What do you often encounter in practice: an organization - large or small - has decided to become data-centric. 'We have so much information about our customers, and yet we approach them every time as if they were walking in for the first time. Both online and in locations.' That's great, such insight. Then you've usually come to the right place with me. And yet, things quickly go wrong...
Stakeholders
The intended decision is made. Stakeholders are brought together. IT, data, marketing, sales. Sometimes multiple roles within one function. Usually a project manager is added and the project can start. The business argues how relationships are better served if you understand them better. This should be reflected in communication, in the message, the offer, the timing, in the channel used. Some examples are given. At least, that applies to smaller organizations, the big guys work out "use cases" in PowerPoint or Excel, including all the Requirements. So far, perfect preparation. No harm done. Then what? Then IT and other support stakeholders take over.
Everything in one data model
Business requirements are melted down to zeros and ones in no time. Customer data turns out to reside in many more systems than expected, and also with many different suppliers. That has to be changed. That must be "flattened" into one model. A data model. Clear. Because once we have unambiguously recorded everything in it, we have our backbone. One platform, one 'single source of truth,' with nice dashboards on it, so we can see exactly how our customers are doing 360-degree and how they are doing. Then our marketers and salespeople can get to work. We flattened it out nicely for them.
Processes take over
In short, if you're not careful, whether you're in a large or medium-sized organization, before you know it, processes take over any sense of business. Processes can be clearly articulated. Business is vague, they just have to follow those processes from now on. Many a marketer has discussions about data that does not fit well into the "data model"... Computer says no. Please let's not forget that all those processes only help. Very well, in fact. To make it all more insightful and manageable. But mostly to promote commercial creativity. Not the other way around.
