Stages are the vocabulary a revenue team uses to talk about deals. When every rep agrees on what moves a deal from Discovery to Proposal, forecasts stop being wishful and reviews stop being arguments.
A deal stage is a milestone in the sales process that marks a buyer's progress toward close. Most teams use a five to seven stage model: Prospect, Qualified, Discovery, Proposal, Negotiation, Closed Won, Closed Lost. Each stage has defined exit criteria, a time expectation, and a probability. Stages make a pipeline forecastable instead of a list of hopeful guesses.
The six things to understand before you design, rename, or defend a stage. Exit criteria is the one that separates a real stage model from a label on a column.
A deal stage marks an observable change in the buyer's situation, not a feeling the rep has. Moving a deal from Qualified to Discovery means specific things have happened: a stakeholder was identified, a problem was named, a next meeting booked. If a rep cannot point at the evidence, the stage move is not real.
Exit criteria are the short list of facts that must be verified before a deal can leave a stage. "Economic buyer confirmed." "Budget range disclosed." "Technical validation signed off." Without them, stages drift because every rep interprets Discovery differently and the forecast becomes a committee opinion.
Each stage carries a probability (10%, 30%, 60%, 80%, 95%) that weights the deal in the forecast. The number is not a vendor-supplied default, it is calibrated against your actual historical conversion rate from that stage to Closed Won. Honest probabilities make the forecast arithmetic honest.
Fewer than five and the pipeline hides important transitions (qualification and discovery collapse into one blob). More than nine and reps stop updating because the taxonomy feels like paperwork. Most B2B teams land on six: Prospect, Qualified, Discovery, Proposal, Negotiation, Closed. Longer sales cycles sometimes add a Procurement or Legal stage.
Stage describes how far through the sales process a deal is. Status describes whether the deal is active, on hold, lost, or won. A deal in Negotiation can have a status of On Hold while the buyer handles a budget freeze. Collapsing the two into one field is the most common pipeline design mistake.
Modern CRMs let each team or product line run its own pipeline with its own stages. New business, renewals, expansion, and partner deals rarely share the same milestones. One stage model forced across every motion is why reps call the pipeline "the field I update to make the manager happy."
Most B2B sales teams converge on the same core stage model, with small variations by industry and deal size. The names below are the most common labels. The job of each stage is the same regardless of what you call it. The important part is not the names, it is agreeing on the exit criteria so a stage move means the same thing to every rep and every manager.
A deal exists because a lead has been qualified for fit (right industry, right size, right role) and the first outreach is scheduled. Exit criteria: a two-way conversation booked and confirmed. Typical probability 10%. If a deal sits here more than two weeks, it is usually a lead, not a deal, and belongs back in the lead pool.
A real conversation has happened and the buyer has acknowledged a problem your product addresses. Exit criteria: budget authority identified, problem named, timeframe hinted at. Typical probability 20-25%. This is the stage where BANT, MEDDIC, or whatever qualification framework your team uses gets filled in on the deal record.
Multiple stakeholders are engaged and the buyer is actively evaluating. Exit criteria: economic buyer confirmed, technical requirements documented, success criteria agreed. Typical probability 35-45%. Deals that linger here usually have a missing stakeholder, which is why Discovery is where champion-building either happens or does not.
A formal proposal, quote, or statement of work is in front of the buyer. Exit criteria: proposal sent, pricing accepted in principle, implementation plan reviewed. Typical probability 60-70%. The jump in probability is why Proposal is the stage auditors scrutinize most, and why fake Proposal moves are the biggest forecast risk.
Terms are being redlined. Exit criteria: contract out for signature, legal review complete, procurement aligned. Typical probability 85-90%. Deals here are rarely lost on the merits, but they are often delayed by weeks over indemnity, data processing addenda, or a late procurement stakeholder nobody warned sales about.
The two terminal stages. Won deals trigger onboarding, invoicing, and commission calculation. Lost deals get a lost reason (price, product gap, competitor, no decision, timing) that becomes the single most useful data point in the whole pipeline. If your lost-reason field is a free-text box, your lost analysis is a story your reps tell themselves.
