When the meta-data matters

I've often found that when dealing with both people and data, the meta-data is more useful, more revealing and more interesting than the data, or the person themselves.

This is not an absolute law of people and data but when I orient my approach to thinking about the meta-data first, I find I uncover more value, faster.

People

Let us start with people, I generally find that discovering:

What drives a person, how they feel they are disadvantaged, their habits, reveal more than what I would term the "official" [If this were a work colleague, this might be an email or meeting, ] interactions I have with them.

I ask myself the question "how does this person want me to feel?" and the answers to that help the way I interact.

There is a sub-text to (almost) every interaction you have with someone (perhaps besides your closest family or friends).

  1. What questions are they not asking that they are trying to get the answer to?
  2. What is important to them that they need me for?
  3. Are they trying to help but don't want to appear like they are handholding?
  4. Are their personal ambitions/goals driving the interaction over the wider goals of the organisation/family/team?

In a strange way, we are all good at what I call "smelling insincerity", it's funny because if people knew how everyone else knew they were being insincere, they wouldn't try half as much to do it because it's only ever a net-loss to them.

This is why I have found, when I am unsure of the subtext or motivations, it's often disarming just to ask, as if the other person can now start to be "themselves". It also saves time as you quickly establish the "ohhh, you just want x".

Data

Meta-data about data that is often not discussed:

  1. How was it collected? (e.g. legally, manually, in an automated way, user submitted, generated)
  2. How accurate it is? (e.g. when was it true, is it out of date, is it predictive data, is is retrospective data, how reliable are the sources)
  3. How hasn't this data been used before? (e.g. everyone uses it like x, can it be used with y. Can it be combined with new data, can it be used to validate other data?)
  4. What is the risk-level of this data? (e.g. if we become dependant on it, can it go away? Can others use it in the same way easily to remove any advantage, will any groups complain if they discover we are using this data?)
  5. How easy is it to manage this data?

I suppose a simple reduction of this concept is "what is this person/data telling me" that I am not listening to, that is or could be important?