Sales management • 12 min

    How to build a reliable B2B sales pipeline: stages defined by buyer-side facts, written exit criteria, a 30-minute weekly review, the metrics that matter and the mistakes to avoid.

    Back to blog

    Short answer

    Effective sales pipeline stages come down to five to seven open stages, each defined by a verifiable fact on the buyer's side rather than something the rep did. Every stage needs a written exit criterion and required fields in the CRM. A 30-minute weekly review focused on stuck deals beats a long meeting about total pipeline value.

    12 min••Updated •By Hugo Granjard

    Founder of Reakly

    Target keyword: sales pipeline stages

    Key takeaways

    • A pipeline stage should be defined by a verifiable fact on the buyer's side, not by something the rep did.
    • Five to seven open stages are enough, each with a written exit criterion and required fields in the CRM.
    • A 30-minute weekly review focused on stuck deals beats a long meeting about total pipeline value.
    • Stage probabilities should be recalculated from your own closed deals, and closing deals as lost with a reason makes the pipeline more reliable.
    Cover of the Reakly article on sales pipeline stages: funnel of bars representing the stages of a sale

    Most advice on sales pipeline stages sounds the same. Break your sales cycle into stages. Attach a probability to each one. Fill the top of the funnel. Watch the total grow. That is exactly why most pipelines are useless for actually running a sales team.

    A $400,000 pipeline makes everyone feel good. Then the quarter closes at $60,000. Half the deals hadn't moved in two months. A quarter of them never really started. Nobody lied. Every rep just staged their deals based on gut feel.

    The problem isn't how many stages you have. The problem is stages defined by what the rep did instead of a verifiable fact on the buyer's side. "Demo done" describes your activity. "The decision maker confirmed the need and the budget" describes the buyer's. Only the second one predicts a signature.

    And sales cycles aren't getting shorter. In Salesforce's State of Sales 2026, a survey of 4,050 sales professionals, 57% of sellers say customers take longer to decide than they used to, and 67% say their customers require extensive education before buying. Longer deals mean more deals quietly stalling without anyone noticing.

    This guide gives you a method to build a pipeline that tells the truth. Two pillars: observable exit criteria for every stage and a short weekly review that cleans the pipeline instead of reassuring the team.

    What a sales pipeline actually is

    A sales pipeline is the list of every open deal, organized by stage, with a value and an expected close date. It answers three questions: how much will we close, when, and where do we need to act this week.

    Pipeline, funnel, forecast: three different things

    These terms get mixed up constantly. They don't measure the same thing.

    ConceptWhat it measuresThe question it answers
    Conversion funnelAggregate volumes at each stepWhere are we losing the most prospects?
    Sales pipelineIndividual deals, each with an owner and a next stepWhat should we do, on which deal, this week?
    ForecastExpected revenue over a periodHow much will we close this quarter?

    The funnel is a snapshot. The pipeline is a work list. The forecast is a commitment. If the pipeline is wrong, the other two are wrong too.

    What belongs in the pipeline, and what doesn't

    A contact who opened an email is not a deal. Neither is someone who accepted your LinkedIn connection request. Putting them in the pipeline inflates the number and buries the real opportunities.

    My rule is simple: a deal enters the pipeline once a real two-way conversation has happened with someone at the target account. Before that, it's prospecting, tracked separately. That's exactly what a lead scoring model is for: deciding which contacts deserve that first conversation.

    A pipeline isn't a list of people you've talked to. It's a list of deals that could actually close.

    Which sales pipeline stages work in B2B?

    There's no universal breakdown. But most SMB and mid-market B2B sales fit into five to seven open stages, plus two closed ones: won and lost.

    The CRM default, and where it falls short

    Every CRM ships with a ready-made pipeline. HubSpot's, described in its official documentation, has seven stages: Appointment scheduled (20%), Qualified to buy (40%), Presentation scheduled (60%), Decision maker bought-in (80%), Contract sent (90%), then Closed won (100%) and Closed lost (0%).

    That's a reasonable starting point. But those percentages are generic. They don't come from your deals. A stage probability is only worth something if it's recalculated from your own history. If only 25% of your "contract sent" deals actually close, showing 90% wrecks the whole forecast.

    A six-stage pipeline template

    Here's a structure that works for a B2B sale in the low thousands to tens of thousands of dollars, with a cycle of a few weeks to a few months.

    StageExit criterion (verifiable fact)Typical next step
    1. First conversationThe prospect replied and agreed to talkPrepare the discovery call
    2. DiscoveryProblem, impact and context logged in the CRMConfirm who decides
    3. QualifiedDecision maker identified, budget range discussed, timeline knownPropose a demo or workshop
    4. Solution fitThe decision maker confirms the solution addresses the needSend the proposal
    5. Proposal sentPriced proposal received, decision date agreedFollow up on the agreed date
    6. NegotiationOpen issues listed, procurement or legal contact identifiedClear the last blockers
    Won / LostSigned or explicit no, with a reasonOnboarding or loss review

    Every criterion describes something the buyer did or said. Not something the rep sent.

