The hidden cost of scattered systems

A cluttered room, a cluttered desk, a cluttered mind… A system made up of disparate applications and processes always carries the same risk of falling apart. So what do we mean by falling apart? A system whose parts cannot talk to each other, and even when they do, cannot agree — a system where no single vantage point gives you the whole picture. Investing in a new application for every need that comes up over time, and folding each solution into the organisation for what looks like a modest per-user fee, eventually leaves you with dozens of applications that have to be updated every time your approach changes. That is why, for a retailer running five separate applications, the real bill is not the licence fee; it is teams trying to get to know the same customer all over again every morning, through different channels and with incomplete data. Asking how many pieces of software a brand runs is really asking how many pieces it has cut its customer into. Here is the picture we see most often in the field: one application is used for campaigns, sales are tracked from another panel, loyalty points are held and awarded in a third system, and the contact centre talks to customers through a fourth, independent screen or team. All four work correctly; none of them knows it is talking about the same customer. So while we build a multi-channel structure and try to give the customer a multi-channel experience, we drift further away from the fact that we cannot actually recognise that one customer. Entering an order into the system from a customer name, phone number and address scribbled on a small notepad in the store brings its own set of errors. Set aside the consents that were never collected; think about what it costs in terms of representing the brand. And it is not even certain that the data was entered correctly. These are scenarios that hurt the customer experience while their financial cost cannot be calculated in the moment. Meanwhile every system that cannot communicate properly with the others causes data to be duplicated and changed information to stay behind in the source system. Duplicated data slows systems down, indirectly increases the need for resources, and consumes more of two of today’s most valuable resources: database capacity and bandwidth. Hidden costs, in short. What is more, data multiplying across different systems is not only an operational burden; it also increases security and data-protection compliance risk. It becomes steadily harder to track which systems hold a customer’s data, who can access it, and whether every copy is genuinely deleted when a deletion request arrives. As systems scatter, responsibility for the data scatters with them. Another invisible cost of this setup is the sustainability of the integrations. Every change made in one application can require the other systems connected to it to be rebuilt and retested. A solution that looked small and self-contained at the outset turns, over time, into a dependency that makes changing the whole structure harder.

The bill is not in the licence, it is in repeated work that leads nowhere

Add up the licence lines and the number usually looks bearable. The invisible line is the reconciliation reports prepared by hand every week, the meetings spent between two reports that do not match, and the time given over to the question “Which number is right?” Either you accept the wrong one as right and carry on blindly, or you simplify and let the right team focus on the right thing. In a measurement we ran with a retail team, preparing the results report for a single campaign took three separate teams a total of two days. Once we brought the same data together in one place, that work came down to a single screen where the report builds itself. Interpreting the report, with AI support, was perhaps a matter of minutes.

Drawing on the experience we gained at Microsoft, this is exactly why we built the INNTHEBOX ecosystem around a single customer profile. CDP, CRM, marketing, e-commerce, loyalty and analytics systems all read the same record; information updated in one channel is instantly current in the next. The goal is not to configure more modules, but to make sure that when the morning meeting reviews the numbers, everyone can look at the same screen. And being able to focus on what your own role requires instead of playing detective with the systems gives every team a great deal of ease and confidence.

Where to begin: identity, ownership, measurement

We start the unification work by defining customer identity, not by picking a product: what makes a customer unique at this brand? The phone number, the e-mail address, the national ID number, or the loyalty number? If every team answers that question differently, whichever software you buy, the data will split apart again.

The second step is ownership. If it is not written down which team updates which field, unified data scatters again within six months. The contact centre corrects the address, the e-commerce system writes the old one back, and nobody knows which is right. In that situation you also lose any chance of making the data gain value in proportion to how current it is.

The third step is measurement: you have to pick a single number that will show whether the unification worked. Our favourite indicator is how many screens you have to go through to get the answer to a question. If a question that took four screens at the start of the project can be answered on one screen by the end of it, you have done the job right.

In short, the cost is not only the licence. The cost is lost time, repeated effort, multiplying data, and trying to get to know the customer all over again at every point of contact.