Wednesday, November 30, 2016

Bake Your Own Pie

Recently I was lecturing a group of high-level Chinese executives, when one asked me: “What do you think of plagiaristic innovation?” Before I could answer, he went on to explain that for China to “catch up,” he felt it needs to have innovation of any kind – even what he called "plagiaristic" innovation.

Don’t worry. I’m not about to rehearse the well-worn arguments about the protection of intellectual property: incentives for continued innovation, just rewards for investors who back authentic creativity, quality guarantees for consumers of branded products, and the like. Nor am I going down the “information must be free” path – indignantly advocating “free as in free speech (not free beer),” “stick it to the man” (the artist is not getting the payments anyway), or the “hackers’ code of ethics.” No, here I’m talking about something else.

My point here is about what “innovation” means. Debates about intellectual property, stealing, and plagiarism are all about who owns the pie. That question is very important, and is obscured when patent “trolls” flood the system with complaints, or when plagiarists masquerade as innovators. But another important point often gets lost in the fray:

Innovation is not about fighting over the pie; it is about baking a new pie.

For example, hybrid vehicles hit the worldwide market starting in 1999 and 2000, and within a few years an echo of patent litigation followed – escalating in 2003. The big car makers battled over who invented what, sometimes with each other and sometimes with small firms, everyone claiming a piece of the pie. Meanwhile, also in 2003 but with far less fanfare, Elon Musk and his team of co-founders created Tesla, the forward-looking innovator that has changed the game in the automobile industry. The noisy pie fights in 2003 were over hybrids; the profound innovations of 2003 were quietly happening at Tesla.

Pie fights extend to all walks of business life, not just battles over intellectual property. For instance, the so-called “browser wars” between Netscape and Microsoft were at their peak in 1998, following Microsoft’s integration of its Internet Explorer browser into its ubiquitous operating system. Advocates of competition howled, and defenders of Microsoft replied with talk of “seamless technology” and “complementarities”. Also in 1998, but unknown to most people at the time, PhD students Larry Page and Sergey Brin created Google – the company that would change the game so thoroughly that we would soon forget about those early browser wars. The noisy pie fights of 1998 were the browser wars; the great innovation of 1998 was quietly taking shape at the newborn Google.

"Wait," you are saying. "After innovating, innovators need to defend their creation." Of course. Take QualComm, for example. That company has an unparalleled track record of continued innovation in wireless technology. As a result, its intellectual property has turned out to be extremely valuable. It has of course defended that property against plagiarists; it owes that to its shareholders. But QualComm transformed its industry by innovating - never mistaking defending IP for creating new, valuable technologies.

All around us, we see real innovators at the cutting edge of knowledge. Have a conversation with my son Burton Barnett, pictured here doing science, and you won't hear about pie fights; you'll hear about amazing new developments in immunology. And similar developments are happening worldwide - in China, Europe, India, the Americas - everywhere forward-looking people are creating new knowledge. This process of innovation is key to our collective future, and it has little to do with plagiarism or pie fighting.

The lesson to innovators: Pie fights are important; we all deserve our piece of the pie. And of course even true innovators often must fight off plagiarists. But being good at pie fighting does not make you good at innovating. Innovation means baking a new pie. 

The lesson to plagiarists: Want to create something useful? Leave the other guy’s pie alone and learn to bake.


Research on the uniqueness of innovators appears in the work of Lee Fleming and Olav Sorenson, among others.

Tuesday, November 15, 2016

The Time-to-Market Strategy

In many industries, products are profitable only during a limited window of time. We see time-sensitive products, of course, in consumer electronics, where new models of phones, computers, and home entertainment products come (and go) frequently. We also see this pattern in fashion markets. Zara, for instance, introduces new clothing products twice weekly across nearly 2,000 stores – amounting to the introduction (and retirement) of over 10,000 designs a year. Timing is key, as well, in the introduction and short shelf life of cultural products, such as popular music and films. The market for video games follows a similar pattern of rapid new-product introduction and short product lives, including flash-in-the-pan hits like Pokemon Go as well as the more cadenced, but still time-sensitive, annual replacements of the various EA Sports games. Innovative pharmaceutical products also compete in a time-sensitive way, since the first to patent enjoys a profit premium that ends abruptly with patent expiration. Consequently, drug companies are in a race against time, first to file the patent, and then to bring the drug to market. Even some durable products, such as automobiles, are introduced and retired on a time-sensitive, if cadenced, basis.

Time-sensitive markets can be identified by the presence of two underlying features. First, these market place a premium on introducing a product early – or at least on a reliable schedule where being late is punished. For instance, releasing a new version of a gaming console like Xbox or Playstation late - after the holiday season – would be unthinkably costly for Microsoft or Sony. Second, the products in these markets also have a limited shelf life. A new clothing style or a new hit song may be wildly popular today, but within weeks (or maybe days) they will be yesterday’s news.

