Published on:

ROI, risk avoided, and value delivered always depend on assumptions, which are probabilistic and problematic

What makes estimations of return on investment intractable is the inherent reliance on unreliable assumptions. All ROI “calculations” depend on givens: “Lawyers spend at least 10 minutes a day looking for documents,” “Paralegals costs us about $25 an hour,” “We prepare more than 30 placement agreements a quarter,” or “Approximately 8 percent of all disbursements charged us exceed our guidelines.” If managers don’t know a number, they estimate or guess it, a risky step.

The assumptions in ROI forumulas rest on presumed facts that are to varying degrees carefully and objectively obtained. In the end the concatenation of assumptions leads to a saving or an improvement in output. Similarly, the input – the investment – harbors some grabbed-from-the-air figures. All ROI estimates, therefore, depend on the assumptions people build into them, which naturally gives opportunity to slant the resulting division of outcome divided by input. And we haven’t even mentioned holding all other factors constant that might influence the results.

Quantifying risks avoided stumbles like calculating ROI: both are exercises based crucially on assumptions, and everything revolves around those key projections or approximations. “Hiring this firm at these rates will avoid a second class action,” “Reducing the number of discrimination complaints by 10% is certain,” “Among these patents are at least two with 7-figure revenue.”

The same holds for “value” from law firms or law departments. “If we get this acquisition done before the year end, EPS will grow at least 15%,” “They can come up with a solution in half the time,” or “The firm’s staffing model is a third or greater more effective.” A comparable slight-of-calculator regarding factual inputs surrounds determinations of the net benefits of legislation or regulations.

You can reach any conclusion you want if you manipulate the assumptions. The quantitative assumptions must be stirred into the mix, but they don’t become empirically true simply for being either necessary or stated as numbers. All assumptions are probabilistic; life if probabilistic. That may be too grand a statement of philosophy for a miniscule blog, but the basic truth is unavoidable. We can create numbers around returns on investment and risk avoided and value delivered, but somewhere inside those numbers nestle one or more assumptions that importantly determine the conclusion.