Use Design Choices to Prevent Imitation

Why would an inventor like Charles Babbage insert deliberate errors into the blueprints of the world’s first computer? And why did Apple mislabel early iPhone prototypes as iPods? Actions like these may not seem intuitive but are in fact central elements in an innovation strategy that has long flown under the radar.

Babbage, like many inventors since, was concerned about a rival gaining access to his blueprints and foreclosing his first-mover advantage. By adding errors to his blueprints, Babbage ensured that any competitor that obtained them would struggle to imitate his design.1 Similarly, Apple’s deliberate mislabeling of iPhone prototypes as iPods reduced the likelihood of the groundbreaking innovation being leaked before launch, which could have hurt the company’s ability to profit from it.

Managers often mistakenly assume that the knowledge underlying an innovation is something that is inherently imitable or inimitable. However, the above cases are examples of innovators proactively using design to manipulate the imitability of knowledge. This approach dramatically widens the scope for managerial choice in crafting an innovation strategy. More choice is critical in a world where the losses from intellectual property theft have reached dizzying heights — including an estimated $600 billion annually in the U.S. economy alone — and legal mechanisms for profiting from innovations are weakening.2

Our research shows that managers can make choices in the design of blueprints, prototypes, and products or services to frustrate the efforts of would-be imitators and improve their organization’s ability to profit from innovation.3 Such design choices can inhibit a rival’s ability to discern the knowledge underpinning an innovation, reduce their motivation to pursue efforts to imitate this knowledge, and increase the capabilities required for successful imitation. There are many design mechanisms that could frustrate a competitor’s attempts at imitation and improve an inventor’s ability to capture value from an innovation, but they can all be thought of as involving one simple decision: adding or removing knowledge components associated with an innovation.4

Four Design Plays for Profiting From Innovation

Joseph Schumpeter theorized that innovations emerge from combinations of different knowledge components — pieces of knowledge that often, but not always, manifest as physical or digital parts of an innovation during the research and development process. For example, Apple’s iPhone was famously announced as a device that combined three products into one: a widescreen iPod with touch controls, a mobile phone, and an internet communications device. While each of these three product categories already had strong, established players, the effective integration of the knowledge components underlying all three into an innovative offering blindsided the competition. While the role of design in creating this new value proposition has been widely recognized, the role of design in protecting it (the prototype mislabeling mentioned earlier) has been an important yet less-well-understood element in Apple’s strategy for profiting from this innovation.

Seeing innovations as combinations of knowledge components that can be intentionally added or removed leads to the following four distinct design plays for protecting and profiting from innovation.

1. Removing component(s). This approach splits an innovation into various knowledge components that the company’s partners can develop or produce separately from one another. In global supply chains, it is a common technique for preventing knowledge from leaking from supply chain partners to rivals.5 While supply chain partners may learn from the different components that they develop, the innovator keeps integration in-house to reduce the likelihood that the innovation as a whole will be imitated. For example, based on their novelty and the perceived threat of misappropriation, integrated circuits can be split into components to be produced by different manufacturers and later assembled and tested in-house or by a trusted production partner.6 This design choice promises the best of both worlds for companies that have patents in this area, such as AMD (Advanced Micro Devices); they can manufacture their most sensitive component(s) at trusted foundries while at the same time benefiting from manufacturing the remaining components at foundries that are untrusted but offer a better deal. We call these approaches careful coordination.

2. Adding component(s). Tagging and cloaking are two approaches to adding knowledge components to an innovation to make imitating it more challenging. Tagging involves incorporating specific components into a product that can serve as proof that an imitator has misappropriated the innovator’s idea. For example, cartographers long ago developed an ingenious way of protecting their ability to profit from costly surveying efforts, by including fictitious entries (such as incorrectly named streets, towns, or landmarks) on their maps.7 Because such “trap streets” do not reflect reality, their appearance in an unscrupulous rival’s map provides credible evidence of plagiarism that the cartographer can use to seek compensation. This very design mechanism helped the U.K.’s Ordnance Survey win a 20 million pound payout in 2001 from a rival that illicitly copied its maps.

Another example comes from Genius Media Group, a digital media company specializing in the crowdsourcing and publication of annotated music lyrics. Genius believed that the lyrics Google was serving up in an information box on its search results page were being taken from Genius’s website without permission, thus depriving the company of traffic to its website and ad revenue. To prove it, Genius used a form of digital watermarking: It added patterns of alternating spaces and types of apostrophes to a sample of the lyrics on its website that spelled out redhanded in Morse code. Seeing this digital watermark appear in Google’s information box provided the company with enough evidence to file suit against Google.8

Cloaking “hides” important knowledge components behind extraneous ones that have been added. A well-known example of cloaking is the use of camouflage by automobile manufacturers when road-testing new models. Another cloaking approach is electronics manufacturers’ use of epoxy resin to conceal parts of circuits.9 For example, the Klon Centaur guitar effect pedal was developed over four years by a small team in Boston before being released in 1994. Rather than patenting the novel design, the designer covered the pedal’s electronic circuit in black epoxy resin during production to prevent it from being reverse engineered. The pedal proved to be immensely popular and fetched high prices on the secondhand market as demand outstripped supply. It took many years and failed attempts before an electronics specialist was able to successfully reverse engineer the circuit in 2007 by removing the epoxy resin and identifying the components. This quickly led to a proliferation of pedals with their own versions of the circuit entering the market. Still, the cloaking approach had allowed the inventor to be the sole provider of the pedal for close to 13 years, despite the product’s growing desirability and the absence of any legally enforceable intellectual property rights.

