Salesforce rarely fails in a single dramatic moment. It drifts. A field here, a workaround there, a report nobody quite trusts — and six months later the team is working around the system instead of with it. The good news is that the drift is easy to spot once you know what to look for.
1. People keep a spreadsheet “just in case”
If your best salesperson runs their week from a personal spreadsheet, that is not a training problem. It is feedback. Something in Salesforce is slower, harder or less trustworthy than a blank grid, and people have voted with their time. Find out what the spreadsheet does that the org does not.
2. Two reports on the same thing disagree
When “pipeline this quarter” gives you three different numbers depending on who ran it, the problem is usually definitions rather than data. Stage names, close-date hygiene and inconsistent record types quietly pull reports apart. Until the numbers agree, nobody will make decisions from them.
3. Every small change needs a developer
Adding a picklist value or adjusting a validation rule should take minutes, not a change request and a fortnight. If routine adjustments always require code, the configuration has usually been over-engineered — automation stacked on automation until no one is sure what will break.
4. Nobody trusts the pipeline
If leadership mentally discounts the forecast by 30% before every board meeting, the CRM has stopped doing its main job. That is normally a process gap, not a tooling one: no shared definition of what each stage means, and no habit of keeping opportunities honest.
5. Onboarding a new starter takes days
A new joiner should be useful in Salesforce within an hour. If it takes days of shadowing to learn which fields matter and which to ignore, the interface is carrying years of accumulated clutter that no one has been given time to clear.
What to do about it
None of these need a rebuild. They need a short, focused review and a prioritised list of fixes:
- Sit with the people who use Salesforce most and watch what they actually do — including the workarounds.
- Agree one definition for each pipeline stage and each core report, and write it down.
- Retire automation and fields that no longer earn their place.
- Fix the two or three things that cause the most daily friction before touching anything else.
A system that has drifted can be brought back. It mostly takes someone willing to look closely, and the discipline to change a few things at a time.