The conversation nearly always starts the same way. Someone takes the annual subscription, multiplies it by five years, compares it against a development quote and draws a conclusion. It looks rigorous and it is not, because it compares two things that are not the same thing.
The subscription is not the cost of the CRM
The subscription is the cost of the licences. The cost of the CRM also includes the people who configure it, the outside consultants you call when the settings panel will not do what you need, and above all the time the company spends bending its own way of working to fit a model somebody else designed.
That last item appears on no invoice, and it is almost always the largest. A concrete example, from an industrial group we work with. In their sector a company attached to a project can be the main contractor, the designer, the material reseller, the broker or the promoter. Five different roles, and the same company can hold a different one on different projects.
Standard CRMs allow one customer per opportunity. One field. The result was that for years contractors were searched on that field, and 90% came back with no projects at all: they were all attached through the other roles, living in secondary fields no search ever looked at.
That missing data was not a software fault. It was a data model wrong for that trade, and no subscription would have fixed it.
The three items that actually decide it
1. What the system cannot do, and what that costs
Every off-the-shelf CRM has a boundary. Inside it, it is excellent and cheap. Outside it, you start with custom fields, then automations, then a third-party module, then a consultant. The curve is not linear: it climbs steeply exactly when the company grows and gets more specific.
2. What it costs to leave
After three years inside a platform the data exists but the logic does not. Automations, approval flows, visibility rules, reports: all of that lives in a configuration that does not export. Changing supplier means rebuilding it, and that is precisely why the subscription can afford to grow every year.
3. What it is worth to know the data is right
This is the item nobody weighs and the one that decides everything later. A system that tells you when something does not add up is worth more than a system that answers quickly. On the project above, the system we are building has an engine that notices on its own when the source platform renames or removes a field: it identifies the successor field, scores its confidence, and if it is not sure it halts the import instead of filling the database with empty records. A loud stop always beats a silent error.
When the subscription is the right answer
Often. Worth saying, otherwise this reads as a sales pitch.
A subscription CRM makes sense when your sales process looks like everyone else's, when user numbers are low, when you have nobody who could look after a system of your own, and when you need something by tomorrow morning. In all those cases buying is smarter than building, and anyone telling you otherwise is selling.
The moment the balance shifts is recognisable: when the part of the job the system does not cover has become more important than the part it does, and when people have started keeping spreadsheets next to the CRM to do the real work.
The transition is not a leap in the dark
The real obstacle is not technical, it is the fear of the switch. Nobody wants to wake up on a Monday without the system their sales team lives in.
That is not how it is done. The new system grows alongside the old one and stays synchronised with it — reading first, then writing too — until it is ready to carry everything on its own. People move in one module at a time. Switch-over day, when it comes, is a boring day: which is the sign the migration went well.