Tim Wood Powell     The value of knowledge

The opportunities revealed by the Knowledge Value Chain derive mainly from taking data, enriching it, and applying it to support decision making in more innovative and value-enhancing ways.

But are there inherent characteristics of certain data that make it more valuable than other data? Absolutely.

Five factors come to mind as distinguishing one set of data from other, even before considering what further processing it has undergone:  timeliness, relevance, accuracy, novelty, and exclusivity.  In order more of mnemonic expediency than importance, we’ll call these TRAN(E).

Timely

Like radioactive atoms, all data have a “half-life” — a period beyond which their usefulness becomes progressively much more limited. They say that yesterday’s newspaper is good only for wrapping fish — because the “news” it contains is no longer “new”. As conditions change, informational descriptions of such conditions must change at a corresponding rate — or lose their potency.

If you’re a stock trader, stock pricing data is essentially useless by the time it’s published “for the rest of us” 20 minutes after the fact. By that time, any market-moving information is already reflected in the stock price. You’re too late to push the button.

Relevant

If value of data is perceived in relation to its applicability in making a particular decision (as the KVC postulates), then the data must be germane to that decision. This is harder to measure than timeliness, and consequently it’s here where many decision processes come unglued. Often this is because the process owners are trying to judge the relevance from the bottom up (data), rather than from the top down (value, result, or benefit).

Accurate

The data must be correct, focused, factual, unbiased, representative. It must be “the whole truth, and nothing but the truth”. Probably no data is 100% accurate — just as no data is 100% timely or relevant. And the requirements for accuracy (and corresponding tolerances for inaccuracy) vary considerably depending on the application of the data.

New

In addition to being timely, the data must be NEW or novel to the recipient. This may be related to, but is not the same as, its timeliness. If data, no matter how fresh, tells us something we already knew, then — regardless of the amount it cost to produce it — it is redundant and has little informative value. (This criterion derives from Shannon’s mathematical theory of communication.)

(Exclusive)

In some, but not all, cases, the data must be EXCLUSIVE to have value. In this way it differs from the other four characteristics. Not all data must be exclusive to have value. Data about, for example, which prescription drugs interact badly with other drugs has value to everyone who needs that information in order to act on it (your doctor or pharmacist, for example). The data scales across applications, and is no less powerful because other people have it too.

Other data, however, derives its core value from exclusivity. This forms the basis of the “insider information” about stock trading that is highly valuable — and highly illegal. If I know for a fact that company X is about to be acquired, and few other people know it too, I can buy up shares of X at a relatively low price.  Once the deal is announced, and the information becomes more widely known, the price of X stock typically goes up considerably (to roughly the acquisition price) — and I have lost my advantage. Economists call these “information asymmetries”.

Timely, relevant, accurate, new, exclusive data starts with a solid basis in value — which can then be further enhanced by processing according to the KVC framework.

---