FAQ: B2B Sales Management: Pipeline, CRM, KPIs and Forecasting

Because the pipeline shows a list of deals, not the probability that any of them will close, and those two things drift apart wherever nobody has defined on what basis a deal drops off the list. It is usually inflated by several things at once: stages assigned out of optimism, when the client only said the proposal looks interesting, deals with no concrete next step and date, values rounded up, and a closing date pushed back for the third time already.

Verification takes fifteen minutes and boils down to one question asked about every deal: what did the client do recently, not the salesperson. If within the last two weeks they replied, scheduled a meeting or pulled someone from their side into the conversation, the deal is really alive. If all the activity is on your side, it is not a sales opportunity, just a hope, which is better moved to a reactivation list and no longer counted in the forecast.

Companies doing this exercise for the first time usually lose half their pipeline, which at first glance looks like a disaster, and is actually the moment from which the numbers start to mean anything and can support decisions about hiring or taking on another project.

I describe how to build and run a pipeline from scratch in my answer to the question how to run a pipeline.

Don’t measure everything. Measure what lets you make decisions.

In the B2B service companies I work with, four metrics are enough to start.

Number of active deals in the pipeline. This is your “inventory level.” If you have too few deals, in a quarter you will have too little revenue. The minimum that gives you comfort is as many active deals as you want to close in a quarter, multiplied by the inverse of your conversion rate. If you close 25% of deals and want to close 5 in a quarter, you need 20 active deals.

Conversion by stage. What percentage of deals move from one stage to the next. This shows where you lose clients. If 80% move from qualification to discovery, but only 20% from discovery to proposal, you know where the problem is.

Average sales cycle. How much time passes from first contact to signing the contract. This lets you forecast. If the average cycle is 8 weeks, a deal that entered the pipeline today will close in two months, not in a week.

Pipeline value. The sum of the values of all active deals multiplied by the probability of closing at the given stage. This gives you a revenue forecast. Imperfect, but better than guessing.

Add more metrics only when these four are measured regularly and someone looks at them. Ten KPIs nobody looks at are worse than four that really influence decisions.

Once a week. 30 to 60 minutes. No more.

A pipeline review is not a “status meeting” where salespeople report in. It is a decision meeting where you go through the active deals and establish what is happening with them.

Every deal should be handled in one of three ways. It is moving forward: there is a next step, a date, a concrete action. Move on. It is stuck: no contact for two weeks, the client isn’t responding, no next step. Decide what you do: one follow-up, a phone call, and if that doesn’t work, close the deal. It is dead: nothing has happened for a month, the client has disappeared, there is no prospect. Close it and move on.

Three questions for every deal: What is the next step? When? What could go wrong?

What not to do at a pipeline review: don’t discuss company strategy, don’t talk about delivery problems, don’t resolve team conflicts. This is a meeting about sales. It should be short, concrete and lead to actions. If a pipeline review takes longer, something is wrong. Either you have too many deals to discuss (split it into two meetings) or you are drifting to other topics (stick to the agenda).

Not based on gut feeling. Based on the pipeline and historical data.

Forecasting in B2B service companies doesn’t have to be complicated. You need three things: a list of active deals with their value, the stage each deal is at, and historical conversion at each stage.

An example. You have 10 deals in the pipeline. Three are at the “proposal sent” stage, worth 80 thousand zł each. Historically you close 40% of deals at this stage. The forecast from those three deals: 3 times 80 thousand times 40% = 96 thousand zł. That is your forecast.

It will be inaccurate. But it will be better than “we’ll probably close something this month.” And with each quarter it will get more accurate, because you will have more historical data.

Two conditions for this to work. The pipeline has to be up to date (dead deals removed, stages matching reality). And you have to measure conversion systematically, not once a year.

Companies that forecast make better decisions. They know when to hire a new person. They know when to step up sales activity. They know when to tell a client “we can start in two months” instead of taking a project they don’t have the resources for. A forecast is not reading tea leaves. It is looking at the numbers you already have.

Because the CRM was implemented without a process. Or there is a process, but nobody enforces its use.

An empty CRM is a signal that people see no value in filling it in. The salesperson has their deals in their head and doesn’t need a system to remember them. Filling in the CRM is extra work for them with no visible benefit. So they don’t. Or they fill it in once a month, before the meeting with the boss, entering whatever makes it look right.

A CRM full of junk is a signal that there are no clear rules about what to enter and when. Everyone interprets the stages their own way. “In progress” for one salesperson means “we talked last week,” for another “I sent an email three months ago and he didn’t reply.” The data is there, but nothing comes of it, because it is inconsistent.

The solution is not buying a better CRM. It comes down to three things. Simplification: four or five stages, a field for the next step, a date. Nothing more. Enforcement: a weekly pipeline review where the CRM is the only source of truth. If a deal isn’t in the CRM, it doesn’t exist. Value: the salesperson has to see that the CRM helps them sell, not just report. When they open the CRM in the morning and see what they have to do today, they start using it. When they open the CRM to please the boss, they stop.

Start with the process, not the tool.

The most common mistake: the company buys a CRM, configures it for three weeks, runs a training for the team and says “from now on we work in the system.” A month later nobody is using it.

Why? Because a CRM is a tool that supports a process. If there is no process, the CRM has nothing to support. Salespeople don’t know which stages to pick, because the stages don’t match how they really sell. The fields are unclear. The reports show things nobody understands.

An effective implementation looks like this. Step one: describe your sales process. Stages, criteria for moving between them, the definition of a next step. Step two: configure the CRM so it reflects that process. A pipeline with your stages. Fields that make sense in your context. Nothing redundant. Step three: start using the CRM at the pipeline review. A weekly meeting where the CRM is the only source of information. If a deal isn’t in the system, it doesn’t exist.

