What it means
A sales pipeline is the operational view of every open opportunity in your business, each represented as a "deal" or "opportunity" record with a stage, a value, an owner, and a close date. Pipelines are usually displayed as a kanban board (columns = stages) but live underneath as a structured set of records you can query, forecast against, and report on.
The pipeline is the difference between "we have a lot of leads" and "we have $1.2M in active opportunity, $480k weighted, with $180k closing this quarter". The former is a vibe; the latter is a forecast.
Three things make a pipeline actually useful (versus a theatrical kanban):
- Stages are defined by exit criteria, not by feelings. "Discovery completed" means specific things happened, not "the rep thinks it is going well".
- Every deal has a dollar value and an expected close date. Without these, forecasting is impossible. Reps update both weekly.
- Stuck deals get killed or unstuck. A deal that has not moved in 30 days is doing damage by inflating pipeline totals. Pipeline hygiene means closing the dead ones aggressively.
The typical stages
Most B2B sales pipelines use a variant of:
- Lead. A real person we can contact. Has a name, email, and a reason to be in the CRM. Exit criteria: outbound made, response received OR auto-disqualify.
- Qualified. Passed initial fit and intent screening, typically a qualification call. Exit criteria: BANT (or similar framework) confirmed.
- Discovery. Deep understanding of the prospect's situation, pain, decision process, and timeline. Exit criteria: discovery notes documented; pain confirmed by the prospect.
- Proposal / Demo. A solution-specific proposal, demo, or scope-of-work has been delivered. Exit criteria: prospect has reviewed and engaged on the proposal.
- Negotiation. Talking terms, price, contract clauses. Exit criteria: verbal agreement; legal review started.
- Closed-won / Closed-lost. The terminal stages. Won = contract signed, first invoice issued. Lost = the prospect explicitly said no, OR went silent for 30+ days after the last touch.
For high-velocity SaaS or ecommerce, this can collapse to three stages (Interested → Active → Closed). For complex enterprise sales, it expands to 8-10 stages with explicit sub-stages for procurement, security review, and legal. The test is always: do the stages reflect how YOUR deals actually move?
Design principles
Five rules that separate a useful pipeline from a vanity board:
- Exit criteria, not feelings. Every stage must have a one-sentence "to leave this stage, X has to have happened" rule. Without it, reps slide deals based on optimism.
- One pipeline per product, not per rep. Reps have views; the team has pipelines. Per-rep pipelines hide bad deals.
- Required fields enforce hygiene. Value, close date, next step, and stage must be filled before a deal can be saved. Anything else can be optional.
- Stale-deal alerts. A deal in one stage longer than its stage's median age gets flagged automatically. Most CRMs do not do this; build it yourself if needed.
- Probability per stage is set by data, not wishful thinking. Look at the last 12 months. What percentage of deals that entered "proposal" actually closed? THAT is your proposal probability. Not 50% because 50% feels right.
Forecasting from a weighted pipeline
The simplest forecast: weighted pipeline = sum of (deal value × stage probability) across every open deal. If you have:
- $10,000 in Qualified (20% probability) = $2,000
- $25,000 in Discovery (30%) = $7,500
- $50,000 in Proposal (50%) = $25,000
- $30,000 in Negotiation (75%) = $22,500
Total weighted pipeline: $57,000. That is the expected revenue from your current open pipeline, regardless of close dates.
For a quarterly forecast, filter to deals with a close date inside the quarter and apply the same formula. Then add a "commit + best-case" view where each rep marks deals as high/medium/low confidence; the commit is the floor, the best case is the ceiling, the weighted number is the central forecast.
Calibration matters more than the formula. If your "proposal" stage win rate is actually 35% (not 50%) historically, your forecasts are systematically overstated by 30%. Pull last year's data and check.
Why it matters
Three reasons.
Cash planning. Without a forecast you are guessing about hiring, runway, and ad spend. Pipeline-based forecasting is the cheapest, most accurate revenue model for sales-led businesses.
Coaching. A manager looking at a pipeline can see "this rep has lots of activity but no deals past discovery; they need coaching on proposal delivery." That kind of diagnosis is impossible without the structure.
Hygiene-as-discipline. A clean pipeline forces a weekly "what's real, what's not" conversation. Teams that run that conversation rigorously close at 2-3x the rate of teams that do not, even with worse leads, because they spend their time on real opportunities.
Real-world examples
- SaaS startup pipeline. 6 stages, avg cycle 28 days, avg deal $9k ARR. The team holds a weekly forecast call where every deal past discovery gets a 30-second review and a commit/best-case rating.
- Agency pipeline. 4 stages (Lead → Discovery → Proposal → Won). Deals carry "service line" tags so the agency can forecast revenue per practice. Stale deals 14+ days idle get auto-archived.
- Real-estate pipeline. Stages are property touchpoints: Inquiry → Viewing → Offer → Under contract → Closed. Each property is a deal; commission is the deal value.
- Coach selling on DMs. 3 stages: DM conversation → Booked call → Paid client. Most deals close within 5 days. Pipeline is small but hygiene is everything No deal sits longer than 7 days.
- Enterprise B2B pipeline. 9 stages with explicit sub-stages for security review, legal, and procurement. Cycles run 6-9 months. Forecasting uses cohort math, not naive weighted pipeline, because the distribution is too long-tailed.
Common mistakes
- Stages that mean different things to different reps. Without explicit exit criteria, "discovery" means whatever each rep wants it to mean. Forecasts become noise.
- Inflating pipeline to look productive. Reps push deals into "proposal" before any proposal has been sent because pipeline coverage is a comp lever. Stage definitions plus weekly hygiene catch this.
- Probabilities pulled from thin air. "Let's call discovery 30%" is fine as a starting point but if you never check the actual historical conversion, the forecast lies to you forever.
- No archive-stuck-deals rule. Without an auto-archive (or weekly hygiene), the pipeline grows indefinitely with deals that died in March. Forecast accuracy collapses.
- Trying to forecast the funnel and pipeline with the same metrics. They are different. The funnel feeds the pipeline; the pipeline is named, dollared, and dated. Treat them differently.
Related concepts
- Conversion funnel: the upstream view; feeds the pipeline.
- Lead scoring: what you use to decide which leads earn pipeline entry.
- Drip campaign: the automation that warms leads before pipeline entry.
- Omnichannel CRM The system that captures pipeline activity from every channel (Telegram, email, calls, live chat).
- Cold outreach: one of the primary inflow sources of new pipeline.
- AI agent: increasingly handles early-stage qualification before deals enter the pipeline at all.
How CRM Solid handles it
CRM Solid ships with a flexible kanban pipeline view: stages and exit criteria are configurable per workspace, required fields enforce hygiene, and stale-deal alerts fire automatically. Pipeline activity is unified across every channel: a Telegram DM, a live-chat message, and an email update the same deal record. Weighted pipeline and revenue forecasting are built in, with per-rep and per-segment views. Calibration is on you, but the data is at your fingertips.