    Fewer stages beat more stages

    Past seven open stages, two things break. Reps no longer know where a deal belongs. And stages get too thin to produce a stable conversion rate. If two stages always convert at the same rate, merge them.

    Write exit criteria, not stage names

    This is the decision that separates a useful pipeline from a decorative one. A stage name isn't enough. You need a written definition every rep applies the same way.

    The yes-or-no test

    For each stage, write a question that can only be answered yes or no. If the answer is yes, the deal moves. If not, it stays.

    Weak "Stage 3: Qualified. Prospect seems interested and has good potential."

    Strong "Stage 3: Qualified. Have we spoken with the decision maker, or someone who gets us to them? Do we know a budget range? Has the prospect given us a timeline? Three yeses to advance."

    The weak version lets every rep decide based on mood. The strong version can be checked by reading the deal record.

    Make the criteria mandatory in your CRM

    A criterion buried in a doc nobody reads does nothing. Your CRM can enforce it. HubSpot, for example, lets you show specific properties when a deal moves to a stage and make them required: until the value is filled in, the record can't be updated.

    Three fields to require first:

    • Next step and its date: at every stage, no exceptions.
    • Decision maker identified: from "Qualified" onward.
    • Loss reason: mandatory to close a deal as lost.

    The loss reason is the one teams skip most. It's also the one that will tell you, six months from now, whether you lose on price, on timing or to the status quo.

    Exit criteria and qualification frameworks

    The exit criteria for "Qualified" overlap with the classic qualification frameworks: need, budget, authority, timeline. You don't need to import a full framework with twenty questions. Pick the three or four facts without which a deal never closes for you, and make them the exit criterion.

    A stage with no exit criterion isn't a stage. It's an opinion with a percentage attached.

    Feed the top of the pipeline with better timing

    A healthy pipeline starts with healthy inputs. Most teams fix an empty pipeline by cranking up prospecting volume. They get more deals in stage one, and the same number of closed deals.

    The 95% problem

    LinkedIn's B2B Institute puts it bluntly with its 95-5 rule: 95% of your potential buyers aren't ready to buy today. They will be, later. A rep who opens deals on that 95% fills the pipeline with opportunities that won't move for months.

    Those deals aren't lost. They're misfiled. They belong on an account watch list, not in the pipeline.

    Open a deal when something changes

    Deals that move tend to start with an event at the account: a funding round, a hiring push for the role your offer supports, a new VP of Sales, a comment on one of your team's posts. Our guide to B2B buying signals breaks each one down.

    Here's what that looks like for a recruiting firm targeting scale-ups:

    Weak "Open a deal for every Head of People we message this month."

    Strong "Open a deal when a target company posts three or more roles in two weeks and the Head of People has replied to our first message."

    In Reakly, intent signal detection surfaces those accounts as soon as the trigger appears. The deal gets created when the conversation starts, not before.

    How many deals do you need at the top?

    Work backward. If you want 10 closed deals per quarter and 20% of qualified deals close, you need 50 qualified deals. If one in two first conversations becomes qualified, you need 100 at stage one. This math only works if the rates come from your own history, which is why exit criteria matter so much.

    Run a weekly pipeline review

    A pipeline decays on its own. Close dates slip. Dead deals stay open. A weekly review is the only way to keep the pipeline honest.

    Rep time is scarce

    The review can't become one more meeting. The same Salesforce report shows reps spend only 40% of their week actually selling. The rest goes to prospecting, data entry, quotes and training. Sales professionals spend 16% of their time on preparation and planning on average. A good review is part of that, as long as it stays short.

    The 30-minute format

    I recommend a fixed format, same day every week:

    1. Deals with no dated next step (5 min): each one gets a next step or gets closed.
    2. Stuck deals (10 min): every deal that has sat in its stage longer than the normal duration for that stage.
    3. Deals past their close date (5 min): a new date backed by a fact, or close the deal.
    4. The three biggest deals of the month (10 min): what could kill them, and what are we doing this week?

    Deals that are moving normally don't get discussed. They don't need a meeting.

    The right questions to ask

    A pipeline review isn't an interrogation. It's the same factual questions every week.

    SituationQuestion to askPossible decision
    Deal hasn't moved in 30 daysWhat was the buyer's last touch, and when?Follow up with a new angle, or close
    Close date pushed twiceWhat changed on the buyer's side?Move back a stage, or close
    No decision maker at stage 4Who signs, and have we met them?Stay in stage 3
    Proposal sent, no decision dateWhen did the buyer say they'd decide?Call to set the date

    A good pipeline review doesn't make the number bigger. It makes it believable.

    The pipeline metrics that matter

    Total pipeline value is the most-watched metric and the least useful. It tells you what you hope for, not what you'll close.

    Four metrics are enough

    • Stage conversion rate: the share of deals that move to the next stage. This shows where you lose.
    • Average time in stage: how long won deals spent in each stage. It's your benchmark for spotting stuck deals.
    • Weighted pipeline: each deal's amount multiplied by its stage probability. HubSpot uses it by default in its forecast tool.
    • Win rate and loss reasons: the share of qualified deals that close, and why the others don't.

