Truly savvy managers know the value of information. It’s the stuff intelligent decisions are borne of. But in recent weeks, the international community and the US Federal Government have been howling over the data collection efforts of the National Security Agency, making arguments as to whether or not those efforts are in the interests of US national security and whether or not data mining is an invasion of individual civil liberties. The concerns being raised may be misplaced. The major concern may not be with the data, but with the information being derived from it.
Information is distilled data. Distillation is a process that profoundly alters the natural state of the data. Anyone who has ever distilled data knows that context, sampling procedures, and data aging all play significant roles in the value of the information derived there from. As managers and executives, we need to examine four key considerations whenever we’re using data and information to make critical business decisions:
- Am I being provided with data or information?
- What’s the context?
- How was the data derived?
- How fresh is the data?
Data or Information?
First, we should always ask if we’re being given data or information. To make the distinction, it’s vital to look at the accompanying information that came with the accounts provided. See adjectives? See drawn conclusions? It’s not data. It’s information. And information disguised as data is frequently propagandistic, presenting a particular perspective or viewpoint. The US federal government has spent much time in recent weeks arguing that the recent data mining episode has strictly involved data, rather than information. If that’s the case, as citizens, the primary concern should be with context and the intentions of the distillation process. As businesspeople regarding the provided data from our respective fields, our concerns should be precisely the same.
What’s the Context?
“FIRE!” Shout that in a crowded theater, it’s a crime. Shout it on a rifle range? It’s a command. Shout it at a water balloon-throwing contest at the beach? It’s a reason to duck and cover. Without context, data is meaningless. But one of the great dangers is assigning context where none truly exists. As the government cell phone data collection scandal broke, I started thinking about the metadata that would be collected from a day’s worth of my calls. It would have revealed calls to three different airlines in the midst of a series of calls to a man to whom I hadn’t spoken in almost a decade. A sleeper agent being awakened in the midst of a plot involving downtown Chicago landmarks? A co-conspirator being briefed on a Chicago meeting?
Neither. In fact, despite the fact that the calls happened in rapid succession, they were totally unrelated. I had simply been attempting to reach a contractor about replacing a broken window. He had done the original installation 13 years earlier. And the Chicago calls? Unrelated! But since the sequence went: Contractor, Chicago, Contractor, Chicago, Contractor, Chicago, Contractor, it would be easy to believe that there was some kind of pattern. Context is crucial, but if we assign context where it doesn’t exist, we may make decisions based on specious, useless information.
How Was the Data Derived?
A big consideration here is what’s called the “observer effect.” That’s the notion that just by analyzing something, you can alter its state. I actually saw this recently with a few friends of mine who have sworn off the iPhone because you cannot remove the battery. They want to be able to be off the grid, and with the prying eyes of government analyzing their phone use, they feared a loss of their privacy. Would they have switched phones if they didn’t feel they were being watched? No.
When others come to us with data, we need to know the collection methodology. The observer effect may actually render the outputs useless. This is particularly true when anecdotal information is presented as raw data, which it’s not. Anecdotes inherently incorporate context, and therefore are a form of information. Information presented as data is inherently suspect.
How Fresh Is the Data?
Data has a very short shelf life. The greater the distance between data and distillation, the greater the challenge in ensuring effective contextual interpretation. Trying to explain why my 1970’s hair was down to my shoulders to my 20-something son is impossible. Too much time has elapsed and the context is far too unfamiliar. When presented with data or information, we need to push early for the full context, or the value will be lost.
As we gather intelligence for our own organizations, we need to be keen observers. Specifically, we need to ensure that we have a clear understanding of whether the inputs we receive are data or information. We need to know the context, the sourcing, and the timing. We need to watch for data presented as information and vice versa. And we need to be acutely aware that the sheer act of observation can sometimes taint what would otherwise be powerful business intelligence.