Decide big, buy small: two questions before you slice a transformation

Written by

in

Everyone still pitches the big transformation. Most of what I actually see getting bought doesn’t look like one.

A colleague asked me recently why we run one big transformation instead of a series of micro-transformations. I’ve been turning that over since. It’s a better question than it first sounds.

In the conversations I sit in, the big target picture hasn’t disappeared. It survives as a budget construct, not as a buying form. What gets signed is usually the extension of work already running. Decide big, buy small.

Where slicing breaks

Two things push back hard against slicing, and neither is soft.

The first is the money. A board approves a business case, not a stack of them. Every slice that has to be justified again spends the sponsor’s political capital again. That cost is real, and it shows up in no plan.

The second is everything without a middle. A core platform replacement. A data migration with a cutover date. A regulatory deadline. Running old and new side by side costs real money every month. Slicing doesn’t remove that bill. It lengthens it. A cutover date doesn’t care how neatly you scoped your pieces.

Two questions before I cut

So before I agree to cut a transformation up, I ask two questions.

One: can old and new coexist for as long as the slicing takes, with no date at which they no longer can?

Two: can the first slice deliver something a business owner would sign for on its own — even if nothing ever comes after it?

If either answer is no, it isn’t a slice. It’s a project phase with a marketing name.

And even with two yeses: pieces without one shared target architecture are a queue, not a program.

That’s the test I apply. I’d like to know yours: what do you check before you agree to slice a transformation — and what makes you insist on cutting it whole?