What Is Pipeline Management? Making Deal Progress Visible to Improve Forecasts
Published:
Last Updated:
Category: CRM, LTV & Customer Management
Published:
Last Updated:
Category: CRM, LTV & Customer Management
Authors: Shusaku Yosa
Asked where the quarter will land, you cannot answer with any confidence. There are deals in progress, but you cannot tell which of them are genuinely moving. The cause is usually not the forecasting formula but the definition of the stages the deals sit in. This article covers how to design pipeline management.
A way of grasping the overall state of the business by arranging live deals into stages. The path to a closed deal is treated as a pipe, and you look at how many sit where.
There are two purposes. One is improving the accuracy of the forecast. The other is finding where deals are stuck.
The second is the more useful one. Forecasts get thrown off by external factors, whereas a stalled deal will move if you act on it.
Whether pipeline management works is decided here.
Most organisations define stages by their own activity: "first visit made," "proposal delivered," "quote submitted."
With that definition, a deal advances simply because the salesperson did something. Sending a quote means nothing if the other side has no intention of evaluating it.
Define each stage by the customer's state instead.
Defined this way, advancing a stage requires drawing something out of the customer. Progress starts reflecting the reality of the deal rather than your own volume of activity.
Finer divisions feel closer to reality, but the opposite holds.
More stages make updating tedious, and things get left diverging from reality. A pipeline that is not updated is the same as no pipeline.
Four to six is the guide. Past seven, check whether each boundary really marks a change in the customer's state. Turning your own internal steps into stages is what makes the number grow.
A pipeline is not finished once the deals are arranged. Look at these three.
How many days each deal has sat in its current stage. Anything longer than the average is stalled.
This metric points at where to act better than value or probability does. A long-stalled deal is sometimes already lost and simply still sitting in the pipeline.
Look only at value and one large deal makes the whole thing look healthy. The moment it disappears, the quarter collapses.
Putting the count alongside it shows whether you have become dependent on a single deal.
Compare how many deals came in this month against how many left through wins and losses.
Let exits outnumber entries for a sustained period and you run out of deals a few months later. Watching this quarter's numbers hides it, so check it separately.
Assigning a probability to each deal and multiplying by value is standard practice, but the input is the problem.
Let each salesperson enter probability by judgment and the criteria differ person to person, making the total meaningless. An optimist and a cautious colleague will differ by more than double on identically-placed deals.
Fix probability by stage rather than leaving it to individuals. Proposal delivered means 30 percent, quote submitted means 50 percent, and so on.
Derive the values from history. Count what share of deals at "proposal delivered" closed over the past year, and that is your probability.
Where individual circumstances need reflecting, write them in a comment field and discuss them separately rather than adjusting the probability.
Even with the mechanics in place, updates often stop within months. There are two usual causes.
Entry that exists only for reporting does not last. Something has to come back — support triggered by what the pipeline shows, or help thinking through priorities.
Where writing down a lack of progress invites interrogation, the information entered diverges from reality.
When a stall gets reported, work out the next move together rather than assigning blame. Only once that response is established does accurate information start coming up.
The pipeline is thought of as a sales artefact, but it is an important source for marketing too.
What to look at is which stage deals from each campaign tend to stall in.
If deals from one campaign drop out heavily at the early stages, the audience you are attracting may be off. If they drop out late instead, there is room to check whether the messaging is raising expectations too far.
Tracking lead count alone makes this judgment impossible.
Set a limit on time in stage and mark anything beyond it as lost. If it revives, register it again as a new deal. Leaving old deals in place makes the whole pipeline look thicker than it is.
Past about twenty deals, you cannot hold it in your head. Below that, simply reviewing the stalled deals once a week does more good than arranging everything into stages.
Check whether the stage definitions are written as your own activity. Treat "we made a proposal" as progress and deals accumulate regardless of the customer's intent. Fixing the definitions alone changes forecast accuracy considerably.
It is. Decision-makers change, budgets fall through — states genuinely do regress. Make it impossible to move backwards and deals travel forward while diverging from reality.
Progress after a deal enters the pipeline may be visible, but which campaign produced it usually lives somewhere else.
Xtrategy manages campaign schedules alongside budget and KPIs on a single screen. With spend linked to results, you can judge where to push the top of the pipeline.
What helps most in pipeline management is revisiting the stage definitions. Start by listing your current stage names and checking whether they are written as your own activity.
How to write a marketing proposal, covering a seven-part structure and the budget section. Explains deciding what decisi...
How to build a KPI tracking sheet, covering column design and the operating pattern. Explains why target, actual, and at...
How smaller companies should choose a CRM when nobody can be assigned to it full time. Covers five criteria — setup time...
