Why 70% of Transformations Fail

I once walked into a program launch where twelve teams each had their own roadmap. Twelve definitions of success. Twelve slide decks, each formatted slightly differently, each containing a Gantt chart that bore no relationship to the Gantt chart in the room next door. Everyone was busy. Nothing was changing.

The CEO asked me how the transformation was going. I said it depended on which of the twelve transformations he was asking about. He did not find this funny. (It wasn’t meant to be.)

Whether the exact failure rate is 70% or 57% or some other number that a consulting firm reverse-engineered from a survey of people who had reasons to exaggerate, the pattern is the same: transformations fail for the same human reasons, repeatedly.

The strategy is usually fine. The execution plan is usually fine. The thing that breaks is the space between the plan and the people who have to carry it out. Here’s what I’ve seen separate the ones that work from the ones that don’t.

Start with one sentence

Before the governance model. Before the org chart. Before the 47-slide kickoff deck. Answer this: what is the one sentence that describes success? Not a paragraph. Not a mission statement that was wordsmithed by a committee until it meant nothing. One sentence. A director should be able to say it without looking at notes.

Is value coming from revenue, cost, or risk reduction? Which one? You have to pick. “All three” is a non-answer that guarantees conflict at the first prioritization meeting, because the initiative that drives revenue growth will eventually collide with the initiative that drives cost reduction, and somebody has to decide which one wins.

That sentence governs every governance discussion, every status update, every prioritization call. Work that doesn’t serve it gets moved to “later.” Later is a fine place. Most things belong there.

Focus is a resource. Spend it deliberately. The first 100 days should have one sentence that every workstream traces back to. If you can’t draw a line from the task to the sentence, the task is either wrong or the sentence is.

The sponsor has to actually sponsor

Sponsorship isn’t a name on the first slide of the kickoff deck. I’ve worked with sponsors who attended every steering committee meeting and sponsors who sent a delegate to the first one and were never seen again. The correlation between sponsor visibility and transformation success is not subtle. (For a deeper look at what real sponsorship requires in practice, see what PROSCI certification doesn’t teach you.)

What “actually sponsor” means: decisions returned within 48 hours, not weeks. Showing up to working sessions, not just steering committees. Willingness to stop work that isn’t delivering, even when stopping it means admitting a bad bet. Especially when stopping it means admitting a bad bet.

When leadership is consistent, resistance fades. Not because people are persuaded. Because people stop waiting to see which way the wind blows. The direction is clear. They can either walk that direction or not, but at least they’re making a real choice instead of hedging.

Adoption, not go-live

Go-live is not success. Adoption is.

I’ve watched organizations spend 18 months building a system, celebrate the go-live with a cake and an all-hands email, and then discover six weeks later that half the organization is running the old process in parallel because nobody showed them how the new one works for their specific job. The system was live. The change was not. (If you work in demand planning and felt a chill just now, that’s because this paragraph is about you specifically. Every demand planning transformation I’ve touched has had a shadow spreadsheet that outlived the new system by years.)

The difference between systems that get used and systems that get worked around comes down to how you design the change. Map what changes for each role, not just which buttons to click. Build training around realistic scenarios, not feature tours. (“Click here, then here, then here” is not training. It’s a screen recording with narration.) Equip managers to coach their teams through the transition, not just forward the training email.

Leading indicators matter: Are people using it consistently after the first two weeks? Are they getting better at it? Has the grumbling shifted from “this is terrible” to “this is annoying but fine”? That shift is the signal. Watch for it.

Governance should be boring

That’s the goal. If your governance is exciting, something is wrong.

One place for documentation (we use a centralized program tracker for this). One cadence for reporting. One log for decisions, timestamped, with the name of the person who made the call. No shadow decision-making in hallway conversations. No surprises in status updates. Finance at the table from day one, so conversations stay grounded in value instead of drifting into opinion.

When everyone sees the same picture, energy goes toward solving problems instead of debating what the problems are. The twelve-roadmap problem I walked into at the beginning of this piece was a governance failure. Not because the governance didn’t exist. It did. It was a gorgeous framework, documented in a 40-page playbook that I’m sure someone was very proud of. But nobody used it, because it was designed to be admired rather than operated.

The best governance systems I’ve built are ugly. They’re a single shared document, a weekly meeting with a hard stop, and a decision log that someone updates in real time. No frameworks. No matrices. Just: who decided what, when, and why. Everything else is decoration.

But even good governance fails when the underlying organizational design hasn’t clarified who has authority. A decision log that says “pending” next to every major call isn’t a governance problem. It’s a structural one.

---

If your transformation feels heavy on motion but light on progress, let’s talk. We help organizations redirect energy toward outcomes that actually stick. For a structured approach to the first 100 days, see the complete integration checklist.