Most brands don’t lose a customer in one go. It’s something that happens over time through various inconsistencies from the company. Fragmented data lets different parts of the business run on different, disconnected pictures of who that customer is, diluting the customer experience. Customers expect the brand to know them consistently across all touchpoints. Meanwhile, this challenge has become common across industries, whether banking, retail, or even B2B.
Segmentation, targeting, and positioning is a marketing discipline that’s been built for stronger customer experiences. Grouping customers by shared traits, finding out what matters to each group, then positioning the brand so it actually connects with them is not a new idea. This three-step process has run for years to spend less and convert more. What’s changed is that AI tools have complicated the first step. Marketing sees one version of the customer and sales sees another, as they are looking at different data fragmented across various departments. A business can’t segment or target without a shared, comprehensive view of the customer. That view starts with a single CRM record each department actually trusts and is willing to invest in. This is a challenge.
The customer only ever experiences the brand as one relationship with the company. A contradiction through any of the communications coming from various departments reads to them as the brand failing to know them.
This is why AI governance belongs in the same conversation as brand strategy, owned by those who hold that customer relationship. While IT teams can keep the data safe, whether the story marketing tells a customer actually matches what sales just told them comes down to brand judgment. Security policies can’t make those calls. Instead, it has to sit with whoever owns the full relationship of the brand and communication with the customer. The tools multiplying across departments right now are a targeting problem, the same one marketers have solved for years, but they are arriving at a scale and speed the old manual fixes can’t keep up with.
It’s been documented that data trapped in silos, with no unified view across departments, takes organisations well over a year to properly untangle. In that same year, AI tools keep multiplying regardless of whether anyone is governing them. IBM’s 2025 Cost of a Data Breach Report found that only 38 per cent of organisations in the region have a formal policy that could even catch this. The gap reflects a broader pattern, which is that tools are arriving faster than the frameworks meant to govern them, and the research shows that this region is no exception. That delay shows up in day-to-day internal friction first, and in lost market share later.
That gap between brand trust and AI trust is documented in Edelman’s 2025 research, which found a 26-point difference between how much people trust the technology sector and how much they trust AI. This is proof that having a strong brand doesn’t mean people trust what you do with AI. Trust is earned slowly, the same way every other part of the brand relationship is earned. It also requires consistency.
I’ve seen the alternative work firsthand, leading marketing through a multi-entity post-merger integration. The change that actually moved the business wasn’t a single team’s tool stack, but getting marketing, sales, customer service, and other departments to look at the same customer record. That unification was also one of the drivers behind the group’s largest carrier sales growth to date. While systems integration was arguably complex, getting every team to actually trust and build on that shared view took more work.
When that shared view remains, segmentation and targeting do their job. When it breaks, customers notice long before the numbers do.
So what can be done by management or boards to solve this?
First, ask whether marketing, sales, and service are working from the same definition of the customer, or whether they are working separately but just sharing the name. That clearly a targeting problem, especially as AI has made it easier for departments to build their own version of who the customer is.
Second, remember that the customer record is meant to act as a single source of unified truth for the business. To apply this, ensure that every AI tool gets adopted the right way, by ensuring that it’s gone through a test where that single customer view has to be proven, tracked, and confirmed to strengthen it while avoiding fragmentation.
If your departments can’t agree on who the holistic view of your customer is, that’s a brand problem and is costing the company the equity your segmentation and targeting work was initially built to protect.
May Neama leads two decades of brand transformation, transition, and alignment across MENA. She’s directed Kalaam Telecom’s post-merger rebrand and carrier sales growth, and headed the regional client servicing department of a consultancy, overseeing accounts such as ADNOC, SABIC, and Qatar Foundation. She is a graduate of Columbia Business School’s Chief Marketing Officer Program and has guest lectured at Columbia, NYU, and Boston University.
&uuid=(email))