You can’t have it both ways: you either want to innovate, or you want a forecastable return on investment.

The only way you can hope to forecast returns — if it’s ever possible — is if all the parameters are known and easily measurable: cutting production cost by 2% for example. 

In these situations you can put a spreadsheet together and can work out the likely payoff with a degree of accuracy, at least in theory.

But when you actually innovate — create a new product, try a bold idea, or approach something in an entirely different way — it’s impossible to know what the payoff will be. It could fail, it could be transformative, it could be somewhere in between. Nobody knows. 

There are two implications. 

First, if you only sign off projects with an obvious return on investment you’ll be limited to a world of incremental improvements whose outcomes are easy to quantify, but whose impact is severely limited. In the long run you’ll be left for dead by competitors who are willing to think bigger.

Second, if you want to innovate, the financial rationale to use is affordable loss, not return on investment — working out an amount you can safely spend to see if an idea works. An approach which, ironically, exposes you to much less financial risk than fantasy ROI spreadsheets that have been fudged to make the numbers work.

See this post on LinkedIn