Tim Wood Powell     The value of knowledge

More ideas from the SCIP09 conference in Chicago…

I’ve worked a lot during my career with business-to-business publishing and information services clients—and consider myself in that business as well.  There are basically two business models for this kind of information—customized and syndicated.

Customized information is that which is developed—often at great effort and expense—and which is highly tailored to the needs of a very small audience or number of “users”.  A consultant’s report is a good example.

Syndicated information is that which is developed—again often at great effort and expense—but that is intended to meet the needs of a much wider audience.  A magazine or a mass-appeal web site are good examples.

The economics are of course quite different.  In both cases, there is a “build once”.  But in the customized case, it’s “build once, sell once”.  In the syndicated model, it’s “build once, sell many times.”

Mass production is one of the key things that distinguishes a cottage industry from a modern business.  In the early 20^th century, Henry Ford revolutionized the auto business by moving from “job shop” one-at-a-time production, to mass production based on assembly lines.  The economics were so different that auto workers, whose former relationship to their product was only to MAKE them, could now also afford to BUY them too.

The net result for the industry was that the hundreds of auto manufacturers using the customized model begin consolidating, and then were forced out of business.  They simply could not compete with the newer model—which we now take for granted as standard operating procedure.

Everybody knows you’ll go broke running a car company the job shop way.  Yet it’s amazing how many of us run our “information businesses” the exact same way—then wonder why we run into trouble proving our value propositions.

Comments made in two Active Dialog sessions at SCIP09 in Chicago (one moderated by SCIP Board member Scott Leeb, the other moderated by me) confirm my own observations that there’s a mix of these kinds of activities in the typical organizational intelligence function.  That is, there are some activities that are “syndicated”—typically, a report that goes to a group of decision-makers.  For example, I work with one company that has a weekly report that goes to the senior management team of about a dozen people.  In another company I work with, there is a report prepared daily for the company’s top 400 executives globally, starting when the New York stock markets close at 4pm ET.

In both of these companies, there is also customized ad hoc work done on request for a given decision-maker.  It’s important to understand the respective value propositions here.  “Return on intelligence” is by definition a function of both the cost to produce intelligence and the value received from it.  While the value for customized work is higher, at least to the individual receiving it, the “per user” production cost is typically MUCH higher.  The result is that the ROI for customized work is typically lower than for syndicated work.

Follow me?  The car analogy is useful here.  When buying a car, you’d probably love to have a car built to your exact specifications.  This would provide the highest possible value received, from your point of view.  But the cost would be prohibitively high, and therefore the ROI would be much lower than buying a “syndicated” configuration.  (The first Fords produced this way were all identical.  “Give ‘em any color they want—as long as it’s black,” Henry is said to have offered.)

What modern auto makers have done as a compromise is to give you limited-range customizability, in that you can choose from options—typically 8-10 exterior  colors, 3-5 interior colors, 2 or 3 transmissions, and several options “packages”.  You feel like you have customized your car, and since the mathematical possibilities are numerous, in a way you have.  (DISCUSSION QUESTION:  What’s the lesson here for information providers?)

In any production-based business model, the “holy grail” is to design once what people will buy many times.  This reduces your effective unit cost of production, in that this cost is amortized over many units.

Since I’ve long argued that “information work” is a type of production process, as a corollary I’d hypothesize that these “economies of scale” apply to information as well as to cars.  And since intelligence is a kind of information, that they apply to intelligence as well.

With information, the economies of scale extend beyond product design to include the actual production process itself.  This is because the “marginal cost of production” is trivial.  Once you write your report, it costs no more to copy 20 people than to copy two people.

There is another dimension of value, namely reactive and “pre-active” delivery of information.  That is, is the work done in response to a user request (reactive), or is it done in anticipation that a user is likely to find the information useful?  This latter is what I call “pre-active”, and encourage all my clients to work toward.

I believe these two dimensions of value are highly correlated, as shown in the diagram.  If you are responding to an individual need, you’re by definition being reactive—and even if you do a great job, your value is focused only on the requesting individual.  (One of my clients calls this “stick-fetching”—something they’re trying to avoid.)

If on the other hand you can anticipate the needs of a group of users, you are playing in a much higher realm of value-creation, from the perspective of the enterprise as a whole.  This is the case, for example, with the daily report to 400 top executives mentioned above.

Bear in mind that there’s also a “network effect” in having a larger number of users—they talk to each other, and can become sources of intelligence.  They have a larger combined organizational footprint than a smaller group.

Is the total “enterprise value” higher for the report to the 400, or for the equivalent effort reporting to one executive?  The answer, of course, is “it depends” on several factors, not the least of which is who that executive is.  But in designing the overall “service line” of offerings for the intelligence function, it’s worth bearing this distinction in mind.

You’ll need to do both, but my hypothesis—based on a lot of experience and observation—is that you’ll create greater TOTAL VALUE if you push toward the right on the continuum in my diagram.

What’s been your experience?

---