A stage model fails for a short list of recognizable reasons. The fix for each is usually smaller than the discussion that gets started about it. If the pipeline feels broken, the problem is almost always one of these.
Teams try to solve ambiguity by adding stages (Pre-Qualified, Qualified, Highly Qualified) and end up with a Kanban that nobody updates. More stages do not clarify the pipeline, they fragment it. If the question is "what should this stage be," the answer is almost never "add a new one."
A stage without exit criteria is a column on a board, not a process step. Reps advance deals based on how much time has passed, how enthusiastic the champion sounds, or how badly they need the deal to look alive in a review. The forecast then weights hopes rather than facts.
When exit criteria are vague, two reps selling the same product place the same deal in different stages. One sees a champion, the other sees a browser. The manager cannot compare pipelines across the team, cannot forecast the roll-up, and starts normalizing by hand on a Sunday night spreadsheet.
A stage field also used to flag On Hold, Lost, Delayed, or Dormant breaks every forecast report. Keep stage focused on process progress. Use a separate status field for health signals. Modern CRMs ship both; teams that only use one eventually invent the other in a free-text note field.
The CRM ships with default probabilities (10/25/50/75/90) that have nothing to do with your actual conversion rates. Teams that keep the defaults get forecasts that are off by 20-40%. The fix is a quarterly calibration: look at the last 90 days of deals, compute the real conversion from each stage to Won, and update the numbers.
Stage names describe what the seller is doing (Qualifying, Pitching, Negotiating) when they should describe what the buyer has agreed to. A pipeline modeled on seller activity drifts every time a rep is optimistic. A pipeline modeled on buyer evidence stays honest because the evidence is checkable.
Designing stages is one of the shortest pieces of work with the longest half-life. A good stage model will serve the team for years with light tuning. A bad one quietly degrades forecasts, reviews, and coaching for every quarter it runs. The four-step method below is the one most ops teams converge on.
Write down the steps a buyer takes from first contact to signature. Not the sales activity, the buyer evidence: when they recognized a problem, when they engaged a stakeholder, when they got budget, when they chose. Those are your candidate stages. The sales motion is a reaction to each one.
For each stage, list the two to four facts that must be true for a deal to be in it. The test is: can a manager look at the deal record and verify the facts without asking the rep? If not, the exit criterion is a feeling, not a milestone. Rewrite until every criterion is checkable.
Decide which stage transitions trigger a handoff: SDR to AE, AE to Solutions Engineer, AE to Customer Success, Finance to Legal. Each handoff needs the receiving team to know what to inherit. Stages are the right place to document this because the stage field is what the automation reads from.
Every quarter, pull stage conversion rates, average age-in-stage, and win rate by stage reached. Deals that die in Proposal reveal a pricing or product gap. Deals that die in Discovery reveal a champion gap. Deals that take twice as long in Negotiation reveal a legal or procurement gap. The data tells you which stage to fix.
Write a one-page playbook: stage names, exit criteria, typical duration, probability, who owns the deal in that stage, and what the next-step action looks like. Pin it in the CRM where reps work. A stage model that lives only in the ops team's head stops being enforced the first time somebody disagrees with it.
Managers who review by stage ("walk me through your Proposal column") get optimism. Managers who review deals ("what has to be true for this to close?") get reality. Stages are the vocabulary of the review, not the agenda. If every deal gets the exit-criteria question, the stages stay honest on their own.
Legacy CRMs make stages a global list that every team shares. Modern CRMs, including Strkr, treat the pipeline as a configurable object: multiple pipelines, custom stages per team, per-stage probability, per-stage exit criteria, and automation that fires on stage change. The point is that your stage model should fit your motion, not the other way around.
Renewals do not have a Discovery stage. Expansion deals do not have a Prospect stage. Running one global stage list across every motion is why reps either skip stages or lie about them. Strkr lets each pipeline run its own stage list with its own probabilities, so the data model matches the work.
