Craig Roth

A member of the Gartner Blog Network

Craig Roth
Managing Vice President: Communication, Collaboration, and Content
4 years at Gartner
25 years IT industry

Craig Roth is a vice president and service director for Gartner Research, in Burton Group's Collaboration and Content Strategies service. Mr. Roth covers a wide range of knowledge and Web-related topics at the intersection of collaboration, content… Read Full Bio

Coverage Areas:

I Dare you to Prove the Business Value of Anything

by Craig Roth  |  November 22, 2013  |  1 Comment

Over the years of covering portals and collaboration technology, I’ve often been involved in discussions over proving the value of these initiatives. I work with the client asking the question to determine the type of arguments they want or are allowed to make, their specific situation, and help as best I can.

But underneath it all, there’s a meta-argument: how do you prove – I mean really, scientifically prove – the value of anything to the business? Yes, I know ROI, NPV, IRR, and all the other acronyms. And I’ve seen them all fudged and led astray too. To cut to the chase, I think the answer is that you can’t really, scientifically prove the value of an initiative to someone whose gut tells them its not a good idea. Even if you make a rational case, they’ll just keep digging deeper into the numbers since they’re sure that somewhere the argument doesn’t hold. Initiatives that got approved just made more intuitive sense – they wouldn’t have passed the deepest scrutiny either.

One way to prove this point is to think about what the deepest scrutiny would be. What would a rock-solid numerical business case look like?

Think of the “hardness” of a business case being like a 1-10 scale, like mineral hardness where 1 is talc and 10 is a diamond. Then rate the hardness as follows:

  • Add 1 point if the improvement is 100% attributable to the process: Great, so sales went up 20% the quarter after you rolled out the new CRM system and you have anecdotes about how the improved data quality and access actually helped close deals. But can you really pin all or any portion of the improvement directly on the new system?
  • Add 1 point if causation directly links the process to the improvement: Gallup research shows highly engaged workforces have 21% higher productivity, and 22% higher profitability. That doesn’t prove causation though – engaged workers or the types of companies and industries that have more engaged workforces may attract workers that are more productive and tend to be more profitable (hence better bonuses and chances for advancement). Another way to look at this: would making the changes to have Gallup rate you a “highly engaged” workforce result in 22% higher profitability?
  • Add 1 point if other factors could not have caused the result. When you rolled out a wiki for sales reps, sales went up 20%. Great, but did you isolate for cyclical factors, or general uptick in the market, or a competitor pulling out of the market? Economists isolate variables they aren’t testing for such as income and educational level, but I’ve never seen that done for more than one variable in a business case.
  • Add 1 point if a substitute process could not have provided higher improvement: A help desk used to carve trouble tickets on stone tablets and get them to level 2 support in a wheelbarrow. A fancy, custom, $500,000 system now allows them to track and forward them online and has proven $1.5mm in time and materials savings. Great, but maybe you could have saved $1.49mm by just creating a simple online form in Word.
  • Add 1 point if the measurement direly impacts the bottom line (in for profit companies, increases income or decreases expenses): Improving employee happiness, Twitter followers, defect-free production, even net promoter score are great, but they don’t pass this test unless they, in turn, can be shown to improve the bottom line in a way that passes the rest of these tests.
  • Add 1 point if benefits can be shown using accepted accounting principles: standard payback period for your business, discounted to standard internal rate of return, use of common infrastructure accounted for. The upfront expense of your system easily pays for itself over 10 years? Big deal if the standard payback period is 5 years and future benefits are devalued due to inflation.
  • Add 2 points for time and place consistency: the benefits are shown over a sufficient period of time to be unlikely to be a random blip, and over numerous locations (if applicable). This is related to the “other factors” test above. Repeatability is even better: if an improvement made in one division saves 15% in a certain expense category, does it do the same when tried in two more divisions?
  • Finally, add 2 points if the person requesting the rock-solid financial proof of value can prove that their own benefits to the company as an employee exceed their salary by more than any other person in their job with the same salary would. I’m kinda kidding on this one, but it makes the extreme point that not everything valuable can have indisputable proof.

I’m sure you can think of more ways to poke holes in financial arguments. Again, my point isn’t that all improvements are phooey. It’s that no company could run only on technology and processes with proven value. Unprovability isn’t an excuse to approve all propjets, but rather an understanding that the bar necessarily needs to be set at a reasonable – not perfect – level. Good leaders need to have a gut feel for risks and make decisions under uncertainty. Otherwise, scientists (or computers) would be running corporations.

It’s ironic that as availability of sensors, big data, process modeling, and lightning fast number crunching increase, good instincts and common sense have never been in higher demand.

1 Comment »

Category: Business cases Collaboration     Tags:

1 response so far ↓

  • 1 josephmartins   November 22, 2013 at 10:24 pm

    This is true of just about any decision-making, personal or professional.

    For anyone interested in the “why?” check out Damasio’s excellent book “Descartes’ Error”.