which crm should we choose?

there are over 300 systems on the market. half of all crm projects do not live up to expectations. it is rarely the platform's fault.

it is rarely the system that fails

it is not because the systems are bad. salesforce, hubspot, dynamics, pipedrive and the rest can solve most problems. but they cannot solve problems that have not been defined. and they cannot create processes that have not been thought through.

"if the choice is system-driven, the cheapest system is usually the best. then you lose the least when the project fails."

per østergaard jacobsen, cbs

the right starting point

the right question is not which system, but what you want to achieve with it.

it sounds obvious. it is not. most companies skip the step or reduce it to "a better overview". a better overview is not a goal. it is a hope.

before choosing a system, three things need concrete answers.

1. the bottleneck. where is the bottleneck in your customer journey right now? is it lead qualification, quote follow-up, onboarding or retention?
2. use cases. which three to five use cases will you actually activate in the first phase? and who will use them day to day?
3. measurability. how will you be able to tell it works six months after go-live? and what does it mean for the bottom line if it does?

when those answers are clear, the choice of system is usually fairly simple. sometimes you do not even need a new system, but a more disciplined use of the existing one.

when the choice has already been made for you

in experience and membership businesses the picture often looks different. you do not have 300 systems to choose from, you have a ticketing system or a membership platform chosen years ago, that the rest of the business depends on.

then the question becomes a different one. not which crm, but where customer data should be brought together when it comes from a system you do not control. that answer rarely depends on which crm tops a review site.

2026 actually makes it easier

there is a reason we say it right now. the last two years have made several things cheaper than they have ever been: prototyping a process, building an integration, testing a data setup. ai tools push the limit of what a small organisation can experiment with, without committing to heavy platforms from day one.

that means the "choose the system first" strategy is even more expensive now than it was before. and that "start small, feel the effect, build up" is a real option, even for companies that would previously have been forced into an enterprise setup.