Michael, Pull the Plug

In January 2018, Liberty Global asked whether I would take a second project in the group. The first — Unitymedia's cloud communication platform — had just landed its first paying B2B customers. The second engagement was different. They had a programme that had been running for several years: a managed SD-WAN product for their B2B customers. Two offshore teams were building it. Three project managers had already cycled through the role and left. The product was not in market.

The ask was simple and entirely non-trivial: take a look, and tell us what to do.

Nine months, January to September 2018. The scope was diagnostic first, decision second — not the mandate to “ship it.” That distinction mattered more than anyone realised at the start.

Before I looked at the product I looked at the team. The pattern was recognisable and, by that point in my career, unsurprising. Posturing in meetings. Backstabbing in the corridors. Smoke and mirrors around status. Trust had been a thing once, and now was a sentence that always ended with “but.”

Three executives owned the decision: the SVP Product Management, the CFO, and the SVP B2B. They were not aligned. They were not even in disagreement in the productive sense. They were puzzled.

I could have made the engagement about the team dynamic. A lot of consultants do. I ran the workshops anyway — the team needed a container — but I treated them as a diagnostic surface, not the destination. What I actually needed was evidence.

I asked for architecture deep dives.

I went to Michael — the VP of Product Management B2B — and said it plainly: pull the plug. David, the SVP above him, had to carry the recommendation to the CFO and the SVP B2B. The conversation was not short and it was not comfortable. A multi-year programme, two offshore teams, a large sunk investment. But the question was no longer “can we finish this?” It was “will finishing this produce a product customers will buy?”

The answer was no. The evidence was specific. The decision was made.

The programme was retired.

1. Culture is a symptom. Architecture is a cause. Teams do not backstab each other for fun. They backstab each other when they are each protecting a position that is incompatible with the truth. Fix the truth and the positions dissolve.
2. Fall in love with the problem, not the solution. The SD-WAN programme had fallen in love with its own architecture. The problem customers had — reliable, managed wide-area connectivity — was entirely solvable. Just not by that specific build. The decision to buy, months later, from a vendor staffed partly by a former team member, proved that the problem had always been real. It was the solution that had to go.
3. The most valuable thing a product leader can sometimes do is write the obituary. Not every engagement ends with a launch. Some end with a clean stop, a recovered team, and a balance sheet that no longer has a €113M problem compounding on it. That is a good outcome. It is rarely the one people want on a CV, which is part of why it needs to be written down.