Analyst: Can we start with the business case? What did the board actually sign off, and what did it assume on ARPU?
Expert: Autumn 2019 the first time. Sent back, went up again the following March with the curve pushed out. Blended mobile ARPU up four point two percent by year five. Sorry — four point two is the second version. The first one had something nearer five and a half, which is part of why it got sent back. Uplift came out of tiering almost entirely. Premium tier, speed guarantee, latency guarantee later on, people trade up, mix carries the line.
Analyst: And what did you realise?
Expert: Down. About six nominal across the same window. Six and a half if you strip the acquisition out.
Analyst: So a ten point miss on the central assumption.
Expert: Yes.
Analyst: How much of that is 5G failing versus the market moving against you?
Expert: Mostly the market, and I'll say that straight. But the case was written as if the market would hold still, and that's the bit I'd want your people to sit with. The downside case in the sensitivity table was minus one percent ARPU. Minus one. We did minus six. No scenario anywhere in that model for somebody buying share four years running and not stopping.
Analyst: Did anybody raise it at the time?
Expert: One person. Head of consumer pricing put it in a pre-read — tiering wouldn't hold, because you can't charge a man for speed when he already thinks his phone is fast. Words to that effect. She was right, and she was gone within about eighteen months, which is how that goes.
Analyst: Did the premium tier get traction?
Expert: Some. Call it nine percent of the base ended up on something carrying a 5G label, incremental about two euros. Most of it was bundled video, so we were selling telly on a network invoice. The radio didn't earn it.
Analyst: If the ARPU didn't arrive, where did the money go? The capex number very much did move.
Expert: Right. This is the part that doesn't get modelled from outside. Peak capex intensity was a shade over nineteen percent of service revenue. Pre-programme we ran thirteen, thirteen-eight. The plan had us back under fifteen by the end. We finished at sixteen and change.
Analyst: And the composition?
Expert: That's the interesting bit. Of what got booked internally as 5G capex in the two heaviest years, about a third wasn't 5G.
Analyst: A third.
Expert: Thirty, thirty-five percent. Ripping out kit we already owned and had paid for, because of the vendor security rules. Working radios. Perfectly good radios, on the back of a truck.
Analyst: I want to come to the enterprise line, because the company published a big number for 5G enterprise customers and I've never tied it to anything in the revenue —
Expert: Right. So. The enterprise number.