    Calculate your own stage probabilities

    CRM default percentages don't match your business. Once you have 30 to 50 closed deals, recalculate them: for each stage, divide the number of won deals by the number of deals that reached that stage.

    StageDeals that reached the stageWonActual probability
    First conversation1201210%
    Qualified481225%
    Proposal sent241250%
    Negotiation151280%

    These numbers illustrate the calculation. They're not a benchmark. Yours will be different, and that's the point.

    Data quality comes first

    These metrics are only as good as the data behind them. In the same Salesforce report, 46% of sales pros using AI agents say data quality issues hurt their sales. A stage left stale, a close date never updated, a made-up amount: every error flows straight into the forecast. A well-kept prospecting CRM beats a fancy dashboard fed by empty records.

    The mistakes that distort a pipeline

    The same mistakes show up on almost every team.

    • Stages defined by rep activity: "demo done" says nothing about buyer intent.
    • Too many deals in stage one: contacts who never replied inflate the total.
    • Nothing ever closed as lost: the pipeline becomes a graveyard nobody dares to clean.
    • Optimistic close dates: pushed every month, never challenged.
    • CRM default probabilities: never recalculated from your own deals.
    • No next step: a deal with no dated next step is a deal nobody is working.
    • Pipeline updated the night before the review: a sign the tool serves the manager, not the rep.

    Close more deals as lost

    It feels backward, but it's the most effective fix. A deal closed as lost with a reason is information. A deal open for four months is noise. Reps hesitate because the number drops. Managers have to show the opposite gets rewarded: a smaller, more accurate pipeline.

    A deal closed too early isn't gone forever. The account goes back on the watch list. If a new signal shows up in six months, you open a new deal with fresh context.

    Move deals backward without shame

    Deals can regress. If the decision maker leaves, the deal goes back to discovery. If budget gets frozen, it leaves the pipeline. A pipeline where deals only ever move forward is a pipeline where nobody updates stages.

    Connect the pipeline to multichannel outreach

    In outbound, the line between prospecting and pipeline is blurry. The first conversation might start on LinkedIn, over email or in a comment thread. If each channel lives in a different tool, the history disappears the moment the deal is created.

    Keep context in one place

    The rep who opens a deal should see why the account surfaced, what messages went out and what the prospect said. Without that, the discovery call starts from zero.

    Reakly's CRM keeps signals, LinkedIn messages and emails, replies and deal stage on the same record. Pipeline and prospecting share one source of truth.

    Follow up on a reason, not a calendar

    Most deals stuck at "proposal sent" are waiting for a relevant follow-up. Not "Just checking in on the proposal". A follow-up that brings something new or builds on a change at the account. Our guide to the sales follow-up message has examples by situation.

    Weak "Hi, just circling back on the proposal I sent over."

    Strong "Hi Claire, you mentioned the decision would come after your board meeting on the 15th. I noticed you just opened two sales roles. Want me to adjust the proposal to cover onboarding them?"

    The pipeline tells you which deals deserve attention. Signals tell you when to give it.

    Conclusion: judge a pipeline by the decisions it lets you make

    Stick to five to seven open stages. Define each stage by a verifiable fact on the buyer's side. Make the key fields mandatory in your CRM. Only open a deal after a real conversation. Review the pipeline every week, in thirty minutes, focused on stuck deals. Recalculate stage probabilities from your own history. Close deals as lost without hesitation, and log the reason.

    An inflated pipeline feels great on Monday and disappoints at quarter end. An honest pipeline sometimes worries you on Monday, but it never surprises you.


    If you want a pipeline fed by deals that started at the right moment rather than by sheer volume, Reakly detects buying signals on your target accounts, runs LinkedIn and email sequences tailored to each segment, and keeps every conversation and deal stage in the same CRM.

    Frequently asked questions

    • Most B2B sales fit into five to seven open stages, for example: first conversation, discovery, qualified, solution fit, proposal sent, negotiation, then closed won or closed lost. The stage name matters less than its exit criterion: a verifiable buyer-side fact that lets the deal move forward.

    • A sales funnel measures aggregate volumes at each step to show where prospects drop off. A sales pipeline lists individual deals, each with an owner, a value, a close date and a next step. The funnel is for analysis; the pipeline is for weekly action.

    • Use a fixed 30-minute weekly format focused on exceptions: deals with no dated next step, deals stuck in a stage too long, deals past their close date and the three biggest deals of the month. Ask factual questions and make a decision each time: follow up, move the deal back or close it.

    • Once you have 30 to 50 closed deals, divide the number of won deals by the number of deals that reached each stage. Replace your CRM's default percentages with those values and recalculate them every quarter.

    • As soon as there is no credible next step: the close date has slipped repeatedly without a reason, the decision maker has gone silent or the budget is frozen. Closing with a loss reason keeps the pipeline accurate, and the account can go back on a watch list until a new signal appears.

    Put this method into practice with Reakly.

    Find the right prospects, reach out on LinkedIn and by email, and keep every reply in one place.

    See pricing

    7-day free trial · Cancel anytime