Triggered by Greg Young's "Stop Over-Engineering" talk at
Buildstuff Odessa,
The recording is currently unavailable,
Here another version of his talk at DDDx in London
https://skillsmatter.com/skillscasts/8082-stop-over-engineering
freq fail * (cost/fail)
------------------------ = CBA ratio
(Investment / period)
One of the cost of failure component at my current client is Reputation.
Some of this Reputation loss definitively has
a clear cost associated to it.
For some other it is less obvious.
Especially, because we have a lot more complaints about the systems in place.
And the kind of complains (not complaints imho) we are getting do require a lot of analysis
- to verify the claim
- find out if they are substantiated or not
This of course, from our point of view.
And we take all of them very seriously.
The worst I've had up to now was, at its core, about whether a mail was sent or not.
That ended up with sending logs of the respective mail servers.
And me explaining to business that SMTP does not guarantee delivery...
The bizarre thing is that most of the complaints we are receiving are about parts of the systems that have not changed fundamentally for a long time.
The question is what has changed then ?
Recently ( a few month ago) we have been a lot in the news, there have been audits.
The conclusion of the audits are, for as much as I can say, quite motivated and have a lot of nuance in them.
But the abridged summary reported in the news, were done in a semi-sensational fashion.
BAP has been literally ruling us for some time.
So the cost per failure may have risen, the investment not.
The frequency is definitively higher.
The question that I have in my mind is
Do we do something about it, in our systems,
some of them Legacy that we'd rather not change.
Or do we just wait for the storm to pass ?
P.S.
This is one of the reasons why my favorite topics lately is dealing with failures.