Twenty-five years of digitalization where the technology, the systems, and the business model all had to reinvent themselves at once.

A fixed-line strategy for Germany that had to design a multi-year, multi-billion-euro investment path and a risk-sharing architecture for a natural monopoly — before the rollout had even begun.

The path I sketched — VDSL → FTTC → FTTB → FTTH — and the contingent-model logic that BNetzA and the Monopoly Commission later reviewed are still load-bearing sixteen years on; along that path, DTAG has cumulatively invested in the order of €17 billion.

Three overlapping partner products, three parallel operations paths, and a looming write-off from a failed state-client project — a cross-functional problem that had been addressed as a portfolio question.

40,000 external customers in twelve months, 20,000 internal connections to EWE itself, largest single deployment of 8,000 employees, roughly two-thirds shorter time-to-provision, and a Best UX Award for the integrated self-service platform; the write-off was absorbed through the partnership architecture.

MiFID II recording, Basel II, the bank's IT-security and codetermination of the works council — for 1,200 employees and ten trading desks at 99.99% high availability across two sites — to be solved simultaneously, in twelve months from bid to acceptance.

Bid won against formally stronger competitors via a persona-driven demo from the daily life of the CIO's assistant; the platform went into production in December 2016, and the codetermination conflict over recording classification was resolved by capturing securities IDs at the trading desk during the call rather than by manual post-processing.

A CEO expectation to follow the competitor on advertised AI functionality — specifically predictive maintenance — without it having been tested whether the competitor had substance behind the website, or whether customers wanted the feature at all.

Customer interviews showed that customers wanted preventive maintenance and a better spare-parts logic, not predictive; the roadmap was adjusted accordingly, the AI feature was not added to the product, and the four-person software-PM team was lifted over six months into a more sustainable position inside the hardware-shaped corporation and handed over to a successor.

A multi-year SD-WAN programme in which three previous project managers had failed — and where an architecture deep dive showed that only about 10% of the bandwidth provisioned per connection was actually usable for paying customers, the rest going to management overhead.

Recommendation to retire the programme; the board carried it, realising in the order of €113 million as sunk-cost write-off and avoiding a substantially larger follow-on investment in a non-viable product — Liberty Global subsequently bought an SD-WAN solution externally.

An early-stage AG whose risk order was financing > market > product — and in which a product function, however cleanly built, cannot substitute for the missing financing and incorporation setup.

Consulting portfolio and PM function built and handed over to the management board in three months, launch event with 150+ participants; the company was subsequently dissolved — the two transferable lessons (risk order in an early-stage AG; AI adoption as a change sale) carry beyond AAACHT.

An established B2B telecommunications carrier wanted a second pillar in the B2C mass market — on a margin and process architecture that doesn't carry B2C unit economics, without the executive board willing to authorise the investments needed to bridge the gap.

The product was built and made available on the website; to the best of my knowledge no meaningful customers were won — the diagnosis dismissed in a C-level meeting as “painting it too black” confirmed itself in the commercial reality.

A best-in-class real-time Java runtime without the product frame that would reach a tier-1 buyer in the connected-vehicle market; internally, the strategy debate over deeper-runtime versus broader-platform capability cut across the executive team.

33 months of repositioning as a V2X platform partner, double-digit-million revenue lift, COVESA industry award, CES 2022 demo built in three months with NXP and AWS; Stellantis re-tiered aicas to tier-1, and the software now ships in more than 5 million vehicles per year.

A PE-backed multi-country integration whose C-level strategy was framed at an abstraction layer that didn't catch a licence server under a product manager's desk — nor the competitive geometry between the existing customer bases.

Sales and operations integrated, ERP rolled out across one country, product synchronisation reduced to the non-conflicting customer bases; the previous consulting firm lost its mandate four weeks after the steerco — Enreach is the market leader in its segment in 2026.

Three years of internal engineering on a cloud B2B-communications product that hadn't shipped — engineering optimised capability, product perfected the spec, no one carried the launch.

Voiceworks integration in nine months: first paying customer at month nine (installed personally, by train), several hundred B2B customers within twelve months of launch; total spend ~€600K against ~€1.2M still required just to finish the original internal build, and the billing-integration gap was reframed as a free-base-fee acquisition window for early adopters.

Competitors with hundreds of high-street shops were selling certainty, speed and a single point of contact — 1&1, as a pure online and phone provider, had no comparable answer to the security gap customers felt before signing a contract.

Four contractually binding service guarantees (overnight delivery, 30-day money-back, priority hotline, 24-hour replacement), monitored against ~75 KPIs and reported daily to senior management for six months — share price up ~25% in 24 months; the Prinzip is still 1&1's public brand promise in 2026.

A new mobile-tariff category that had found its market acceptance before the delivery, fraud and service machinery to support it — 5,000 orders sitting on a shelf, 1,500 orders inbound per day, one week after launch.

Delivery, fraud-prevention and customer-support functions built from a manual mail-merge on day one to a productive operation within months; each of those functions is the precursor to what 1&1 still runs in mobile in 2026, two decades on.

A complete technological renewal of the German fixed-line network, in which voice and data could not coexist within a thirty-year-trained sales logic — a workshop of four internal participants and sixty external consultants marked the moment when training as a lever was ruled out.

The end-to-end process setup across all channels, encapsulating the new voice-and-data dependencies in systems rather than in heads, still carries Germany's mass-market fixed-line business sixteen years later — through two technology generations and without observable structural adjustment.

A sales organisation whose weekly complaints list — missing materials, demos, test licences, lead quality — systematically prevented the performance of the individual salesperson from being attributable.

The full list cleared in twelve weeks; the strong salespeople performed better, the weaker ones stayed where they were — the diagnosis of the sales organisation was on the table without anyone having to speak it.