Almost every CRM on the market ships with a preconfigured sales pipeline out of the box, with stages like "New lead", "Contacted", "Proposal sent", "Negotiation" and "Won/Lost". It's a reasonable starting point, and it's also, almost always, a mistake to leave it as is. That generic pipeline is designed for "some business" in the abstract, and that's exactly why it doesn't reflect how any specific business, yours included, actually sells. When the pipeline doesn't match your real sales process, the team starts filling it in however, the data stops being reliable, and the CRM turns into a nice-looking database nobody actually uses to make decisions.
Why the generic template doesn't work
The underlying problem is that every business sells differently, with different milestones that mark whether an opportunity is moving forward or stalling. A services agency that closes contracts after a meeting and a proposal has a short sales cycle with very different stages from a company selling industrial machinery, where months can pass between first contact and signature, with several technical visits and approval from a purchasing committee along the way. If both use the same five generic stages, in one case the stages fall short (they miss the real nuances of the process) and in the other they run long (stages that in practice never get used separately).
How to define the right stages for your business
The starting point isn't opening the CRM, it's looking back at your last 20 or 30 closed deals (both won and lost) and asking: what steps did each one actually go through? Not the steps that "should" happen according to some manual, the ones that actually happened. A pattern almost always emerges with clear, recognisable milestones: the moment the customer confirms they have a real problem they want solved, the moment they see a concrete proposal with a price, the moment there's a specific objection to overcome (budget, timing, comparison with a competitor), and the moment there's a decision, whether yes or no.
Those milestones, not the pretty wording of a template, are your pipeline's stages. A stage in your CRM should always represent a real change in the probability of closing the sale, not just the passage of time. If you have a "follow-up" stage where opportunities can sit for weeks with nothing changing about whether they'll buy or not, that stage isn't contributing useful information, it's just there to have something to show in the weekly meeting.
A practical guide: if your sales cycle takes less than two weeks, 4 stages is usually plenty. If it takes several months and passes through several decision-makers at the client's company, you'll need more stages (maybe 6 or 8) to reflect that complexity, but never so many that the team has to think for two minutes to decide which one a given opportunity belongs in.
What information every opportunity should carry
A pipeline with the right stages is only half the job. The other half is what data each opportunity holds within those stages, because that data is what lets you draw useful conclusions later. At a minimum, every opportunity should have: the estimated deal value (even a rough range works, you don't need perfect precision from day one), the estimated close date (which forces whoever owns the opportunity to commit to a forecast, instead of leaving it floating with no date forever), the lead source (where it came from: website, referral, trade show, cold call), and a short, current note on what the next concrete step is and who needs to take it.
That last field, the next step, is what separates a CRM that works from one that doesn't more than anything else. An opportunity with no clear next step and no date is, in practice, stalled even if nobody consciously decided that. When a salesperson reviews their list of opportunities and each one says "next step: call Thursday to resolve the delivery-time question", they know exactly what to do every morning. When the note just says "follow up", it says nothing.
How to spot bottlenecks by looking at the pipeline
Once the pipeline genuinely reflects your process, it becomes a diagnostic tool, not just a tracking one. Looking at how many opportunities pile up in each stage and how long they sit there on average is the most direct way to find where your sales process gets stuck. If you have 40 opportunities in "proposal sent" and only 5 in "negotiation", the problem isn't closing sales, it's that the proposals aren't convincing or there isn't enough follow-up after sending them. If, on the other hand, few opportunities are entering the top of the pipeline but the ones that do move forward well, the problem isn't your sales team, it's lead generation, and the effort belongs in marketing, not in training salespeople harder.
Another useful signal is the average time an opportunity spends in each stage compared with the ones that actually end up won. If won deals typically spend 5 days in negotiation and you see opportunities that have been sitting there for 40 days with no movement, those 40 are almost certainly deals that aren't going to close and are eating up the team's attention for nothing. Cleaning those "zombie" opportunities out of the pipeline (honestly marking them as lost) isn't an administrative chore, it's what makes the pipeline's conversion numbers mean something real instead of being inflated with dead deals.
Reviewing it isn't a one-time task
The pipeline doesn't get defined once and forgotten. As your business changes (new products, new types of customer, a sales cycle that shortens or stretches), it's worth reviewing every six months or a year whether the stages still represent reality well, whether there's a stage almost nobody uses anymore, or conversely, a real step in the process that has no stage of its own and is therefore invisible in your reports.
An example of a real restructuring
A home renovation company set up its CRM using the generic five-stage pipeline that came with the tool by default, and had spent months unable to explain why it only closed one out of every eight opportunities that came in. On reviewing their recent sales, they discovered that the real decisive step wasn't captured by any stage at all: it was the in-home technical visit to measure the space and put together an accurate quote, a step that in the generic pipeline got lumped into "proposal sent" together with quotes made without a visit, which almost never converted. By splitting those two situations into separate stages ("quote without visit" and "quote after technical visit"), the team could immediately see that technical visits converted almost three times better, and they reshaped the sales process to offer the visit earlier in the conversation, instead of leaving it as an optional step at the end.
Frequently asked questions
How many stages should my sales pipeline have?
There's no single correct number for every business. What matters is that each stage represents a real change in the probability of closing, not just the passage of time. Most businesses work well with between 4 and 7 stages; more than that tends to create confusion about where to classify each opportunity.
What should I do with opportunities that haven't moved in months?
Mark them as lost if there's genuinely no sign they'll move forward, even if it's uncomfortable to admit. Keeping them open "just in case" only artificially inflates your pipeline and distorts your real sales forecasts.
Should the pipeline be the same for every type of product or service I sell?
Not necessarily. If you sell products or services with very different sales cycles (for example, a quick low-cost sale and a large custom project), it usually works better to have separate pipelines in the CRM for each type, rather than forcing them into a single process that doesn't fit either one.