Posted by: rolfsky | August 28, 2008

leadership does not mean optimizing for ROI

Choosing projects based on projected ROI is a dangerously simplistic way of running your business.

If you take a look at the actual acronym: “ROI” return on investment, it seems like a perfectly logical way of directing your business activity. After all, who wouldn’t want to invest in the things that bring them the greatest returns?

The unfortunate simplification in action is that “return” is generally taken to mean revenue or cashflow, which is but one of the important aspects of running a business.

The problem here is that while revenue can be easily counted, recorded, multiplied and divided; other intangible dimensions cannot be. How do you quantify “trust”? How do you measure “excitement”?

What would an ROI of 20% on trust actually mean? Because the intangibles cannot easily be typed into Excel, they can’t be utilized on pivot charts, or factored into equations.

And because MBA’s live and die by Excel, anything you can’t count, doesn’t count.

ROI based on revenue or other “quantifiable” metrics prove to be an overly blunt way of looking at the world, missing the nuanced and very real ways that vectors like “image” and “brand” profoundly affect your bottom line. If you only have one “real” data point, you tend to optimize to increase that value.

This starts you off making very poor business decisions. Take, for example, 3rd party advertising on your site. Investment is minimal (just open a few content spaces on your pages), and return could be huge. BINGO, the ROI meter goes off, let’s do this!

Did anyone consider what it might look or feel like to have 3rd party advertising on the site? What type of message does this ROI-driven decision send to our customers and what do they think about it?

If you lined up a row of junior MBA’s and really forced them to “prove” it to you, their minds would ultimately jump to questionnaires, surveys, and satisfaction reports. Scratching at the walls to come up with NUMBERS they can give you, they’ll throw out all sort of tests and questions they can ask. Because they want to fit it into their Excel spreadsheet.

And this all takes time and money, and the results can only really be measured after you do something. Not to mention that more than likely, your “results” are indelibly tainted by asking the wrong questions.

so if you can’t measure it, what do you do?

Because so many of the truly important aspects of your business can’t be measured directly, you probably shouldn’t try. Measure the ones you can and then jump in to get a “feel” for the rest. No Excel magic to calculate ROI is going to help you here any more than it would help you learn to surf, or fly a helicopter. You’ve gone through ground school, now it’s your chance to really show what a good pilot you are.



  1. Well said. There are indeed many things in business that you need to do on the fly before you can measure or prove the results.

    You do need to have an appreciation of both tangible and intangible aspects of the business, but don’t take it out on us scientifically-minded MBAs. It’s not excel, it’s clarity. It’s linking A directly, or at least measurably, definably to B.

    It is hard to argue against the fact that you need to somehow measure what you do in order to tell how well you are doing it , even less definable elements of business (especially if there is someone who is investing in your business and might like to see a return). Even if it is not a short or clear path from something like trust to the bottom line, you need to have a story, a reason, a defensible postion. Who are you responsible to and how do you tell the story? Should they just trust you until you? Can you tell a story withou linking A to B in definable ways?

    So, if building customer trust is part of what your company wants to do, then that should also be clearly stated, and you should find ways of measuring whether or not you are doing it in order to tell the story. There are ways of measuring trust, I think, but you are inevitably stuck with the Observation Effect. By observing or measuring, you affect the outcome. The question then is, by how much do you affect the outcome compared to the scale of the observation. Anyway, that’s a whole other story.

    I’ll add one more point: what is too complicated or too many degrees removed to measure now becomes measurable and relationships more concretely definable (say between trust and revenue) in the not so distant future. Could fMRI come up with a trust measurement of people using the site? Could this be related to spending of the customer over the long and short haul? It’s unsettling, even a bit creepy, but these things that were once intangible won’t be forever. Privacy is quickly evaporating because people are (unwittingly or not) volunteering for it.

    How do you see patterns if you aren’t measuring? How do you know if your inclinations about the world are real? How do you define a pattern if the veracity of the pattern is not testable? Should we throw out the scientific method?

  2. @leo You make some very good points. 🙂

    I agree that we are amassing more and more information about consumer preferences and etc. every day, the real question is if we’ll have the intelligence to convert that data to actual information any time soon. Sensing is the beginning, parsing and reacting are still critical.

    When I mentioned that you “probably shouldn’t try” to measure I was specifically referring to the intangible. There is still a lot more art than science at work with managing many of the intangibles and their effects.

    60 years ago, creating an autopilot that could land an airplane was wholly unattainable because we didn’t have the correct measurement devices (sensors) or a full understanding of flight dynamics. Now that we’ve refined both our sensing capabilities and our knowledge of how planes actually fly, auto-landing planes are not so far away in the commercial sphere.

    In the interim between then and now, it would be wiser to invest in smart, trained pilots who could practice the “art” of flying rather than trying to build such a system yourself.

    I always appreciate the discussion. 🙂

  3. I think we are arriving at the same spot: to lead means being ahead of the curve, on the face of the wave. At this point, the tools are not available to do very good prediction, that happens on the tail end of the wave or the market, at the point where you have switched to new market entry to honing processes and performance optimization).

    Wow, re: autopilots, I guess Bill Joy and Ray Kurzweil were right, the future won’t need us very soon.

    Speaking of which, What will the professions of the future be? What will we be able to do that machines cannot? Are we reaching an age when we will be ever more easily swayed? How can we maintain individuality and freedom of choice in the face of concepts such as neuromarketing (

  4. […] generated as a side effect.  Rolf Skyberg has written an excellent post on a similar theme, called “Leadership doesn’t mean optimizing ROI”. Read the article, it’s excellent. Here is an excerpt. Choosing projects based on projected […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: