On the types of innovation

As usual, this Friday’s The Economist holds some thought-provoking pieces. One, unrelated to our subject, mentioned (at last) the blatant political leaning of a Spanish newspaper they have often blindly relied on for analysis.

The second and more relevant one was a very confused piece about innovation in Great Britain.

I’m not blaming them, they’re journalists (however better prepared than the usual sort), but the picture they manage to convey is as muddled as possible. Which is quite a bit, with good reason: there is no wide agreement on the very basic concepts of the field, so The Economist ends up mixing patents (non-economically meaningful indicators of leading innovation) with investment in software (economically meaningful indicator of the adoption of previous innovations) and even marketing expenses in the same pot… and bringing in «intellectua assets» as reinforcement.

I’m not going to reinvent a field that already claims hundreds of specialised consultants, but it seems to me that «innovation accountants» could do with a better look at the subject of their accounting.

Not just off the cuff…

There was a time, not long ago, when I was asked to comment on the nature of innovation for a regional innovation-management body. It got useated by elections, which was sad, but I kept my notes… and I badly wanted to get it off my chest. So here’s a proposal for those interested in a bit of clarity.

First, let’s look at the beast: we need a clear narrow definition of «innovation», so we can then append adjectives that make sense.

Innovation is literally change. Purposeful change. And it spans two main varieties: relative innovation (when the agent becomes different from the rest of the field) and absolute innovation (when the agent changes, period).

Economically relevant innovation is that which enables the agent to improve its performance as an economic agent. If a company, it will improve results (in any way). If a country, it gets more complicated.

Not just arbitrary

There are four large groups of innovations available to an economic agent, according to the aspect that gets changed. It can be productive, it can be managerial, it can be financial, and it can affect distribution.

• If productive, it affects the competitive qualities of the product or service delivered. This can comprise product design improvements as much as production methodology improvements. It is embodied as much in designs (intellectual property) as in expertise.
• If managerial, it affects the way the agent’s resources are used to make the most return out of them. This comprises every aspect of the use of existing productive assets (including people, knowledge, time and money, as well as the rest) and is enbodied in procedures, methods, policies and information systems… and yes, it does include marketing practices and activities.
• If financial, it affects the ability to acquire and make use of the resources needed to run and build the agent’s operations. This comprises everything from the cost of operating capital to the ability to execute project finance deals. It is embodied in the financial capabilities of the firm, both exercised and potential.
• If distribution, it affects the range of action of the economic agent, or in other words the geographical expanse of its market. This comprises the network of partners and channels used to do business in each and every market. It is embodied in the range of subsidiaries, intermediaries, suppliers, dealerships and contacts, as much as in the outsourcing, support, agency or other contract agreements and the means to enforce them.

No, I’m not quoting. Unless you count «from observation».

Real innovation and potential innovation

Where does this classification leave such staples of innovation accounting as patents and basic scientific research? Well, not quite off the picture, of course.

Basic research is not innovation, per se. Least of all, economically-meaningful innovation. It does have a long-term effect in the ability to innovate, in some ways that have not been as thoroughly described as I’d like. But it is a different beast and I would not advocate putting it in the same basket.

Between basic research and engineering changes lies «applied research» which is often branded «R&D«. There is a strong link between applied research and innovation, especially the productive kind, but it is important to point out that the relationship is not a full dependence. We will return to this later, but at present let it suffice to consider that the main result of R&D is patents.

A patent is just a way to enforce rights on a piece of engineering (at least, outside the US where you can attempt to patent business models). So a registered patent is a sign that there is potential innovation, but not neccesarily that it has been applied or will ever become relevant to the economy; so many patents languish unused and unlicensed that it would be crass to think differently.

As we’ve seen, all three are bad proxies for actual innovation. This has not prevented them being used as such, but should help us unclutter the picture.

Riding the wave or sailing along: leading innovation and adopting it

Another issue adds to the confusion: mixing cutting-edge innovation with follow-up adoption. They are quite different, but possibly in different ways than the accountants have been touting.

Let us put it bluntly: is it more innovative to plug seven different hitherto-unused patents into a blockbuster product (iPhone) or to spread the use of weaving ants for pest control in thousands of African mango groves? Neither technology is absolutely new, no new patents are involved, the market values derived from the innovations are just as great, and the logistical difficulties are enormous everywhere.

Personally, I think they are just as innovative. But they innovate in different ways.

Pest control with weaver ants is indeed change, and innovation in absolute terms. But it is not very innovative in relative terms: it has been done before elsewhere, in only slightly different ways. Just as offshoring manufacturing is, nowadays, hardly leading-edge, and lean manufacturing would not shock many.

If you look at the economic impact of the change, you can’t grudge these «adoption» innovations their title. They’re innovations, and not just because none is just like its predecessor. And they’re key to a company’s survival, or a country’s competitiveness. Putting this kind of innovation out of the picture would be self-defeating for those driving the «innovation accounting» as a competitive tool.

On the other hand, leading innovation (the relative kind) is a different beast. It doesn’t always make economic sense (market failures are not uncommon) but when it does, it often gives competitive advantage that lasts a bit more than mere adoption: after all, adoptable methodology is more or less available to all, whereas new methodology can often be warded from the competition by intellectual-rights protection such as patents, and by good marketing practices.

In other words, any accurate innovation accounting would need to measure different types of innovation, clearly separate but all clearly relevant. Leaders who launch or open new fields or markets are rare but significant. Early adopters are grabbing a competitive advantage that will likely last. Late adopters are skipping the hard part and jumping on the benefits. All of them are doing better than anybody else.

Practical consequences

Beyond the niceties of table-painting and report-writing, there are some political consequences to this. They would require a long report themselves, but two stand out:

• Making it easier or more profitable to be first in a market or technology (by making it easier for knowledge to become product or methodology, and by protecting the rights of those who bet on it) is bound to increase the chances of innovation leaders.
• Economically relevant innovation is not just the one on the headlines. It can be as prosaic as extending your distribution network in yet one more country, using tried and true methods. Making those tools and methods available and affordable to local enterprise is bound to help the local economy more than having a couple of worldwide leaders around.

In a nutshell

You have two large groups of innovations: economically relevant, and not. Some are potentially relevant, which is another beast again.

Among the economically relevant you can count four families: productive, managerial, financial, and distribution changes.

You have two large types of innovations: relative, and absolute (leading and adoption) which can be economically relevant or not.

I’m sure you can draw the matrix in a matter of seconds :-). Or even a more evocative figure.

Last but not least, maybe «absolute» and «relative» are not the best picked adjectives for these ends. Can you think of a better pair?