Measuring innovation returns: A Q&A with Alexander Kandybin
Why do companies have so much trouble realizing a return on their innovation efforts?
Alexander Kandybin: Very few companies, if any, calculate actual returns. Instead, they assess their innovation efforts from a high-level, strategic, conceptual perspective. Sometimes they calculate a net present value (NPV) for certain projects, but NPV can’t be directly connected with either expected growth or the overall revenue companies would like to get out of their innovation portfolios. NPV is a metric that many people have difficulty making sense of, and it is very difficult to connect it to something tangible.
In addition to being a comprehensive measure of innovation performance, the Return on Innovation Investment (ROI2) methodology allows companies to compare innovation returns with returns from other types of investments, such as marketing, or returns from small projects versus large projects. It makes comparison across innovation initiatives much easier, and that allows you to manage innovation returns very explicitly because you’re measuring them. One of the most useful characteristics of ROI2 is that it correlates with organic growth and can be effectively used for managing innovation for growth.
How hard is it for companies to accurately assess potential returns on innovation projects before they’ve actually managed to get those projects off the ground and into the marketplace?
Kandybin: Everywhere we implement innovation portfolio management, using ROI2 as a methodology, that question arises. Potential returns can be assessed based on a very high-level analysis. If I know how much I am planning to spend on a particular R&D project, and can estimate the product’s future margins and third-year sales — information that, in most cases, is available very early on — I can calculate at least the approximate returns. If I can extend that to my entire portfolio, I will have a lot of information about where I should be spending my money.
As long as you perform these calculations consistently across all the projects in your portfolio, the absolute accuracy of each individual project calculation becomes less relevant. What is important is how that calculation compares with another project and another idea. And as long as all of them use the same approximate methodology, that’s all you need, because you are making the assessment on a comparative basis.
How can that information be used to analyze a company’s overall returns on its innovation investments?
Kandybin: That’s where the innovation effectiveness curve comes in. It’s a way of representing a collection of returns for all the individual projects in the portfolio. It allows you to actually see which projects promise high returns and which promise low returns. And that, in turn, can tell you very informatively what the effect of changes in your R&D expenditures will be. (see Exhibit 1.)
If your innovation effectiveness curve is at a reasonably high level, but you’re already at the limit of your R&D budget, then you should probably increase R&D expenditures. Those dollars would be very effectively spent because you can use them to continue to realize higher returns, generating incremental sales and more growth.But if you have a long tail of low-return projects, the chances are that any increase in R&D investments will only generate an even longer tail of low-return projects and will create neither growth nor value for the company. You can’t simply spend more, because the additional dollars you spend will not improve your effectiveness, your ability to grow. What you need to do is to raise the curve first — to generate more high-return projects and fewer low-return ones. That earns you the right to invest more money in your R&D portfolio. But doing that, of course, is the difficult part.
Note: This graphic was originally published by Booz & Company.
How does the innovation effectiveness curve vary by company?
Kandybin: Just about every company’s effectiveness curve has three sections. The first section includes the high-return projects. Some of these projects are large, some small; some are ongoing efforts, some one-off deals. Such projects cannot be replicated easily, and that makes it very difficult to build an innovation strategy based on them. It differs company by company, and industry by industry, but typically companies spend just 3 to 10 percent on projects with very, very high returns. It seldom goes beyond that.
Then you have the projects in the middle section with solid returns — what we call the healthy portion of the effectiveness curve. The projects in this section can be replicated; this is the part of the curve that you want to try to influence. And there is the tail of less effective projects, whose contribution to growth rapidly declines. The length of that tail really varies by company.
Is it possible for a company to have too many high-risk projects?
Kandybin: Yes. When we first build an effectiveness curve for a company, we don’t take risk into account, because we’re just trying to figure out the potential of each project. But when we then try to assess the effectiveness of the portfolio as a whole, we do want to bring risk into the equation. If all the projects are too risky, then absolutely, the portfolio should be rebalanced across high- and low-risk projects. Portfolio balance as a function of risk comes into the equation once you’re at the point where you’re actually managing the portfolio.
Indeed, when we’re at the stage where we are actually managing the portfolio, ROI2 is just one of the metrics we take into account. A comprehensive innovation portfolio assessment typically includes four components: overall strategy, ROI2, incremental revenue, and risk profile.
Can you provide an example of how the ROI2 analysis works in a particular case?
Kandybin: We recently performed an ROI2 analysis for Bayer Material Science. Looking at the company’s effectiveness curve, we noticed a very long tail. Then we analyzed the types of projects in the tail versus the types of projects in the healthy portion of the curve, and we found something interesting. The projects in the tail were predominantly focused on two customer segments, neither of which was really present in the healthy portion, which was focused on different customer segment.
We concluded that Bayer was realizing much higher returns through new-product introductions for very specific customer segments, and that focusing on them would generate higher growth. So we recommended that the company reallocate some of its innovation resources to the more promising customer segments, thus extending the healthy portion of the curve and shortening the tail. As a result, the company’s total return on innovation investments has gone up.
Based on this analysis, we helped the company redeploy its resources from these two customer segments into others, and the company changed its strategic priorities. This also demonstrates how this analysis can inform a company’s overall strategy. Not only should strategy drive your portfolio decisions, but good innovation portfolio analysis can also provide useful feedback for your strategy.
Have you also used this technique to transform a company’s entire innovation strategy?
Kandybin: We have. With Bayer, we essentially executed a tactical redeployment of resources, which improved its returns and raised its growth rates. But Bayer was upgrading its innovation efforts within existing categories and customer segments, a tactic that has limits in terms of how much value it can deliver.
In the work we did with Pfizer Consumer Healthcare, we took a very different approach. PCH came to us looking to transform its entire approach to innovation, to shift its entire effectiveness curve upward toward higher returns. The only way to do that is by linking the portfolio directly with strategy, building the capabilities to improve the company’s strategic priorities, and then using those priorities to determine which innovation projects enter the R&D portfolio. So PCH refocused its strategy on growth and invested heavily in creating an innovation portfolio management capability that really is best in class in terms of selecting the right ideas. The company executed this plan over a period of two to three years, and ultimately lifted its entire effectiveness curve. In 2001, PCH came in last in a consumer healthcare benchmarking study we conducted; in our 2006 study it was number one.
This approach to innovation portfolio management is very much driven by metrics. Is there any room left for the “magic” of innovation?
Kandybin: Absolutely. No amount of analytics will replace the magic in R&D—the ingenuity of scientists and marketers who come up with great ideas. The analytics enable you to differentiate between good and bad projects. You can never be 100 percent sure, but even small improvements in your ability to tell the difference between good and bad projects and to increase innovation effectiveness can have a tremendous impact on your ability to drive growth. So although the ROI2 methodology can help you focus the magic in the right direction, ultimately it can’t improve the magic involved in generating great ideas — that’s still the hardest part of innovation.