In such time-sensitive markets, success depends on a distinct logic: the time-to-market strategy. To understand this strategy, consider a firm that introduces a successful new product, “Product 1,” as in the plot below. Initially the product takes off slowly, and then it catches on, finally reaching a maximum market size. This “S-shaped” diffusion curve is typical of successful products.
The story does not end there, however, because this company is facing a market where the shelf life of a product is limited in time. Consequently, it is preparing to introduce a second product, “Product 2,” that will compete directly with its own Product 1. Note that if the firm does not introduce Product 2, somebody else will. That is what is compelling the company to continue with the introduction of another product even though it will cannibalize its first product, as shown here.
Now focusing in on Product 1, we can see that there is a limited window of time during which the firm can make money selling the product. It is interesting to think about the price that can be charged for Product 1 over the life of the product. Typically, the price that can be charged will be much higher earlier in the product’s life, for two reasons. First, the earliest buyers of the newly introduced product will be those with a greater willingness to pay for the product. For instance, if Intel comes to market with a new chip that is extremely valuable to cloud service providers running state-of-the-art server farms, these eager buyers will be the among the first to buy this new chip and they have a high willingness to pay. Less enthusiastic buyers will also buy at some point, but only if the price falls. Second, the price will begin to fall over time as other competitors come out with a product that is a direct rival to Product 1. For instance, perhaps AMD will introduce a competitor to Intel’s new chip. As the price falls, now more and more chips are sold as other buyers come into the market with lower levels of willingness to pay. Ultimately, once Product 2 is on the market, the price for Product 1 falls away completely. This pricing dynamic is pictured below.
At this point, the logic of a time-to-market strategy is clear. If we introduce Product 1 as shown, our firm will make a great deal of revenue. To see this, look at both the price and sales volume curves in the plot above. Revenue from Product 1 is found by multiplying these two curves, and total revenue over the life of the product is just the cumulative revenue over time. However, what if we are another firm and we release a competitor to Product 1 – but we do so late. A competitor entering near the end of Product 1’s life may sell at high volumes, but only for a short time and only during the period when the price of the product is very low. This firm will have introduced a product that, over its life, will make very little cumulative revenue. So the earlier a firm enters into this competition, the more it makes in cumulative revenue. In fact, entering earlier increases total cumulative revenue at an increasing rate.

By this logic, it is obvious that we should release Product 1 at the soonest possible date. However, this may not be profitable. To see this, consider now the costs of developing Product 1. If we could take all the time we want, we could carefully research and develop Product 1 and, when it is ready, we will have run up total costs equal to C1. But by taking our time, we might end up introducing the product late, and this will hurt our revenue. So instead of paying C1 over a long period of development, say, 2 years, what if we accelerate development and pay the same amount but all in one year – or even in six months?

Well, here is the bad news. It turns out that compressing development costs into a short period does not give you the same result. This problem, famously dubbed the “mythical man month” by Frederick Brooks, occurs for two reasons. First, compressing the amount spent on development causes many to work simultaneously and in parallel, which results in coordination diseconomies. Second, there is the problem sometimes called “gestation,” where development is inherently sequential and cannot be made to happen all at once. Concretely, gestation requires time in development because answers obtained at one point in time decide the questions asked at the next point in time. Doing all this development at once leaves us asking many questions that we will never need to know the answer to – and failing to answer questions that we wish we had researched.

Consequently, to speed development, it is not enough to concentrate C1 into a more compressed period of time. Instead, we will have to pay more than C1, and the more we compress the development process, this additional amount will continue to escalate until, at some point, we could spend an infinite amount and not improve our release date any more. So development costs increase at an increasing rate as time-to-market shortens, as shown below.
As the figure shows, the time-to-market strategy confronts the firm with a dynamic cost-benefit comparison. To be profitable, the firm needs to introduce the product early enough to benefit from higher revenue over a longer time, but not so early that its development costs skyrocket.

How is a firm to achieve this balance? The answer comes down to organization. Firms that are good at the time to market strategy are organized in a way that minimizes development costs while maximizing the reliability of product introduction. The key factor to be managed by these firms is uncertainty. These firms typically design around customer-defined time milestones, track progress toward meeting those release dates, and hold employees accountable for staying on time. As uncertainties arise, routines such as phase-review processes are used to update release dates or revisit go/no-go decisions. And wherever, possible, the firm contracts with other firms in order to solve “weak link” problems that are slowing its ability to deliver on time.



Time-based competition is discussed in my book on the Red Queen.