Version 138

The VP of Product Management gave me a clear, two-stage brief.

First, a detailed market analysis: what was happening with fibre rollout in other countries? Which business models were prevailing? What role did open-access models play in markets that were ahead of us? Which models had not prevailed, and why? International, careful, with the understanding that Germany must not miss the cycle — and that we were still early enough to choose the strategy rather than be overtaken by it.

From that research, the second step: a portfolio and rollout strategy for the German fixed-line network, including the question of how that strategy should be financed and how its risk should be borne.

The result was a strategy paper that ran through 138 versions during the months of preparation before going to the board for decision. The version number isn't accidental — it is the most honest description of what producing a strategy of that magnitude in a corporation of that complexity actually costs.

Protect the existing copper network. But bring fibre, step by step and under economic discipline, closer to the customer.

Operationally: VDSL as the first extension. Then FTTC — fibre to the cabinet. Then FTTB — fibre to the building. Then FTTH — fibre to the home. Incrementally, not as a big bang, because the investment blocks and refinancing cycles of fibre infrastructure don't permit a big bang.

The second, distinct contribution of this engagement was the contingent model. The logic in one sentence: if DTAG invests hundreds of millions in fibre, and fibre is a natural monopoly — which under regulation must be opened to competitors sooner or later — then DTAG carries the rollout risk alone for an infrastructure that will be shared anyway.

That is an unhealthy risk asymmetry. It leads either to too little rollout, because the risk deters, or to overinvestment, because the risk isn't yet taken seriously enough in the early phase.

The contingent model redistributes the asymmetry. Companies that participate in the rollout — through capital, through advance commercial commitments, through operational contributions — receive better terms per line in service. Skin in the game is the entry ticket to preferential conditions. Those who carry the risk benefit. Those who don't enter at standard terms after regulatory opening.

In effect: the model turns a monopolist's rollout risk into a marketable asset class, in which the Telekom finds co-investors without giving up operational control.

We discussed this model very early. At the time, there was no widespread fibre rollout — only individual pilot areas. Dresden is the anchor I remember from those discussions.

1. A version number is a receipt, not a flaw. Reaching v138 isn't a sign of indecision; it is the honest measure of how many assumptions, sensitivities and stakeholders a fixed-line strategy for a country with Germany's political, regulatory and market structure actually contains. Whoever takes v3 to the board hasn't understood the complexity. Whoever takes v138 has absorbed it.
2. The contingent model is, in effect, a general answer to a recurring question. Whenever a company has to make substantial investments in infrastructure that cannot, regulatorily or structurally, remain its exclusive use later, it pays to distribute the risk at the front rather than absorb it at the back. The mechanism — preferential terms for early co-risk-bearers — transfers well beyond fibre.
3. A rejected recommendation is not necessarily a wrong recommendation — and the market is the most honest court of appeal. Open access in the form I proposed in 2009 didn't find a majority within the Telekom in 2010. With the benefit of subsequent market development, it could have played a larger role than it did. To document that — including the rejected portion — in a strategy is part of professional responsibility. To write strategy means to record what the organisation, in that moment, doesn't want to hear.