# Sales Process Design

> Use when the user wants to design or fix their sales process — the stages a deal moves through, what has to be true to advance, and what each side does. Also use when the user mentions sales process, sales stages, pipeline stages, deal stages, exit criteria, sales playbook, qualification, MEDDIC, BANT, or "our pipeline is messy." Produces a defined stage-by-stage sales process with entry triggers, verifiable exit criteria, owner actions, and a fitting qualification framework.

- Skill: `sarojkjha/sales-process-design` (Agent Skill)
- Install (CLI): `npx skillmds@latest add sarojkjha/sales-process-design`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sarojkjha/sales-process-design/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- License: MIT
- Author: sarojkjha (https://skillmd.com/u/sarojkjha)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/sarojkjha/sales-process-design

---


# Sales Process Design

Turn an ad-hoc sales motion into defined stages with clear exit criteria — so deals move for real reasons and the pipeline actually means something.

## When to use

When pipeline stages are vague or every rep uses them differently, reps disagree on what "qualified" means, forecasts are unreliable because the stages aren't defined, or the team is scaling and the motion needs to be repeatable.

## Before you start

1. **Read the brand/product context first.** Pull the ICP from `.agents/product-marketing.md`. If none exists, ask the user.
2. **Gather inputs:** how deals actually flow today, the average deal size and sales cycle, and who's involved (SDR / AE / SE).
3. **Confirm the objective:** a process the team can follow that makes the pipeline trustworthy. If the user needs forecasting math, that's `pipeline-and-forecast`; if they need the first-call script, that's `discovery-call-framework` — this defines the stages those live inside.

## Method

Design the process around how customers actually buy, not your internal steps. Every stage needs a verifiable exit criterion — something objective that must be true to advance — so two reps would agree a deal has moved. Bake a qualification framework sized to the deal into those criteria, and keep the whole thing to a few sharp stages.

## Workflow

1. **Map the buyer's journey first** — how customers decide and buy — then design stages to mirror it.
2. **Give every stage a verifiable exit criterion** ("economic buyer confirmed," "technical validation complete"), never a feeling.
3. **Pick a qualification framework** sized to the deal (lightweight for SMB, MEDDIC-style for enterprise) and bake its checks into the exit criteria.
4. **Assign owner actions** per stage: what the rep does, and what the buyer does to advance.
5. **Set stall rules** — a default "age-out" for deals that stop moving.
6. **Keep it to 4–6 stages** and present the process.

## Reference

This skill bundles no data files. Pull the ICP and motion specifics from `.agents/product-marketing.md` and the user's description of how deals flow today.

## Present the result

- The **stage list**, each with: entry trigger, exit criterion, owner actions, typical duration.
- The **qualification framework** and where each check is satisfied.
- **Stall / age-out rules.**
- A one-line definition of a **"qualified opportunity."**

## Guardrails & common mistakes

- **Reject vague exit criteria.** "Had a good call" isn't a criterion — it must be objectively verifiable, or the stage is meaningless.
- **Mirror the buyer, not the org chart.** Stages built around internal handoffs instead of buyer decisions produce dead pipeline.
- **Keep it few.** More than ~6 stages and reps stop using them faithfully.
- **Make "qualified" mean one thing.** If the team can't agree what qualified means, the forecast can't be trusted.

## Related AAJ resources

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

## Related skills

`target-account-list` · `discovery-call-framework` · `pipeline-and-forecast` · `win-loss-analysis`

## 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.