CRM problems

CRM problems: why they keep coming back after every fix.

Poor adoption, unreliable data, weak forecasting and work happening outside the CRM can be technical problems. They can also be signs that the system is trying to represent business logic that has never been made clear.

Before diagnosing the problem

Which of these looks most like your CRM?

Not which platform you use. What the system actually represents when people work in it.

One more question

When does possible future work become visible to the business?

This is often more revealing than whether the CRM is tidy.

01As soon as someone thinks it may happen.

There is somewhere to represent uncertainty before a formal proposal or deal exists.

02When somebody creates an opportunity.

Anything before that tends to live in conversations, notes or memory.

03Usually when a proposal is sent.

The business only sees possible work once it has already moved quite far.

04It doesn't. Someone remembers to follow up.

The relationship may be well documented while the possible commercial change is not represented at all.

A CRM can contain a rich history of a relationship and still have no explicit place for what might happen next.

If deals move through stages

Your CRM already represents uncertainty - but only inside the pipeline.

The useful question is what disappears before an opportunity is created, what the stage actually means while it is open, and what becomes invisible after the deal is won.

Two deals can sit at the same stage while procurement, technical approval, delivery capacity or stakeholder change make them fundamentally different. Existing customers may also fall outside the pipeline until somebody decides to create another opportunity.

The pipeline can be well maintained and still represent only part of the commercial reality.

If the CRM mainly stores relationships

The history can be detailed while the next commercial change stays implicit.

The system may know every meeting, email and contact and still have no separate object for "they may buy more", "renewal is becoming uncertain", "a new stakeholder has appeared" or "this commitment changes the relationship".

Management then learns about possible work, risk or change from the person holding the relationship rather than from the system.

The CRM is not necessarily failing. It may simply have been designed to remember relationships, not represent what could happen next.

If you do both

New-business uncertainty may be explicit. Existing-customer uncertainty may not be.

A £30k prospect may have an opportunity, stage, expected close date and owner. A £200k existing customer that may renew, expand, shrink or leave can still appear as a stable account until somebody creates a new deal.

This is one reason existing revenue can look safer in the CRM than it really is.

An existing customer is not existing future revenue.

The common complaints

The same CRM complaint can come from very different systems.

Before fixing adoption, data or forecasting, establish what the CRM was designed to represent - and what commercial reality sits outside that model.

01 · Adoption

People do not use the CRM consistently.

02 · Data

Nobody fully trusts the information already in it.

03 · Handover

Sales and Delivery hold different versions of the customer.

04 · Forecasting

The forecast looks reasonable until reality arrives.

05 · Economics

Revenue is visible, but the work created by the customer is not.

These symptoms do not prove one root cause. The useful first step is to separate a technical fault from a mismatch between the CRM model and the business.

The distinction

The CRM may be the problem. Or it may be where the problem becomes visible.

A broken integration, duplicate records, poor permissions or an unusable interface are real CRM problems. They can and should be fixed as system problems.

But if the same issue returns after cleanup, retraining or reconfiguration, I look at the business logic underneath it: definitions, ownership, handovers, decision points and when information is captured.

01 · Adoption

People do not use the CRM consistently.

Training may help if people do not understand a system that otherwise fits the work.

But low adoption can also mean the CRM duplicates another task, asks for information too late, does not support the decision people are making, or reflects a process that exists on paper but not in practice.

Before adding mandatory fields or another training session, I would separate a knowledge problem from a workflow problem and a genuine usability problem.

02 · Data

"Dirty data" can have more than one cause.

Technical

Duplicates, failed integrations, import errors, permissions or configuration can damage otherwise well-defined data.

Definitions

Two people can enter different values correctly if the business has never agreed what a qualified lead, won deal or active customer means.

Timing

Records reconstructed after the work happened are less reliable than information captured when the relevant decision or event occurs.

A cleanup project can correct the current record without changing the mechanism that recreates the problem.

03 · Handover

A customer record can move while the commercial context does not.

Sales can record the deal correctly and Delivery can still inherit information it cannot use.

What was promised, what remains uncertain, what is outside standard scope, which requirements affect effort and who owns the next decision may not travel with the customer.

The CRM then looks complete at the point of sale while the operational cost appears later.

04 · Forecasting

The forecast can only use uncertainty the business has made observable.

Two opportunities can sit at the same pipeline stage and have very different reasons they may not close: procurement, technical approval, delivery capacity, stakeholder change or a commercial dependency held outside the CRM.

If those differences have nowhere to be recorded, changing the probability percentage does not solve the underlying information problem.

05 · Customer economics

Most CRMs are better at showing revenue than the cost of producing it.

A £50,000 customer and another £50,000 customer can look identical in a sales system while requiring very different delivery effort.

Hours, rework, exceptions, senior involvement and coordination may live elsewhere. The problem is not necessarily that the CRM should become the delivery system. It is whether the business can connect the customer identity across the systems that hold the economic evidence.

What I look for

Three questions before deciding what to change.

Definitions

Does the business agree what the important customer, opportunity, stage, risk and handover states mean?

Evidence

Are the events needed to observe those states captured consistently, at the point they happen?

Interfaces

Does the information needed by the next function move with the customer, or have to be reconstructed later?

Before automation

Do not automate an ambiguity you have not defined.

The same foundation matters for CRM automation and AI. A system needs explicit definitions, observable evidence and clear decision logic before scoring or automation has reliable business context.

Questions people ask

CRM problems, separated.

Is poor CRM adoption usually a training problem?

Sometimes. But inconsistent use can also come from duplicate work, a workflow that does not match the way the business operates, unclear ownership or a system that asks people to record information after the decision has already happened.

Is dirty CRM data always a data-quality problem?

No. Duplicate records, failed integrations and poor configuration are genuine data problems. Unreliable data can also appear when teams use different definitions or reconstruct records after the work has happened.

When is it genuinely a CRM problem?

When configuration, permissions, integrations, duplicate records, usability or technical constraints prevent the system from doing a job the business has already defined clearly.

Why do the same CRM problems return after cleanup or retraining?

If the original cause sits in definitions, handovers, ownership or capture timing, improving the current records can remove the symptom without changing the mechanism that recreates it.

Start here

Not sure whether this is actually a CRM problem?

I research the business before the first call, then use the conversation to separate a system problem from a process, data or commercial one.

If the CRM works locally but management cannot connect it to Delivery and Finance, the Revenue Visibility Review is designed for that question.

If you are already using a named platform and want to know how this work fits around it, see the CRM and systems pages.

Free · 30 minutes · No presentation to prepare.