Thread by @colinrtalbot

100k tests.

A great example of why, sometimes, targets are extremely damaging.

Not, I hasten to add, because measuring performance is always wrong.

Targets and performance measurement are not identical.

Targets need performance measurement, but PM does not need targets.

Badly designed and used targets can encourage dysfunctional attempts to meet the target, but miss the point.

In this case an arbitrary target - 100k tests - replaced a clear idea of what, and who, we needed widespread testing for.

We needed enough to test, trace and isolate.

This obsession with 100k detracted from the second and third objectives. About which nothing much has been done for the last two months.

It has led to ridiculous measures to achieve 100k - such as a small number of large swab-taking hubs and even fewer test-processing centres.

We needed was a careful analysis of who needed testing to suppress the virus, and what action should follow testing.

We needed testing for broader surveillance purposes - to see where and when the virus spread. This was supposed to be being done but it's not clear it has been.

Once we had clear objectives we needed to measure how well we were doing getting there.

Instead we have this ridiculous distraction and diversion of Hancocks 100k (or Boris Johnson's 250k, which everyone seems to have forgotten).

I wrote a book about all this once....