The sales pipeline in a mid-sized company: stages, forecasts and reporting the board can trust

Take a hypothetical distribution company with 150 employees. The management board sees PLN 2.4 million in opportunities for the quarter in the CRM and plans hiring on that basis. At the end of the quarter, the company signs contracts worth just under a million. Sales did not fail – the method of counting did.

Such a gap is usually caused by the same three mechanisms. Each of them can be removed without changing the system, simply by changing the rules:

In a company that sells to other businesses, with a sales cycle of a few weeks to a few months, six stages are usually enough. With a very short cycle, they can be merged into four or five. Each additional stage only makes sense if it carries new information.

The most important column in the table is the second one: the entry criterion. It is a fact that can be checked, not the salesperson’s feeling. For example, an opportunity enters the “Quote” stage only once the quote has been sent, not when the customer asked for it.

The probabilities in the table are typical starting values. In your company they may look completely different – we explain below how to calculate them. Every lost opportunity should have its reason recorded, because without it you will not find out why the company is losing.

The hardest step is moving on from needs analysis, because this is where opportunities that were never real most often end up. Qualification methods help here: lists of questions to which a salesperson must know the answers before moving an opportunity forward.

BANT consists of four questions: does the customer have a budget, are we talking to the decision-maker, is the need real, and is there a decision date? The method is simple and is enough for a sales cycle lasting a few weeks.

MEDDIC is more detailed and is useful for sales that take months and in which several people on the customer’s side make the decision. It requires you to establish the six things listed below. Whichever method you choose, all salespeople must use the same one, because otherwise their pipelines cannot be compared.

The simple sum of all opportunity values almost always inflates the picture, because it treats a first conversation the same as a contract out for signature. The weighted value multiplies each opportunity by the probability of its stage and only then adds up the results.

In the example in the table below, the simple sum is PLN 2.4 million and the weighted value PLN 830,000 – just over a third. If the quarterly target is PLN 1 million, the management board sees the gap immediately, instead of discovering it at the end of the quarter.

The weighted value is only the first step. The second is the salesperson’s commit: the list of opportunities they are confident they will close this quarter. The third is the sales manager’s forecast, which combines both figures with the manager’s own risk assessment; this is the figure passed to the management board.

Starting probabilities are only an assumption. After a few quarters of working with clear stages, you can calculate your own rates from the history of closed opportunities. These are what make the forecast start to match the result.

The calculation is simple. Take all the opportunities closed in the past year that reached at least the “Quote” stage, and check how many of them were won. If 80 customers received a quote and 28 signed a contract, the real probability for that stage is 35%, not 50%.

Repeat the same calculation for each stage, separately for the main product or customer groups if they differ a lot. Update the rates once a quarter. When they differ sharply from the starting values, it often means the stage entry criteria are not being followed.

Even the best-designed stages will not help if the CRM data is out of date. A few simple rules, enforced at the weekly review, keep the pipeline in good shape. Some of them can be enforced through CRM settings, for example with a mandatory field when an opportunity is closed.

A sales pipeline only stays alive if someone comes back to it regularly. Companies that review it only once a quarter find out about a weak result when it is too late to react. The table below shows a sensible rhythm for a mid-sized company.

At the monthly review, the management board does not need a list of all opportunities. A few figures are enough: the weighted value compared with the target, and the number of opportunities at each stage with its change since the previous month. Add the average time from first contact to contract signature and a list of the key opportunities with their close dates. The most important figure, however, is accuracy – a comparison of last month’s forecast with actual sales.

The choice of system matters less than the rules described above, but it affects how much work reporting takes. In monday CRM there is a ready-made opportunities board with a forecast value column calculated from value and probability. Salespeople usually find their way around it quickly, and the stages appear as columns on the board.

Dynamics 365 Sales works better in companies on Microsoft 365 with a larger team and more extensive reporting. Emails from Outlook are easy to link to a specific opportunity, and the data can be moved into Power BI, Microsoft’s tool for reports and results dashboards. Other popular systems, such as HubSpot or Pipedrive, have similar core features.

Whatever the system, it is worth having the management report generated automatically from CRM data. When the sales manager copies the figures into Excel every month, the report is out of date before it reaches the table, and errors creep in easily.

AI tools in CRM systems can score each opportunity based on the history of similar opportunities and the salesperson’s activity, such as how often they are in contact. They can also flag opportunities at risk of slipping or being lost before the manager notices.

The precondition is data. In Dynamics 365 Sales, the opportunity scoring model will not start unless the company has at least 40 won and 40 lost opportunities from the chosen period. A mid-sized company with a few salespeople and large contracts may not have that many lost opportunities a year, especially if it has not been closing them until now.

So the order is clear: first stage criteria and clean data, then your own rates, and only at the end AI. An AI forecast works best as a second opinion – when it differs from the manager’s assessment, it is worth taking a closer look at that particular opportunity.