An email to a friend: Why the Boston Consulting Group found returns on "innovation" aren't acceptable

A friend recently emailed me a question:

According to a Boston Consulting Group study, 3 out of 4 companies will increase spending on innovation this year but 1 out of 2 are unhappy with the return on investment. One critical reason for that is that no one knows how to measure the success of "innovation."

What are your thoughts on measuring innovation? How would you show a boss that the million dollars he spent on innovation last year was worth it?

We observed there are three kinds of innovation: selling more to existing customers, improving the productivity of existing processes, and creating new markets of customers. The question was whether if we could measure innovation better, we would get better results. After some back and forth, here's the conclusion I came to in the last email in the conversation:

In my personal experience here is what I have seen happen.

A) Certain people are responsible for increasing sales from “new products to existing customers.” This is a combination of sales and R&D people. At times there were increases in the budget for this, and sales out the back end of the process were tracked. We knew the ROI on this.

B) Certain other people are responsible for “reduced costs.” This is a combination of operations, IT and R&D in terms of reducing the cost of producing the product, and primarily operations, IT, and sales in terms of trying to reduce the cost of other processes. At times there were specific budgets for this, and cost reductions were tracked. We knew the ROI on this too.

In KEMET’s specific example, the CEO (and everyone else) would tell you that the return on investment on these investments was not acceptable. We could measure it till the cows came home and the returns would still be unacceptable. The problem we had mature products and a mature business model and we kept squeezing the orange but getting less juice out of it. We could measure the juice coming out, there just wasn’t enough, and more precise measurement of the juice or of the effort to squeeze the orange wasn’t going to help much.

C) That leaves us with, “selling products to new customers.” Most companies don’t know how to do this. Reference Clayton Christensen’s The Innovator's Dilemma. The problem that Christensen identifies is that while A) and B) are high data activities where we can make data decisions, there is very little data about “new customers,” so inherently it is a low data environment. This is not a question of not measuring the right things, because there is not much to measure. When investing I used to remind people that we could do all the due diligence we wanted, and we should, but there was still more that was unknowable than was knowable. So Christensen says that in addition to teaching mangers to make data driven decisions in later stage businesses, we should also teach managers to make theory driven decisions in early stage businesses. This is the informed intuition that the entrepreneurs I spoke to were depending on. They were not placing wild-eye, unknowledgeable bets, it was just that their knowledge had been gathered informally over a long time and they were doing what humans do so well, which is find patterns in widely disbursed data. It's why experienced VCs bet on the jockey, not the horse. A great book on this phenomenon is Blink.

I have an informed guess about what the CEOs you are talking to are telling you. For the past couple of decades, major companies have invested a great deal in quality and cost reduction programs like Six Sigma. It is not that they have not executed this well, that is, if they measured it better they would get better results. Twenty years ago, there was a lot of juice in the quality and cost oranges. Today, in major corporations most of the juice has been squeezed out and so their returns on continuing to squeeze are diminishing. What they need to now focus on is entrepreneurial opportunities “selling products to new customers” but they don’t know how to do that.

“Innovation management” is at the same place “quality management” was 20 years ago. You and I have discussed this before. I don’t think the problem of “innovation management” is A) and B), as much as it is C).

John