So an exchange marked as AO is meaningless , to a hell of a lot of potential customers,
In one sense, tommy45, you'hit the nail on the head here... the number of exchanges "going AO" isn't very meaningful in the big picture. And you get the reason right too - it isn't meaningful for most people individually.
The *only* way it is useful is that it acts as a marker in a marketing sense - it is one of the few ways that the average Joe Public can relate to the telecoms infrastructure (ie the exchange name).
It doesn't help us understand how the project is doing in the grand scheme of things - which is about what number of lines could now choose to order SFBB if they wanted to. For this we want a single number (the number of properties passed) or, at a pinch, the number of cabinets converted.
However, you make the same mistake as Bob_s2 except, where Bob uses the number of exchanges in AO as his indicator, you use the number of delayed cabinets to pick fault with.
Unfortunately, the number of cabinets that have been delayed tells us nothing about the number that have been converted - For example, 10% being delayed looks bad alone, but if 20% extra were converted early then things look better.
Even worse, we are only going by anecdotal evidence on here - how many people have complained. We're too small a dataset to have any meaningful statistical significance.
The only number we've got with any reliability is the number of homes passed, compared with the target numbers. Here's what I can find for those numbers:
2010: July: 1.5 million premises passed, 100,000 per week
2010: October: 3 million premises passed
2011: April: 4 million homes passed, 10,000 cabinets, 400 exchanges
2011: May: Target is 6 million by end 2011
2011: May: Target is now 16 million by end 2015, to complete commercial rollout
2011: June: 5 million premises passed (11 million to go)
2011: June: Passing 80,000 properties per week
2011: October: 6 million homes passed (10 million to go)
2011: October: Target is 10 million by end 2012
2011: October: Target is 16 million, now by end 2014, to complete commercial rollout.
2012: February: 7 million passed (9 million to go)
2012: May: 10 million passed (6 million to go)
2012: September: 11 million passed (5 million to go)
2012: October: 12 million passed (4 million to go)
2012: October: Passing more than 1 million per quarter.
2012: October: Target is 16 million by spring 2014, to complete commercial rollout.
2012: October: Target is 19 million after the current BDUK contract wins, no date.
2012: November: Press releases now refer to 19 million as the "two-thirds" target.
Over the 12 months to October 2012, they went from 6 million lines to 12 million.
In the eighteen months to Spring 2014, if they replicate that rate of progress, they could reach 21 million lines - enough to cope with the commercial rollout (16m) and all the BDUK contracts so far (now more than 19m presumably). Another year, another 6m, would take us well over the BDUK target, and there'd still be a year left after that.
It does seem that they can easily maintain 1 million every quarter. At that rate, they'd be able to reach 25 million by the end of 2015.
One thing of note in the stats above was how things slowed in the Oct-Feb period, and how they accelerated in the Feb-May period. Does that suggest weather hold-ups?
I'm not convinced they can quite keep up the 6m/yr rate - we're getting into the harder rural areas after all. But I don't think that the evidence suggests that we should worry unduly - and that wild speculation about penalty clauses (latching onto one semi-plausible factor) is, right now, just that - wild.