Take control of your sales pipeline in 30 minutes a week

A sales pipeline is every open deal you're working on, sorted by the step it has reached. Managing it means keeping that list honest: the right deals in it, each one moving, and the dead ones out. When the list is honest, its total tells you what's likely to close and how much you still need to find.

People mix up three words here. They describe different things:

You manage the pipeline and measure the funnel. HubSpot's sales blog makes a fair warning about the funnel picture: it suggests prospects always thin out at a steady rate, which leads people to believe they need a fixed multiple at the top. Your own numbers are a better guide, and the coverage section below shows how to get them.

There's no single right list. Salesforce's guide names seven stages: prospecting, lead qualification, demo or meeting, proposal, negotiation and commitment, opportunity won, and post-purchase. That's a good map of a big team's sale.

A small business usually needs fewer, and the stages should describe what the buyer agreed to rather than what you did. A deal moves only when something is true on the buyer's side. Here's a set of six that fits most people who sell one to one:

Two rules make the stages work. Every deal has one owner, because shared deals are the ones nobody chases. And every stage change is dated, so you can see how long deals sit where.

If your deals live in chats and your head today, you can build a first pipeline in a few hours. Work through it in this order:

1. Write down every person who's waiting on you for something: a quote, a callback, a second meeting.
2. Put each one at the stage the buyer has reached, using the exit rules above, not where you hope they are.
3. Give each a rough value. A range is fine; a blank isn't.
4. Add the next step and its date. If you can't think of one, the deal probably belongs in lost.
5. Note where each customer came from and which product or service they want, so in three months the data tells you which source brings deals that close.

You don't need a CRM to start. A spreadsheet with one row per deal does all five, and the sales pipeline template is set up that way, with a column for the date you last spoke. When the sheet gets too long to sort every day, a CRM does the sorting and puts each follow-up on its day.

Enough that your win rate turns it into the target. This ratio is called pipeline coverage, and the amount you need is 1 divided by your win rate by value. At a 25% win rate you need four times the target. At 33% you need about three times. Salesloft's guide says the same thing from the other side: coverage targets depend on your own win rate, so set your own benchmark rather than borrowing someone else's.

The sums below use made-up numbers. Say next month's target is S$40,000 and your win rate by value over the last 12 weeks is 30%.

• Coverage needed: 1 / 30% = 3.33 times.
• Pipeline needed: S$40,000 x 3.33 = S$133,333 due to close next month.
• Open pipeline due next month today: S$90,000.
• The gap: S$133,333 - S$90,000 = S$43,333.
• At an average deal of S$5,000, that's 8.7, so 9 new deals to find.
• Over the 4.33 weeks of a month, that's about 2 new deals a week.

Do this sum for next month, not this one. If your sales cycle is six weeks, this month is mostly decided already, and next month's goal is the one you can still move. The win rate calculator gives you the rate to start from.

Hold one review a week on the same day, with every open deal on screen sorted by stage. Talk only about what changed since last time.

Spend the most time on the stuck deals. That's where forecasts go wrong. Once a month, also work out your win rate and each stage's conversion again from the last 12 weeks, so the coverage sum uses fresh numbers. For each one, the owner names the next step and its date, or the deal comes out. If you sell on your own, run the same five blocks once a week. It's quicker alone.

A pipeline full of stale deals feels safe and closes badly. These rules keep it honest:

• Every open deal has a next step and a date. A deal without one moves out until the buyer comes back.
• When a decision date slips, write down why. After two slips in a row the deal drops a stage.
• Work out how long deals usually sit at each stage, and look at anything that has sat there twice as long.
• Log every lost deal with the reason in the buyer's words. They tell you more about your pitch than the wins do.
• A deal moves on something the buyer did. A good feeling after a call doesn't count.

You don't need a dashboard of 20 numbers to track a pipeline. These seven tell you almost everything you need to understand it:

Sales velocity rolls four of these into one number, which helps when you compare months. The sales velocity calculator does the sum.

Weight each deal by your own history, not the percentage a CRM came with. Over the last 12 weeks, work out what share of deals at each stage went on to win, then multiply each open deal by its stage's share.

With made-up numbers: if 40% of your deals at the quote stage went on to win, and three quotes are out at S$6,000 each, that stage is worth 3 x S$6,000 x 40% = S$7,200. Add every stage the same way and you have a forecast built from what your pipeline has actually done.

Keep a second number beside it: the deals you'd bet on, called commit. When the two are far apart, one of them is wrong, and the stuck block of the weekly review is where you find out which.

• Keeping dead deals because the total looks healthy. Five times coverage made of stale deals closes like one.
• Naming stages after your activity ("called twice") instead of the buyer's.
• Reviewing every deal every week, when only the changes and the stuck ones need talking about.
• Only looking at this month. By the time it's short, it's too late to fill.
• Letting deals fade away instead of marking them lost. A fade teaches you nothing.

Palby CRM, which we make, puts the pipeline on a board or a grid, with your own stages. A new pipeline can start from one of eight stage sets by trade, from general sales to tuition and home services. Each contact can have several deals, each with its own close date, and they add up to a weighted forecast. Cards that sit too long in one stage get a flag, which is the stuck block of the review done for you. The pipeline and deals page shows the board.

• HubSpot, Sales pipeline guide, read 2 November 2026
• Salesforce, What are the stages of a sales pipeline?, read 2 November 2026
• Salesloft, Sales pipeline management, read 2 November 2026