And most important: the owner has to use the CRM themselves. If the owner keeps their deals in their head or in a spreadsheet, but demands the CRM from the team, the message is clear: the CRM isn’t important. And nobody will take it seriously.

Answers to four questions. How many active deals we have. How many we will close this month. Where we are losing clients. And what we are doing to fix it.

A report that answers these questions fits on one page. You don’t need a twenty-page analysis with charts. You need information you can base a decision on.

Specifically. The number of active deals in the pipeline and their total value. A comparison with the previous month. The forecast of closings for the current month (the value of deals at advanced stages multiplied by historical conversion). A list of deals that are stuck, with an action plan for each of them. Conversion by stage: where deals drop out.

What it should not contain: a list of all the salespeople’s activities (how many emails they sent, how many calls they made). That is operational information, not strategic. The owner doesn’t need to know that a salesperson made 47 calls. They need to know that the pipeline has 15 deals, 4 of which are at the negotiation stage and should close within three weeks.

The report should go out once a week or once a month. Not daily. Daily reports are noise. Weekly ones are information.

Bottom up, not top down.

Top-down targets look like this: “I want 2 million in revenue this year, so the salesperson has to close X deals a month.” Sounds logical. The problem is it ignores reality. How many leads are there? What is the conversion rate? What is the average sales cycle? If targets have no connection to the pipeline, they are fiction.

Bottom-up targets look like this: we get an average of 20 new leads a month. Conversion from lead to close is 15%. The average deal value is 60 thousand zł. So realistically we will close 3 deals a month with a total value of 180 thousand zł. That is a target that comes from data, not from wishes.

If 180 thousand zł a month is too little, you won’t fix it with a higher target. You fix it with more leads, higher conversion or a higher average deal value. Each of these elements requires a concrete action, not “motivating the salesperson.”

Targets should be ambitious but achievable. A target 20 to 30% higher than current results motivates. A target twice as high with no change in conditions demotivates. Because the salesperson knows it is unreachable. And stops trying.

Short, concrete, with decisions.

A sales meeting is not a “status update.” It is a meeting where you establish what is happening with every active deal and what you do next.

A format that works. Max 60 minutes, once a week. Each salesperson goes through their deals in the pipeline. For every deal: what stage it is at, what happened this week, what the next step is and when. If a deal is stuck, you set an action. If a deal is dead, you close it. If a deal needs help (escalation to the owner, support with the proposal), you decide who does what.

What to avoid. Meetings where one salesperson talks for 20 minutes about one deal. Meetings that turn into a discussion about company strategy. Meetings without the pipeline open on the screen (if you talk about deals from memory, not from the CRM, the CRM is fiction). Meetings that produce no concrete actions.

One rule: every meeting ends with a list of actions. Who does what and by when. The next meeting starts by checking whether the actions were done. That builds habit and accountability.

Managing salespeople is managing people. Motivation, development, feedback, compensation. Managing sales is managing a process. Pipeline, conversion, forecasting, optimization.

You can have great salespeople and weak sales management. Results will depend on individual talent. One salesperson will close 30%, another 5%, and nobody will know why.

You can have average salespeople and good sales management. Results will be stable, predictable and repeatable. Because the system gets more out of people than their natural talent alone.

The ideal situation is both at once. But if you have to choose where to start, start with sales management. Build the process, measure results, set a rhythm of reviews. Only then work on developing people. Because developing people without a system is like training pilots without a plane. They have the skills, but nothing to fly.

In the B2B service companies I work with, owners usually manage salespeople (they talk to them, motivate them, give feedback) but don’t manage sales (they don’t measure, don’t forecast, don’t optimize). And then they are surprised that results are unpredictable.

Not at the moment you lose. Systematically, once a quarter.

A one-off analysis of “why did we lose this deal” is of little value. Because every deal has its own story, and it’s easy to find an explanation that sounds sensible but changes nothing. “The client had no budget.” “They chose a cheaper option.” “The decision was postponed.” These are answers that don’t lead to improvement.

What is valuable is looking for patterns. Take all the deals lost in the last quarter. Group them. At what stage did they drop out? Who was the client? What was the lead source? Who ran the deal? What was the reason for losing?

The patterns that emerge will be concrete. “60% of lost deals dropped out after the proposal was sent.” That means the proposal is the problem. Or: “80% of lost deals were companies under 20 people.” That means qualification is too loose.

One caveat: don’t count on the client telling you the truth about why they didn’t buy. “We decided on another vendor” says nothing. The real reason usually lies in what you didn’t do: you didn’t reach the decision maker, you didn’t dig deep enough into the problem, you didn’t address the risk the client had in mind.

Not when “sales aren’t going well.” When sales are going well but need someone to run them day to day.

A sales director makes sense when the company has a working sales system and needs someone to execute it, develop it and keep it on track. Someone who runs the pipeline reviews. Who develops the salespeople’s craft. Someone who analyzes conversion and looks for places to optimize. Someone who is accountable for the result.

A sales director makes no sense when the company has no system. Because then the director walks into an organization with no process, no metrics, no clear definition of the client. And has to build all of it alone. Some can. Most can’t. Because they were sales directors in companies where the system already existed. They managed, they didn’t build.

The most common scenario I see: a company hires a sales director for 15 to 25 thousand zł a month because it “needs someone to get sales under control.” After six months the director leaves frustrated, because there were no foundations to work on. The company loses money and is in the same place.

The right order: first the system, then the director. Build the process, measure results, create a sales rhythm. Only then hire someone to run it. A sales director should walk into a company where they know what to do. Not a company where they have to figure out what to do.