Designing a Sales Pipeline That Fits Your Business
Every CRM ships with default deal stages, and most teams keep them forever. Clientverse defaults to Lead, Qualified, Proposal sent, Won, and Lost, which is a reasonable starting point and a poor finishing one.
A pipeline is a model of how you actually sell. If the model is wrong, every number you derive from it is wrong too.
The Test for a Good Stage
A stage is useful when it passes three tests:
It is observable. You can tell from the outside whether a deal is in it. “Proposal sent” passes - either you sent one or you did not. “Interested” fails, because it describes a feeling you inferred.
It is exclusive. A deal is in exactly one stage at a time. If two stages could both be true, they are not stages, they are attributes.
It changes what happens next. If your next action is identical in two stages, the distinction is decorative. Merge them.
Most bloated pipelines fail the third test. Eleven stages where six would do means every deal sits ambiguously between two of them and nobody updates anything.
Build It From Your Actual Process
Take your last ten closed deals, won and lost, and write down what actually happened in each, in order. Ignore what the sales playbook says. You are looking for the events that recurred.
You will typically find three to six real transitions. Something like:
- Someone got in touch or you reached out.
- You had a conversation that established whether this was real.
- You scoped what they needed.
- You sent numbers.
- They decided.
That is your pipeline. Name the stages after the event that moves a deal into them, not the feeling associated with it. “Proposal sent” beats “Negotiation” because it has a date attached.
Qualification Is the Stage That Earns Its Keep
The gap between “someone emailed us” and “this is a real opportunity” is where most wasted effort lives. A distinct qualified stage forces the question and gives you the single most useful ratio in your pipeline: how many enquiries become real deals.
Write down what qualified means for you - a budget signal, an identified decision maker, a stated timeline, whatever genuinely predicts a close - and apply it consistently. Consistency matters more than which criteria you picked.
Losing Is Part of the Model
A pipeline with no exit but “Won” is a list of things nobody wants to admit are dead. You need a Lost stage and a habit of using it. We go into when to use it in when to mark a deal lost.
If you can, record why. Even three or four reasons - price, timing, competitor, no decision - turn your lost deals from a graveyard into a report about your pricing and positioning.
Expected Value and Close Date Are Not Optional
A stage on its own tells you order, not size. The two fields that make a pipeline forecastable are the expected value and the expected close date. If half your deals have neither, your pipeline is a to do list rather than a forecast.
Put a number in even when it is a guess. A rough figure you correct later is more useful than an empty field, and forecasting from stage conversion rates is covered in forecasting revenue from your pipeline.
Review It Once a Year, Not Once a Quarter
Pipelines should be stable. Every change breaks comparability with your historical data, so a pipeline you restructure every quarter can never tell you whether anything improved.
Design it carefully, use it for a year, then look at whether any stage is a formality every deal passes through in a day - and cut it if so.
Need Professional Help?
Buy managed hosting, white label licensing, or technical support online - or ask for a quote on custom development and data migration.
Explore Services