Pipeline And Forecast

Pipeline & Forecast

See whether the pipeline can actually hit the target — weighted forecast, coverage, the gap, and how much new pipeline to build.

When to use

When forecasts are sandbagged or wishful and you want a grounded number, when you need to know whether there's enough pipeline to make quota, when deciding whether to add pipeline or push the deals you have, or for board/quota reporting.

Before you start

1. Read the brand/product context first. Pull any relevant goals/targets from .agents/product-marketing.md. If none exists, ask the user.
2. Gather inputs: the target/quota for the period, the open deals (amount + stage), the stage→probability mapping (from sales-process-design or historical conversion — not optimism), and the average win rate. If stages and probabilities aren't defined, run sales-process-design first.
3. Confirm the objective: a defensible number and the right lever to pull. (For LTV/CAC/payback questions, that's unit-economics.)

Method

Weight open pipeline by real stage probabilities to get an expected forecast, then compare open pipeline to the target as a coverage ratio against the ~1/win-rate benchmark. Separate commit (near-certain) from best case (everything open) so the number is honest, and name the lever: build pipeline if coverage is thin, or improve conversion if coverage is fine but the forecast is short.

Workflow

1. Gather the target and the open pipeline — deals with amount and stage.
2. Apply stage→probability weights from real conversion history to get the weighted forecast.
3. Compute coverage = open pipeline ÷ target (the engine flags it against ~1/win-rate).
4. Compute the gap and the new pipeline needed = gap ÷ win rate.
5. Separate commit from best case.
6. Name the lever and present the result.

The framework

weighted forecast = Σ (deal amount × stage probability)
coverage          = open pipeline ÷ target
required coverage ≈ 1 ÷ win rate
gap               = target − weighted forecast
new pipeline need = gap ÷ win rate
commit = Σ amount for deals at prob ≥ 0.75   ·   best case = Σ amount (all open)

Run the tool

Paths assume you installed with npx skills add. From a clone of this repo, use skills/pipeline-and-forecast/resources/… instead.

node .agents/skills/pipeline-and-forecast/resources/forecast.js --input=pipeline.json
node .agents/skills/pipeline-and-forecast/resources/forecast.js  # built-in demo

Input JSON: { "target": 500000, "winRate": 0.25, "stageProbabilities": { "Proposal": 0.5, ... }, "deals": [ { "name": "Acme", "amount": 60000, "stage": "Proposal" } ] }. It prints the weighted forecast, commit vs best case, coverage ratio with a healthy/thin flag, the gap, the new pipeline needed, and the recommended lever. Use these exact numbers — don't restate them from memory.

Reference

The forecast engine and its input shape live in resources/forecast.js. Stage probabilities should come from sales-process-design or historical conversion data.

Present the result

• Weighted forecast, with commit vs best case.
• Coverage ratio against target, with a healthy/thin flag.
• Gap to target and new pipeline needed at the current win rate.
• The recommended lever — build pipeline vs. improve conversion.

Guardrails & common mistakes

• Probabilities must come from real conversion history. Flag clearly when they're assumed — an optimistic forecast is just a wish.
• Keep commit and best case separate. Reporting only the upside is how forecasts miss.
• Fit the benchmark to the win rate. A 3× coverage rule is wrong at a 10% win rate.
• Mark assumed inputs as assumptions so no one mistakes the model for a promise.

Related AAJ resources

• AAJ tools — https://aajconsult.com/tools.

Related skills

sales-process-design · target-account-list · unit-economics · marketing-budget-planning

Credits

Original AAJ skill. The Agent Skills format and the marketing-skills catalog by Corey Haines (coreyhaines31/marketingskills, MIT) were references for structure and coverage; this skill is independently written. See the repository README for the full reference list.