Ask most small businesses what’s in their pipeline and you’ll get a confident answer followed by a much less confident search through an inbox. The deals exist. The pipeline, as a thing you can look at and reason about, usually doesn’t.
That matters more than it sounds. Without one you can’t tell whether a quiet month is a blip or the start of something, you can’t see which stage is leaking, and you end up doing the two most expensive things in sales: chasing dead deals and forgetting live ones.
A pipeline isn’t software. It’s an agreement about what the stages mean and a habit of keeping them honest. The software just makes it visible.
Define stages around what the buyer does
The most common mistake is building stages around your own activity — “called”, “emailed”, “followed up”. Those describe effort, not progress. You can call someone five times and be no closer to a sale.
Stages should describe what the buyer has done, because that’s the only reliable evidence of movement. A deal advances when something changes on their side, not when you’ve done another task.
- Qualified — they have the problem, the budget and the authority, and you’ve confirmed it rather than assumed it
- Discovery done — you understand their situation well enough to propose something specific
- Proposal sent — a real quote or scope is with them
- Negotiation — they’re engaging with terms, which means they’re seriously considering yes
- Won or lost — and lost should always carry a reason
Five or six stages is plenty. More than that and people stop updating them, which produces a pipeline that’s worse than none because it’s wrong in a way that looks authoritative.
A stage should be something the buyer did, not something you did. Otherwise you’re measuring how busy you are and calling it progress.
Write down what each stage means
This sounds bureaucratic and it prevents the single biggest source of pipeline fiction. If “qualified” means one thing to you and another to a colleague, your pipeline total is meaningless.
Define the exit criteria for each stage in a sentence: what must be true before a deal can move forward. “Qualified means we’ve confirmed a budget range and spoken to someone who can sign.” Now there’s nothing to argue about, and nobody can quietly promote a hopeful conversation into a forecast.
Keeping it clean is most of the work
Pipelines rot in a predictable way. Deals get added and never removed. Someone goes quiet and their deal just sits there at 60% for four months. Eventually the total becomes a number nobody believes, and people stop using it.
The fix is unglamorous: a standing rule that every deal has a next step with a date, and a weekly pass to clear anything that doesn’t. If you can’t say what happens next and when, it isn’t a live deal — it’s a hope with a value attached.
Be willing to mark things lost. There’s a strong pull towards leaving deals open because closing them lost feels like admitting failure, but a pipeline padded with zombies makes every forecast wrong and hides the fact that you need more new conversations.
The age test
Look at how long each deal has sat in its current stage compared with your normal. A deal that usually clears proposal in two weeks and has been there for nine is not slow — it’s almost certainly lost, and treating it as live means you’re under-prospecting without realising.
Every comment and DM in one inbox.
Reply across all your platforms from a single unified inbox. Free for 30 days, no card required.
Unify your inbox →Forecast without fiction
Two approaches, and the difference matters.
Weighted forecasting multiplies each deal by the historical close rate of its stage. If deals at proposal close 40% of the time, a ₹100,000 proposal counts as ₹40,000. Across enough deals this is reasonably accurate, and it stops a single large deal from dominating the picture.
Commit forecasting asks the person closest to the deal a blunter question: will this close this month, yes or no? It’s less mathematically tidy and often more accurate, because a salesperson who has just spoken to the buyer knows things no stage percentage captures.
Use both. When they disagree sharply, that disagreement is the signal — either the stage percentages are stale or someone’s optimism needs testing.
The one thing that ruins both is using guessed percentages. If “proposal” is set to 60% because it feels about right rather than because that’s what your last hundred proposals actually did, you’re forecasting your own mood.
Find where deals actually get stuck
The value of a pipeline isn’t the total. It’s the conversion rate between each pair of stages, because that’s what tells you where to fix things.
- Stage-to-stage conversion — where the biggest drop happens
- Average time in each stage, compared against deals that eventually won
- Average deal size, and whether it varies by source
- Win rate overall and by lead source
- Pipeline coverage — total open value against your target for the period
Coverage is the early warning system. If you need ₹10 lakh this quarter and close roughly a quarter of what’s open, you need around ₹40 lakh in the pipeline. Falling below that means the problem is already six weeks old and living at the top of the funnel — which is a marketing problem, not a closing problem. Our guide to the marketing funnel covers the stages above this one.
Read the losses
Loss reasons are the most underused data in most businesses. Record them as a fixed set of options rather than free text — price, timing, chose a competitor, no decision, lost contact — and review them quarterly.
The patterns are usually blunt. Lots of “no decision” means you’re not creating urgency or you’re talking to people without authority. Lots of “price” often means you’re selling to the wrong segment rather than that you’re too expensive. Lots of “lost contact” means your follow-up process has a hole in it.
Follow up better than everyone else
Most deals are lost to silence rather than to competitors. Someone gets busy, the proposal drifts down the inbox, and nobody ever says no — the deal just quietly stops.
The remedy is a follow-up cadence you actually stick to, with something useful in most touches. A relevant case study, an answer to a question they raised, a note that something they mentioned has changed. Pure “just checking in” messages are easy to ignore because they contain nothing.
And always agree the next step before the current conversation ends. “I’ll send the proposal today, shall we speak Thursday?” beats “I’ll send it over and let you know”, because the second one puts the entire burden of momentum on the person with least incentive to carry it. There’s more on this in social selling without being pushy.
When one pipeline isn’t enough
A single pipeline works until you’re selling genuinely different things, at which point the averages start lying to you.
The signal is that your stages stop fitting. If a ₹5,000 self-serve sale closes in three days and a ₹5 lakh contract takes four months, forcing both through the same five stages produces a conversion rate that describes neither. Your average time-in-stage becomes a number that no real deal has ever matched.
Split when the buying process is genuinely different, not when the product is. Two products sold the same way to the same kind of buyer belong in one pipeline. The same product sold self-serve to individuals and via a procurement process to large organisations belongs in two, because the stages, the timelines and the people involved have almost nothing in common.
Resist splitting further than that. Every extra pipeline is another set of stage definitions to maintain and another place for deals to be filed inconsistently. Two or three is manageable; six means nobody has an overview of anything.
Where the pipeline should live
A spreadsheet is a legitimate starting point, and better than nothing by a wide margin. It stops working once more than one person touches it, or once you want history — when a deal moved stage, what it was worth last month, why it was lost.
The moment you want to answer “how has our win rate changed since we raised prices?”, you need something that records changes over time rather than just current state. That’s the point at which a proper CRM earns its keep — not because spreadsheets are bad, but because they only ever hold the present tense.
A weekly routine that keeps it honest
Thirty minutes, once a week, and the pipeline stays usable.
- Every open deal has a next step with a date — no exceptions
- Anything untouched beyond your normal cycle length gets a decision: advance, or mark lost
- New deals entered with a realistic value rather than a hopeful one
- Check coverage against the target for the period
- Read last week’s loss reasons and look for a repeat
None of that is sophisticated. But a pipeline that’s slightly simplified and completely honest will beat an elaborate one that everybody quietly knows is fiction — because the whole point is to be able to trust the number when you’re deciding what to do next.