The system was chosen carefully, configured properly and everybody was trained. Six months later, half the team keeps a spreadsheet and the records are unreliable.
This is almost never a discipline problem, and more training will not fix it.
What resistance is telling you
People adopt tools that make their own day easier. They avoid tools that make it harder for somebody else's benefit. That is rational, and it is the single fact that explains most adoption failure.
So when a system is not being used, the question is not "how do we make them use it" but "what is it costing them".
The usual answers:
It is slower than what they did before. Logging a call took ten seconds on a notepad and now takes ninety seconds and four clicks.
The benefit is invisible to them. The data feeds a management report. The person entering it gets nothing back.
It does not fit the actual work. The form assumes a sequence that does not happen in that order, or has no field for the thing that matters most.
It is not available where the work happens. On a phone, on site, with no signal.
They were not asked. A system chosen without the people who use it will be resented before it is opened.
Adoption failure is a design failure wearing a training costume. If people are avoiding the system, the system is asking too much of them.
The changes that work
Make the daily view answer their question
Most systems open on a dashboard designed for a manager. Open instead on the thing the user actually wants: my jobs today, my overdue follow-ups, what is waiting on me.
If opening the system answers a question they had anyway, they will open it.
Cut what must be entered
Go through every required field and ask what decision it supports. Delete the ones with no answer. Most implementations can lose half their mandatory fields with no loss of anything.
This single exercise rescues more systems than any other intervention.
Capture automatically wherever you can
Email that attaches itself. Enquiries that create records. Addresses looked up from a postcode. Payment status arriving from accounting.
Every field that fills itself is a field nobody resents and a field that is always accurate — see the data worth capturing.
Give something back immediately
The person entering data should get something for it in the same session. Their next actions listed. A quote generated in seconds from what they just entered. The customer's full history when they take a call.
Systems that only take are systems that get abandoned.
Fit the work, not the org chart
If the form requires a stage that does not exist in the real process, people will pick something arbitrary and the data will be worthless. Watch the work being done before configuring it — this is the value of a proper discovery.
Rolling out without a mandate
Involve the sceptics early. The person most likely to object is usually the person who knows the process best. Bring them into the selection, take their objections seriously, and they become the strongest advocate.
Start with one team for a month. Fix what they complain about before anyone else sees it. Then roll out with them saying it is good, which carries far more weight than management saying so.
Do not run two systems in parallel indefinitely. A brief overlap builds confidence; a permanent one guarantees the old way survives. Set a date and hold it.
Migrate the history. A system starting empty feels like extra work for no benefit. One containing three years of customer history is immediately useful — see migrating without losing history.
Fix complaints visibly and quickly. The first fortnight sets the tone. A request raised and resolved in two days tells everyone the system is theirs. One that disappears tells them it is not.
What not to do
Mandating it. Produces records that exist and are not true, which is worse than no records, because now decisions are made from them.
Reporting on compliance. Measuring who updated the system produces updates, not accuracy.
Blaming the users. If capable people are avoiding a tool, they have a reason. Find it.
More training. Training fixes a knowledge gap. It does not fix a system that is slower than the alternative.
How to tell whether it is working
Not by login counts. Ask instead: is anyone maintaining a private spreadsheet? Are the records true when you spot-check them? Do people use the system to answer their own questions, or do they ring a colleague?
That last question is the best test. A system people consult voluntarily has been adopted. One they only feed has not.
Our CRM team configures and rescues systems for UK businesses, starting with what the daily user is being asked to do. Start a conversation if you have a system everybody works around.
