Okhp3 Outcome Modeling Sales

okhp3-outcome-modeling-sales

OverKill Hill P³ · overkillhill.com · github.com/OKHP3

Apply outcome modeling to sales and business decisions where raw wins can hide discounting, easy-opportunity selection, churn, service cost, or money left on the table. The primary object is economically efficient contribution, not a flattering headline metric such as close rate.

Scope

Reframe the sales outcome

Do not treat a 100% close rate as proof of superior performance. First examine opportunity assignment, customer fit, deal size, competition, discounts, cycle time, retention, expansion, service burden, and capacity consumed.

Use an economic objective such as:

expected contribution margin
- discount cost
- acquisition and service cost
- time and capacity cost
+ retention and expansion value
+ incremental lift above opportunity baseline

The exact objective must be agreed before ranking salespeople or allocating resources.

Computational payload

Read references/computational-model.md for expected contribution, incremental lift, and constrained allocation formulas. Read references/glossary.md before using sales abbreviations. Reproduce the synthetic allocation with scripts/calculate-sales-allocation.py examples/sales-example.json. The helper uses local JSON, exhaustive search for small fixtures, and no CRM connection or file writes.

Workflow

1. Load okhp3-outcome-modeling-core and define the decision horizon, owner, and target.
2. Identify the opportunity, account, rep, territory, product, and time grains.
3. Build pipeline or account states using only information available at each forecast date.
4. Separate assigned opportunity quality from rep actions and outcomes.
5. Normalize revenue, margin, discount, cycle time, retention, and expansion measures.
6. Compare raw conversion with expected value, contribution, and incremental lift.
7. Check selection bias, territory effects, customer segment, competitive intensity, and reverse causation.
8. Use experiments or holdouts when the recommendation is an intervention such as discounting, routing, or outreach.
9. Allocate under capacity, budget, service, and risk constraints.

Validation gates

Before ranking people or accounts, verify the as-of boundary, opportunity-assignment fields, margin definitions, and retention window. Reject a recommendation when the economic objective or causal comparison is undefined.

Output contract

Return:

• target and forecast horizon;
• data quality and assignment boundary;
• pipeline or account state definition;
• raw outcome versus economic-value comparison;
• margin and discount analysis;
• selection-bias and confounding assessment;
• calibrated forecast or ranking;
• recommended allocation with constraints;
• retention, expansion, and downside risks;
• next measurement or experiment.

Never punish a salesperson for a low close rate until opportunity difficulty and assignment quality are modeled. Never reward a high close rate without testing whether price, fit, volume, margin, and retention justify it.

References

• references/computational-model.md -- sales equations and example.
• references/glossary.md -- sales terms and abbreviations.
• examples/sales-example.json -- synthetic opportunity fixture.
• scripts/calculate-sales-allocation.py -- transparent small-allocation helper.
• ../okhp3-outcome-modeling-core/SKILL.md -- shared objective, state, and validation contract.

About

Built by Jamie Hill · OverKill Hill P³ Published at github.com/OKHP3 Part of the OKHP3/skillz Agent Skill library. MIT License -- free to use, fork, and adapt. A nod to the source is appreciated.