When a multi-industry conglomerate decides to go digital, the story almost always begins the same way: a reputable ERP platform is selected, a large budget is approved, a "go-live" date is set. Then, 12 to 18 months later, the project slips, the budget balloons, and the most frustrating part is that no one can point to exactly where it went wrong. 

After years of working with businesses structured as holdings with many member companies, we've come to one conclusion: most projects don't collapse because of technology. The technology almost always works. They collapse because of three governance blind spots that leadership tends not to see until it's too late.

Blind spot #1: Treating digital transformation as an IT project, not an operational change

This is the root mistake. When the project is handed entirely to the IT function, it's implicitly treated as a one-time "software swap." But at a conglomerate, digital transformation is fundamentally about changing how departments and member companies record data, make decisions, and coordinate with one another. The consequence: the software gets installed on schedule, but the operational staff keep running their old Excel files because the real-world process never actually changed. The system becomes an expensive shell that no one truly uses. How to avoid it: the project needs an owner at the operational level - not the IT level - accountable for business outcomes, not just for getting the system "live."

Blind spot #2: Each member company is a data island

At a multi-industry group, each line of business - manufacturing, trade, logistics, real estate - typically builds its own systems for its own needs, at different points in time. By the time leadership wants a consolidated, group-wide picture, the data is scattered across systems that don't talk to each other. At that point, answering a deceptively simple question - "how much did the whole group earn this month, and where" - can take days of manual aggregation, and the final number is still suspect. How to avoid it: before standardizing core systems, unify the data layer first - shared definitions of customer, product, and supplier used across the entire group. Skip this step and every consolidated report you build later sits on a crooked foundation.

Blind spot #3: The ambition to "do everything at once"

A comprehensive transformation plan usually looks beautiful on paper: infrastructure, data, core systems, artificial intelligence, customer experience - all rolled out in parallel. The problem is that an organization can only absorb a certain amount of change in a given window of time. When everything changes at once, no unit can keep up, and the project bogs down on every front. How to avoid it: sequence the roadmap by value. Start with items that deliver visible results within a few months - automating a labor-intensive process, turning a report that used to take a week into one that takes minutes. These early wins build the trust and momentum the organization needs to take on the harder parts.

Closing thought

Digital transformation at a multi-industry group is not a problem of buying the right technology. It's a problem of managing change across a structure made of many moving pieces. None of the three blind spots above appear in any software vendor's technical documentation - yet they are what decide whether the project lives or dies. Dotbase distilled these observations from what we've seen and resolved while working alongside businesses with multi-company structures in Vietnam.