[Sigia-l] IA Mythbusters - impact of redesign in traffic

[Ziya Oz]

Paola Kathuria:

> Without knowing the before/after design, there's no way
> of telling if a BAD redesign was the cause of the loss in
> revenue rather than just being redesigned.

Bingo, bingo, bingo!

'Failure' (whatever that means) could come as a result of a good design
executed badly, a bad business strategy executed well, a bad business
strategy to begin with, rapidly changing market conditions, or one of
1,678,355 contextual variables and permutations thereof. Or the opportunity
cost of not doing the redesign at all.

People looking in from the outside are rarely in a position to divine
causality and pinpoint blame. They may at best see the results of the action
but usually don't have access to the why and the how.

This pseudo-science, quant stuff really pisses me off royally. In general,
people who don't grok or care about causality, regression analysis or
randomized testing just shouldn't bother doing 'studies'.

-- 
Ziya

It depends.
If it didn't, you'd be out of a job.