Principal Analyst – Data Analytics
Summary Bullets:
- To be ‘data driven’ is to be story driven.
- Understanding data depends on the knowledge brought to it.
It’s stylish these days to be ‘data driven’ even while almost no one talks about what that really means. Data is just a proxy, the spokes on the wheel, the shorthand for what’s really going on. What’s the real driver?
During a blizzard, for example, we may talk about degrees Fahrenheit or Celsius. But that’s shorthand for the cold and snow, which is what actually drives decisions about what to wear or whether to go out at all.
One vendor that seems to look behind the proxy is Narrative Science. At in early May event, several speakers helped explain what it is to be truly data driven. Two in particular were notable.
Narrative Science co-founder Kristian Hammond pointed out seeing behind the data proxy requires us to infer the facts that drove the data. We understand the data based on our own inferences that spring from the schema in our heads. For example, suppose you’re told that Mary Smith is Stanford University’s premier roboticist, she’s on her way home at the end of one of her usual 80-hour weeks, and she stops at a bar for a drink. You’re told nothing about her age or whether she’s able to be served. You just infer it — which is a big part of the story.
His point: The human does the work of understanding the facts behind the data. No matter how advanced the data tool may be, it can’t do the understanding for you. To be truly data driven is actually to be driven by inferences that arise from a lifetime of knowledge.
Data plus inference makes a story. This is the basis of data storytelling. Stories are the meat on data’s bones.
Data expressed in data stories becomes valuable when it’s shared. Storytelling is conversation, according to Donald Farmer, principal consultant at Treehive Consulting and renowned veteran of product development in the data-intelligence industry.
Every conversation, over any medium, requires some degree of trust. That trust gives underlying data its credibility or lack of it.
Trust is so critical that to be driven by data alone is worth nothing. Data is so plentiful now that one could prove just about anything with it. Only trusted data is valuable.
Do we audit all the data that drives us? No, of course not. There’s no time. We just trust the source, the people who delivered it. Farmer, channeling retired data industry eminence Scott Davis, observed that people don’t trust data; they trust other people.
To be driven by data requires trusting the source. We might call that the data’s lineage, but informally, the lineage is simply every person who managed the data from its origin.
Trusted data used in decision making is actually compiled stories made of trusted data plus inferences.
To be data driven is to be story driven.
IT Connection 