Projects should have a positive return on investment – wisdom shared so often that our extra-ocular musculature has probably thrown in the towel by now.

Those less schooled in the mysteries of management decision-making might be forgiven for thinking this means projects should return more money to the corporate coffers than they company invests in them.

Those with a bit more financial sophistication add opportunity cost to the calculation. Projects, in this more-robust view, should return not only the initial investment, but also the dividends and interest that would have been earned on that money had it been invested in a financial instrument of some kind.

This threshold is called the hurdle rate. Not the hurl rate, although many discussions about project desirability contribute to this. Project governance mavens insist that proposed projects promise to clear a set rate of return – a hurdle in the run-fast-and-jump-high-enough sense of the word.

It’s a superficially plausible criterion that isn’t so much wrong as it is, as someone once observed, insufficiently right. Why it’s insufficiently right is something any chess player who has progressed beyond the novice level of play would recognize.

Novice chess players are schooled in ROI-based decision-making. Each chess piece is, according to this model, worth a given number of points. Why does it work that way? Don’t worry about it unless you’re just curious.

Anyway, ROI-based chess players will cheerfully trade any piece for an opponent’s piece or pieces that are worth more in total than the piece they’re sacrificing – trades, that is, that have a positive chess-piece-point-count ROI.

It’s a formula that’s as plausible and wrong for chess-playing as ROI-based decision-making is for project governance decisions.

The fault in ROI-driven decision-making logic stems from this characteristic of business (and chess): Strategies don’t have ROIs.

In chess, strategic decisions are based on whether a move will increase the likelihood of beating the opponent. Removing an opponent’s most powerful pieces certainly can contribute to this, but so can other moves.

In business, strategic decisions should, in similar fashion, be rooted in beating opponents – in a word (okay, in two words) – competitive advantage.

This is, by the way, the flaw in stock buy-backs. When a board of directors decides to buy back stock it’s spending money that could have been used to make products more appealing or customer-care more loyalty-building. Instead, the board reduces the number of stock shares profits are allocated to, artificially … and temporarily … inflating the company’s earnings-per-share calculation.

Nothing about this analysis makes a focus on ROI wrong. Sure, a project that delivers untold wealth to the corporate coffers is, more often than not, a good idea.

But not always. A project that, for example, makes a colossal profit by posting a few million more cat videos to YouTube is sufficiently horrific that it should be vetoed by all right-thinking (and, for that matter, left-thinking) individuals, ROI or no ROI.

But I digress. Getting back to the point, strategy doesn’t have an ROI. It might seem to – you’d sure think competitive advantage should generate countable currency – but that’s rarely the case. One reason is something that, in evolutionary theory, is called the Red Queen hypothesis. It proposes that newly evolved adaptive advantages don’t always confer lasting results because a species that evolves an adaptive advantage leads its predators, prey, or competitors to adapt to their adaptation with their own now-advantageous adaptations.

Bob’s last word: I trust the business parallel is clear. But we need to take this one step further: As with so many instances of organizational dysfunction, the insistence on ROI stems from an unhealthy emphasis on measurement.

ROI makes value measurable. Not really, but it looks like it. Competitive advantage, for example, generates a financial return, but the size of the financial return can’t be predicted in advance. It isn’t just that anyone who tries to predict future customer behavior is about as reliable a source as Nostradamus, although they are.

It’s also that predicting how competitors will respond to a company’s strategy is almost as hard, and arguably more important.

Bob’s sales pitch: About once a month I publish a piece on CIO.com under the heading “CIO Survival Guide.” They’re a bit longer than KJR. And as the title implies they have a more overt CIO focus. You can see them all at Bob Lewis | CIO .

New on CIO.com’s CIO Survival Guide: “Why IT communications fail to communicate.” The point? Never confuse documentation with communication. The purpose of documentation is to remind, not to communicate.