Of Mess and Trust in a System
The amount of trust you put in a system directly relates to the amount of care you dedicate to it.
Let me introduce you to this relation through a few examples.
Incomplete calendars
Be it the family planner on your fridge or a digital calendar system, the amount of trust you put in it depends on whether you update it regularly, whether everyone puts all the expected stuff on it. That is, whether the stuff on it represents the reality of the family.
Past a certain threshold of mess or lacking in there, the trust goes down. Way down. And this has nothing to do with a linear decrease.
Past a certain threshold of mess— and, although I have no precise data, my best guess is that it lies around 2% to 5% of wrong or missing entries —we lose most trust in a system.
It’s the 5% of mess that kills the 80% of trust.
Back to our example: say the family calendar holds more than 100 events, I believe it suffices that 5 of them are skewed or missing to have most of us distrusting the calendar and, most likely, consequently stop updating it, thus increasing the amount of mess, and resulting in greater distrust.
Partial todo-lists
Same for your todo-list, it suffices that you are aware of a handful of tasks that do not lie in there to make you lose your trust in it because you know it does not reflect in reality what you need to do.
Slightly buggy pieces of software
For the programmers among us: at what point will you consider that a whole software code is a piece of rubbish? Often, only after having found a handful of glitches, naughty bugs and ugly lines of code. A handful of buggy lines among, most likely, thousands of working lines of code.
For software users, all the same: when do we give up faith in a software? After which number of files lost or buggy functionalities? Probably below 10%.
Other systems
Examples abound, also in a less knowledge-work related context. Think for yourself:
- After how many hiccups do you lose trust in your car’s ability to do the next ride?
- After how many delays do you lose faith in <name your not so favorite train company or airline>?
- After how many errors or incoherences will you distrust this article and tis author? ;-)
My bet is that, in all those thought experiments, you will land under 10%. You don’t need many hiccups or delays in order to lose trust, just a handful of them.
How many unfortunate acts from your trusted ones before you lose trust in them? Here also, the loss curve isn’t linear at all.
So what?
There are systems you don’t control, like the train system, and systems you do control, like your calendar.
For the systems you control, do realize that the amount of care you dedicate to maintain the system quality will have a vast impact on your ability to trust it, and thus the value you see in it, thus the care you give to it, …
The “5% mess — 80% distrust” hypothesis stated previously makes the above system cycle a tough one. Better not hit the 5% mess threshold.
The above words might make you think that any system ends up being distrusted and useless.
Systems need care
Now, suddenly decide, against all odds, to take 30' once in a while to revise all coming entries of the family calendar with all parties involved, or clean up your todo-list. Or “accidentally” spend a bug-fixing sprint with your teammates. Or invest in a trip to the garage for your car and a new engine.
What would you be doing? You would be taking care of the system. A strong, timely investment. Which will end up in both regaining trust and lowering messiness.
Pay attention to your systems, notice when mess builds up, notice how shortly after trust goes down and ultimately realize, that it is time to take care of the system.
Artwork : National Archives Catalog
