How to define pipeline stages
Stages are the backbone of a pipeline. Get them right and the forecast, win-rate analysis, and coaching all become reliable; get them wrong and every downstream number is noise.
The single most important rule is to name stages after what the buyer has committed to, not what the seller just did. That one choice separates a pipeline you can trust from one you cannot.
Short answer
Define pipeline stages by naming each one after a buyer commitment, using 5 to 7 stages total, and writing a single objective exit criterion for each. Attach a default win probability per stage for weighting. Buyer-based stages with clear exit tests keep every rep consistent and make the forecast trustworthy.
Step by step
Interview recent won and lost deals
Talk to 5 to 10 recent deals and write down the real sequence of buyer commitments: took a meeting, admitted a problem, looped in budget, agreed to a proposal, signed. Your stages come from this reality, not a template.
Choose 5 to 7 stages
Fewer than 5 hides signal; more than 7 creates busywork and forces reps to guess. A common backbone is New, Qualified, Discovery, Proposal, Negotiation, Closed.
Name each stage for the buyer commitment
Reject seller-task names like "Demo scheduled" and use commitment names like "Problem confirmed" or "Proposal accepted." Buyer-based names make progress mean something.
Write one exit criterion per stage
Each stage needs an objective, one-line test a deal must pass to advance. If two reps would disagree about whether a deal qualifies, the criterion is too vague.
- Qualified exit: budget owner and problem confirmed
- Proposal exit: pricing sent and champion confirmed
- Negotiation exit: verbal yes and paper in legal
Attach a default win probability
Give each stage a baseline close probability for weighting, grounded in your historical stage-to-close conversion, not gut feel. These feed weighted pipeline and the forecast.
Buyer stages vs seller stages
A seller stage like "Demo delivered" tells you what your rep did. A buyer stage like "Buyer agreed to evaluate" tells you what the customer committed to. Only the second predicts a close.
When stages describe buyer commitments, an advancing deal is real progress, not just activity. That is what makes stage-based forecasting accurate.
How Ardovo helps
Ardovo lets you define stages with explicit exit criteria and default probabilities, and Rook updates a deal's stage from meeting notes and activity, so stages reflect what the buyer actually did instead of wishful data entry.
Frequently asked questions
How many pipeline stages should I have?
Five to seven for most B2B teams. Fewer hides where deals stall; more creates data-entry busywork and forces reps to guess. Start with New, Qualified, Discovery, Proposal, Negotiation, Closed and adjust to your motion.
What is a stage exit criterion?
A specific, verifiable condition a deal must meet to advance, such as "economic buyer confirmed" or "proposal sent." Exit criteria keep stage data consistent across reps, which is what makes the forecast trustworthy.
Should stages be based on buyer or seller actions?
Buyer actions. "Demo delivered" is a seller action; "buyer agreed to a proposal" reflects real progress. Buyer-based stages produce far more accurate forecasts because advancement means genuine commitment.