So there we are: it is and can be awesome. But in too many corporations games -- especially lack of openness -- prevents things from going anywhere. And that, friends, is a darn shame.
As I mention time to time on HN the best, short, spot on reads that take this to its core avoiding simplistic conclusions while arriving at doable, realistic things are two:
https://www.amazon.com/Human-Element-Productivity-Self-Estee...
https://www.amazon.com/What-Total-Quality-Control-Japanese/d...
In different, self reinforcing ways, there's a confluence: those with self awareness will figure it out nicely summarized by Schutz's Jung quote:
"All the greatest and most important problems in life are fundamentally insoluble ... they can never be solved but only outgrown. This 'outgrowing' proves upon further investigation to require a new level of consciousness."
This dovetails with one of the central insight's in Human Element: the biggest impediment to teamwork are rigid individuals. The how, why, and what next? Gotta hit the books. Good grief: used copies of the book can come to your front door for less than a Padron cigar or 15 year Laphroaig single. That's value.
Off subject but a particularly funny example of can't-change is the Angelica Huston part in Buffalo-66, an absolutely outstanding film.
Now, the author rightly argues that coming up with RCA isn't easy. In particular it tends to follow a twin inverted V shape:
<--- starting wide various technical/operation issues considered
+ <--- narrowing to the root"ist" technical cause
/-\
<--- widening back into organization issues as to why an
agent of change could do or did do something
Even the book on TQM in 6-sigma work:
https://www.amazon.com/What-Total-Quality-Control-Japanese/d...
points out that technical issues often dissolve into organizational issues. See page 57-58 then elsewhere.
So where does that leave one? Dealing with the many contributing factors in an organization that impinge on decisions which may lead to bugs/outages/defects:
- Eliminate opportunity defects. Simplify.
- Inputs to the next process should be controlled (in manufacturing parlance they are X-sigma quality). It's harder for me to screw something up if what I start with conforms to its requirements aligned ultimately with customer satisfaction.
- quality is everybody's problem. See again pg57-58. Sectionalism is a major impediment to enterprise wide improvement
- the ultimate aim is continuous improvement of which RCA (fish bone diagrams and the rest) are but tools. The salient business question is: ok, a client outage occurred. Not OK, but does it repeat?
Root cause analysis? Plants are doing just fine. Human organizations can be helped by root cause analysis. That's just how it is.
However, it takes street smarts, experience, and being burned a time or two to know when it's two guys (associates) talking trying to get it right and when you're being played by a player.
Flailing individual trust precedes the decline in institutional competency. And once that happens the prospect for legit corrective action is largely gone. In-fighting will dominate.
Clues that maybe something is off (not exhaustive) based on some guys I worked with a few years ago:
* The boss talks with team individuals only then brings the team together to report what everybody said, and what everybody agreed to. Actually, no individual knows what anyone else said and didn't have a chance to ask questions.
* You're afraid to bring something up, or notice privately team members complain or share concerns but never say anything to decision makers or in a team setting
* You're encouraged to make changes because of "reputation," team image or other nebulous sounding reasons. Change may indeed need to happen. But those are not reasons
* You're instructed to remain technical to not engage or discuss issues that seem more pertinent to management, organizational values. Being boxed in is a bad sign.
* Being brow beat into doing things out of guilt or entitlement are bad signs.
* Putting internal politics or good light on the boss' project ahead of customer satisfaction even though the work will not satisfy intended customers is a bad sign.
* Abusing through manipulation internal service supplier status is bad. One guy I knew slickly sold his team's capability knowing once tied in they were practically married. The code was implemented such that the customer had no insight or control to it. By running around the floor and vacuuming up as many internal teams as he could, he parlayed his position to "take over the floor, but taking over the data." Guess what? It failed in a spectacular fashion; he never took over control of that system and never gave his "customers" a better solution. He treated his customers as suckers.
* The above point has a corollary: internal service suppliers get implicit management support, and it's implied internal customers have no choice but to work with them. This is NOT how it works on the outside. I choose my roofer and if I am not happy payment may not be given or is disputed or I fire the roofer. The hole in my roof is my motivation to treat the roofer right. My checkbook is why the roofer needs to treat me right. Either can walk away if the other is a jerk. But not in large corporations. That's why, for example, deprecating private data centers for the cloud are horror shows. The internal service suppliers are used to being in charge even entitled no matter how crappy or expensive their service is. And when they find out they might be replaced, they fight it all the way.
See:
https://www.amazon.com/What-Total-Quality-Control-Japanese/d...
https://www.amazon.com/Human-Element-Productivity-Self-Estee...
https://www.amazon.com/FIRO-Three-Dimensional-Theory-Interpe...
and any work from Deming, Crosby, or Drucker which do far better at integrating business, profit, people, and customers.