Most CRM business cases compare licence costs, implementation fees and perhaps a year of support. They rarely include the line that decides whether the project lands on time: how many hours your own people will spend on it, on top of the jobs they already have.
Those hours are real, they are predictable, and they are almost always underestimated. Here is where they go.
Data cleansing is a staff job, not a supplier one
A supplier can write scripts to deduplicate records and standardise addresses. What they cannot do is decide which of two conflicting records is right, whether a supporter who gave once in 2014 should come across at all, or what an undocumented activity code meant to the team who created it.
Those decisions need people who know the supporters, and in most charities those people are also running the next appeal. Plan for their time explicitly, in weeks, and protect it.
- Deciding what to migrate, archive or leave behind
- Reviewing duplicates and conflicting records a script cannot resolve
- Explaining what old fields and codes actually mean
- Testing trial migrations against what the team knows to be true
- Running two systems in parallel during cutover
- Training, and the slower weeks that follow go-live
Parallel running doubles the admin
For a period, gifts, Gift Aid declarations and consent changes may have to be recorded in both systems, or reconciled between them. That period always feels short in the plan and long in practice. Keep it as short as you safely can, agree in advance exactly what gets entered where, and have one person accountable for reconciliation.
Every week of parallel running is a week of doing the same job twice.
Training is the start of the cost, not the end
A day of training teaches people where the buttons are. It does not make them as fast as they were in the old system, where they knew every shortcut and workaround. Expect a dip in productivity for the first few months, particularly for supporter care and data teams, and avoid launching into your busiest fundraising season.
How to make the number smaller
- Start cleaning data now, in the old system, before the project begins
- Agree a retention rule for old records early, so the migration scope shrinks
- Rehearse the migration several times, so cutover weekend holds no surprises
- Give named staff protected time for the project, and backfill their day jobs if you can
- Choose a go-live date away from Christmas, year end and your major campaigns
None of this makes a CRM migration cheap. It makes it honest. A plan that counts staff time from the start is one the organisation can actually deliver, and one that does not quietly burn out the people you need most once it goes live.
