Five Thousand SIM Cards on a Shelf

1&1 launched a new mobile tariff. The All Net Flat. They were not sure how it would sell, so they had built the minimum required to take orders — the online sales funnel, the phone sales channel, and nothing behind it. No fulfilment. No fraud controls. No customer support flow. No delivery process worth the name. The plan, as plans of this shape usually are, was to scale the back of the house once they knew the product had a market.

The product had a market immediately.

I was brought in as an external consultant to the product management team one week after the launch. On arrival, there were 5,000 orders sitting on a shelf. Inbound was running at around 1,500 per day. The frontend was still accepting them.

A case of pure Product-Market Fit with zero Product-Company Fit. The market loved the product. The company could not deliver it.

The first question was not “how do we optimise this?” It was “how do we ship even one of these?”

Once the initial backlog was clearing, the actual brief opened up: build the processes that would have been built if the launch hadn't caught the company by surprise.

• Delivery. A repeatable, non-manual process for getting a SIM (and later a device) from the warehouse to the customer. Logistics pick-up windows, order cut-off times, tracking, returns flow.
• Fraud prevention and detection. The tariff attracted the kind of customer base that finds gaps in the rules. Fraud was not a theoretical risk — it was an active, ongoing cost. Two characteristic episodes: we received a phone return box, apparently containing a defective device, of exactly the same weight as the original. It contained chocolate. The weight match was deliberate. Separately, a postal-service worker was arrested for stealing several dozen phones in transit. We involved the police regularly. Fraud detection became a permanent, staffed part of the operation — not a project.
• Customer support. End-to-end flow: intake, triage, resolution, escalation. Built from nothing to a first-class support organisation.

1. The serial letter is the strategy. The instinct when you walk into an emergency is to reach for a framework. The correct move is often to reach for a working process, however crude. A mail-merge that gets a hundred SIMs into envelopes today is worth more than a delivery-architecture deck that ships next quarter. The framework can come after the envelopes.
2. PMF without plumbing is its own failure mode. A product that sells faster than the company can deliver is not a success story, yet. It is a scaling emergency with a nice headline. The win here wasn't finding the market — the market found us. The win was building, under pressure and inside a very small window, the company the product already needed to belong to.
3. Fraud is a permanent discipline, not a project. Any product attractive enough to find PMF quickly will find people who want to game it quickly. The chocolate-in-the-phone-box is funny until you count how many identical boxes arrived that month. Fraud work got built as a standing capability from the start — and the fact that 1&1 is still running a version of it, two decades later, is the proof that treating it as infrastructure rather than an incident was the right call.