Every stage carries a configurable probability that weights the deal in the forecast. The weighted pipeline total is a single reportable number: pipeline value times stage probability, summed. When probabilities are calibrated against real conversion, the forecast becomes a defensible math problem, not a vibes-based summary.
Stages can require specific fields before a deal advances. Moving a deal into Proposal requires an amount, a close date, and an economic buyer. Moving into Closed Lost requires a lost reason. The enforcement is in the CRM, not in a manager's head, which keeps the data clean without weekly reminders.
A stage change can create a task, notify a stakeholder, generate a document, update a related record, or kick off a workflow. "Moved to Proposal" generates the proposal doc. "Moved to Closed Won" notifies finance and onboarding. The automation turns stages into the engine the whole revenue motion runs on.
The pipeline report shows how many deals entered each stage, how many advanced, how many were lost, and how long they spent there. The leaky-stage view surfaces the single biggest process problem on the team: the stage where you lose more deals than you should. That report pays for the CRM on its own.
Strkr records every stage transition with timestamp and user, so the audit trail can show a deal that was shoved into Proposal on quarter close and shoved back on Monday. Managers who care about forecast integrity read this log weekly, because it is where fake stage moves become visible.
Strkr ships multiple pipelines, per-stage probability, per-stage exit criteria, and stage-triggered automation out of the box. Pricing is published. The features page shows exactly how the pipeline, stages, and forecast fit together.
A deal stage describes how far through the sales process a deal has progressed (Qualified, Discovery, Proposal, Negotiation, Closed). A deal status describes the current state of the deal regardless of stage (Active, On Hold, Dormant, Won, Lost). A deal in Negotiation can have a status of On Hold while a buyer handles a budget freeze. Keeping the two fields separate is the single biggest clarity win in pipeline design.
Five to seven is the sweet spot for most B2B teams. Fewer than five hides important transitions and makes the forecast coarse. More than nine turns the pipeline into a taxonomy exercise and reps stop updating it. Complex enterprise motions sometimes add a Procurement or Legal stage. The right number is the smallest set of stages that each have distinct exit criteria.
Exit criteria are the short list of facts that must be verified before a deal can leave a stage. They matter because they turn stages from opinions into observations. Without exit criteria, two reps interpret Discovery differently, forecasts drift, and deal reviews become arguments. With them, a stage move is a checkable event and the forecast is a math problem.
The most common B2B model is six stages: Prospect (first outreach booked), Qualified (problem acknowledged), Discovery (stakeholders engaged, requirements documented), Proposal (formal quote delivered), Negotiation (contract redlining), and Closed Won or Closed Lost. Names vary by industry, but the underlying buyer evidence rarely does. The stages describe what the buyer has agreed to, not what the seller is doing.
Calibrate against your own history, not the CRM defaults. Pull the last 90 days of deals and compute, for each stage, the percentage that reached Closed Won. That is the real probability. Update it quarterly. Vendor-supplied defaults (10/25/50/75/90) are starting points, not answers. Teams that keep the defaults get forecasts that are off by 20 to 40 percent.
Yes, and they usually should. New business, renewals, expansion, and partner deals follow different buyer journeys and need different stage lists. Modern CRMs like Strkr support multiple pipelines, each with its own stages and probabilities, so a renewal does not have to pretend to go through Discovery. Forcing one global stage list across every motion is a top-three cause of pipeline rot.
A stage conversion rate is the percentage of deals that advance from one stage to the next. If 100 deals enter Discovery and 60 reach Proposal, the Discovery-to-Proposal conversion is 60%. These rates are the diagnostic for pipeline health: the stage with the lowest conversion is the stage that most needs coaching, product work, or qualification tightening. Measuring them is why stages exist.
Review quarterly. Pull stage conversion rates, average age-in-stage, and win rate by stage reached. Rename a stage, merge two stages, or recalibrate a probability only when the data demands it. Rewriting the stage model more often than once a quarter forces reps to relearn the vocabulary and breaks the historical reporting that makes the next review possible.
No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.
