares to equivalent partnerships in the CORTEX industry benchmarks — is the base case at the peer median, lagging, or leading? If you're claiming top-quartile revenue per partnership from this deal, the model flags the assumption as aggressive.

7Assumptions register

Every number in the projection is backed by an explicit assumption — rate × volume × duration — which can be inspected and overridden. The register shows source ("CORTEX benchmark", "your historical average", "user-supplied"), value, and a confidence level.

8Risk adjustments

Factors that warrant discount: deal-stage risk (Series A partner = higher execution risk), regulatory risk (partnership type with recent enforcement action), concentration risk (overweight reliance on a single partner channel).

What TECH uses to seed the model

The projection is not generated from the air. Inputs:

When to run it

ROI Projection has a clear place in the partnership lifecycle:

  1. After Fit Analysis, before the first commercial discussion. The Fit tells you whether the partnership makes sense; the ROI tells you what it's worth if it does.
  2. When sizing the pilot budget. The onboarding cost section is designed to be ported directly into your pilot budget line item.
  3. Before a committee presentation. The PDF export is designed for committee consumption: one page per scenario with the base-case as the primary slide.
  4. When re-forecasting mid-pilot. Update the assumptions with observed pilot performance, regenerate — compare to the original projection.

Worked example structure

An ROI Projection for a mid-sized fintech pursuing a co-brand card partnership with a regional bank might look like:

Year 1 revenue (base)
$0 (pilot phase)
Year 2 revenue (base)
$180K (first commercial year at modest volume)
Year 3 revenue (base)
$420K (scale reached on the base path)
Onboarding cost
$85K
Annual ongoing cost
$38K
Payback (base)
18 months
Most sensitive assumption
End-customer conversion rate — 20% down-move extends payback to 28 months
Peer benchmark
Base-case Year 2 revenue is at peer p50; Year 3 reaches p60

Export

ROI Projections export as a PDF in an executive-ready format: cover page, one page per scenario, sensitivity analysis, assumptions register, and an appendix with the peer benchmark visuals. Designed to drop into a committee deck without reformatting.

Not a forward-looking statement
ROI Projections are model-generated scenarios based on the inputs and benchmarks described above. They are not financial forecasts in the SEC Regulation S-X sense, are not audited, and are not suitable as the sole basis for a commercial decision. They are labelled AI-generated and presented with explicit confidence bands. Treat them as a structured starting point for committee discussion — never a committee substitute.

API surface

POST /intelligence/roi/<org_id>/
Generate an ROI Projection for that target
GET /intelligence/roi/<org_id>/
Retrieve the most recent projection
Get the number for your committee

Run ROI before the ask.

Open Intelligence →