I have led transformation programs where the technology was flawless and the outcome was still a failure. The variable that decided it was almost never the platform. It was what happened in the first 90 days, before a single process was redesigned.
Most programs are lost in the first quarter, not because leaders do the wrong things, but because they start building before they understand the ground. Here is the sequence I run, in order, every time.
Days 1–30: Listen before you touch anything
The temptation is to arrive with a plan. Resist it. Your first month is for evidence, not action.
On my last ERP rollout I ran 34 structured interviews across finance, supply chain, and the plant floor in the first three weeks. No slides, no solutioning — just questions and silence. Two patterns emerged that were absent from the official project brief. One of them would have sunk the go-live if we had built to the brief.
Listening is not soft. It is the cheapest risk-reduction you will ever buy. An hour of interviews costs you an hour. Discovering the same problem in UAT costs you a month.
- Interview across levels, not just sponsors. The person who keys the invoices knows things the VP does not.
- Ask what breaks today, not what they want tomorrow.
- Write down verbatim quotes. You will need them later to prove you heard.
Days 15–45: Map the real system, not the org chart
Every organisation has two structures. The formal one on the wall, and the real one that gets work done. Transformation lives or dies in the second.
I map three things in parallel: the process flow as it actually runs, the data as it actually moves, and the influence network — who people genuinely trust when they are stuck. That last map is the one nobody hands you and the one that matters most.
The org chart tells you who has authority. The influence map tells you who has permission. You need both, and they are rarely the same people.
Days 30–60: Build the baseline before you promise anything
You cannot claim a win you cannot measure. Before I commit to a single target, I lock a baseline: cycle times, error rates, manual touchpoints, cost per transaction — whatever the program is meant to move.
On one program the "known" invoice-processing time was quoted at two days. The measured baseline was 6.5 days end to end. That gap became the single most powerful number in the entire business case, and we would never have had it if we had trusted the anecdote.
Baseline discipline does two things. It protects you from over-promising, and it gives you undeniable proof when the numbers move.
Days 45–75: Ship one quick win that people can feel
Credibility is a currency, and in a transformation you start with almost none. You earn it by delivering something visible and useful before the big system arrives.
I look for a fix that is small in effort, high in daily annoyance, and owned by the team that is most skeptical. On a recent program we automated a weekly reconciliation that four people hated — three hours of manual work gone every Monday. Cost: two days of build. Return: those four people became advocates who defended the program in rooms I was not in.
- Pick something users touch weekly, not the executive dashboard.
- Solve it fully, not partly. A half-fix erodes trust faster than no fix.
- Let the team take the credit. Your goal is momentum, not applause.
Days 60–90: Form the coalition that outlives you
No change leader scales alone. By day 90 you need a coalition — a small, cross-functional group with real credibility who will carry the message when you are not in the room.
I keep it small: six to eight people, chosen for trust rather than title, drawn straight from the influence map. I give them early information, a genuine voice in decisions, and public ownership of wins. In return they absorb resistance long before it reaches me.
A coalition is not a steering committee. A steering committee approves. A coalition advocates. One meets monthly to review status; the other defends the program in the corridor on a Tuesday afternoon.
What the first 90 days actually buy you
At the end of the quarter I have not shipped the transformation. I have something more valuable: a program built on evidence instead of assumption, a measurable baseline, one proven win, and a coalition of trusted people pulling in the same direction.
Everything after this is execution — and execution is easy when the foundation is honest. Skip these 90 days and you spend the next 900 paying for it.
The first 90 days are not the warm-up. They are the program.
Cédric Bignet is an AI & ERP Change Management expert at Novartis and founder of AInspire. He writes about change management, AI adoption and enterprise transformation.