3. Adding and removing components. Spiking and feinting combine the two approaches above. In a spiking approach, key components are replaced with others in a way that causes imitations to malfunction, making effective knowledge theft difficult. A recent example is Nightshade, a tool designed to help artists protect their work from being used to train visual generative AI models without their permission. By imperceptibly changing pixels within their images, Nightshade can “poison” AI models that attempt to use these images in their training data, such that these models are not able to correctly respond to user prompts. While the effectiveness of Nightshade in practice remains to be seen, it appears to have caught the attention of artists, with over 250,000 downloads within a few weeks of its release. Spiking can also be seen in Babbage’s blueprints and in most digital rights management software.

Feinting involves attempting to misdirect potential imitators by intentionally leaking knowledge components that could plausibly be part of the innovation but are in fact intentionally suboptimal. An example of this approach is the filing of decoy patents designed to direct imitators down unproductive paths.10 For instance, in the early 20th century, German dye manufacturers filed such patents to prevent rivals from working out their dye compositions.11 Rivals like DuPont struggled to imitate new dye products introduced by German companies: Even when they managed to replicate the color, the quality of the resulting dye was inferior. A key reason for this was the use of Umgehungs Patente, or “evasion patents,” by German producers that were strategically designed to send competitors down dead ends. Thus, even after World War I provided DuPont with the legal right to imitate dyes patented by its German rivals, “making sense of them was well-nigh impossible.”12 The head of DuPont’s technical laboratory stated that “it takes nearly as much effort to decipher the correlation between patent and commercial dye as it does to discover the color originally, and this has been the chief cause of the delay in making the newest dyes.”13 Further examples include Apple’s mislabeling of iPhone prototypes as iPods, and companies being hired by media producers to “poison” the illegal peer-to-peer sharing of unauthorized copies of their products by mixing file fragments containing corrupted data into the pool of peer-provided files.

4. Not adding or removing components. What we call ground staking involves choosing to keep your innovation complete and unaltered as part of a strategy for profiting from it. This is often the approach used when more traditional protection mechanisms are being relied on, such as filing a patent that is a complete and accurate representation of the innovation as a whole in a bid to acquire temporary legal monopoly rights to commercially profit from it. Alternatively, organizations that own the specialized downstream assets required to realize the value of innovations in their industry may find it advantageous to publish some of their discoveries in full to preclude other companies from being able to patent them. For instance, pharmaceutical company Merck’s decision to put a database of expressed human gene sequences (the Merck Gene Index) into the public domain made these genes unpatentable. This ensured that Merck could continue to develop and sell drugs targeting cardiovascular diseases without the threat of another company claiming ownership of genes associated with such diseases.14 More recently, companies developing open-source software have made millions of dollars and have reached valuations in the billions by charging for additional services that support the deployment of the software in commercial enterprises.

So when should you use each play? The most sophisticated companies combine multiple mechanisms to improve their ability to profit from innovation.15 Surveys have long shown that there are substantial and persistent differences across sectors when it comes to how useful legal mechanisms are for protecting knowledge.16 For example, industries like pharmaceuticals and medical devices have long reported that patent protection is an effective means of protecting knowledge. In these industries, companies will most likely find that ground staking and tagging easily complement the legal mechanisms on which they primarily rely.

However, these legal mechanisms are not viewed as particularly effective across other industries. Rather, surveys have found that some combination of secrecy and lead time is the most significant factor in profiting from innovation.17 This factor often presents a challenge for managers looking for solutions: How does one generate more lead time or ensure secrecy? In these settings, effective means of profiting from innovation may include the removal of components for a careful coordination approach, or both the addition and removal of components for a feinting or spiking approach. For example, to help keep its innovative tire-curing process secret, Michelin spiked it by using thermostats that were calibrated to a proprietary temperature scale whose connection to commonly used scales was known to only a few Michelin executives.18

Competitive Advantage Through Innovation Strategy

The knowledge underpinning an innovation has never been more vulnerable to multimodal codification and instant dissemination. Legal mechanisms for isolating innovations from imitation have struggled to effectively respond to this onslaught. Integrating design into your strategy for protecting and profiting from innovation provides an important new avenue for sustaining advantage in today’s highly competitive markets.

Still, no isolating design mechanism is impervious. Imitators will eventually be able to reverse engineer, or innovate around, your means of isolation. But for many innovators, being able to hold off imitators for that extra day, week, or month is the difference between boom and bust.

Generating and sustaining this advantage requires the use of all of the isolating mechanisms at your disposal. To date, without design, many companies have been having this fight with one hand tied behind their backs and, too often, without harnessing the engineering and strategic expertise at their disposal. Innovating across the functional silos in your organization can not only generate new ways to create value but also new ways to protect it. We hope that our research helps make it easier for innovators to have a fair fight and, in doing so, help sustain and further incentivize investments in innovation.