Variables Determining the Rate of Adoption of Innovations (008)

By Bill Henderson on May 29, 2017

If you have been readings the foundational posts for Legal Evolution, this installment (Post 008) will reward you with something of clear, practical value: An empirically grounded model that identifies specific factors that influence the rate of adoption of an innovation.

What is the specific practical value?

If you are an innovator, this model can be used as a functional checklist to assess whether your innovation is ready for market; and if so, where to focus your limited bandwidth to maximize the odds of successful adoption.
If you are an early adopter, this model helps you assess whether you want to cast your lot with a specific innovation or, instead, hold your powder until the innovation is more developed or another innovator produces something better.

The graphic above is adapted from Chapter 6 of Everett Rogers, Diffusion of Innovations (5^th ed. 2003). As noted earlier, this is one of the most cited books in all of the social sciences. Although the graphic does not look quantitative, it is actually a user-friendly presentation of a multivariate regression model.

The left column of the graphic lists five groups of variables that influence the rate of adoption of an innovation. The rate of adoption is the dependent variable, which is listed in the right column. The rate of adoption is a dependent variable because its value depends on the value of the other variables. In the parlance of statistics, the other variables are called “independent” or “predictor” variables. The five groups of variables on the left have been shown by Rogers and others researchers to be valid and reliable predictors of the rate of adoption of an innovation.

If you’re investing a lot of time and money in an innovation, this is a profoundly useful model.

I. Perceived Attributes of the Innovation

Among the five categories of predictor variables, the most important is the first category, the “perceived attributes of innovation”. Rogers reports that between “49 and 87 percent” of the variance in the rate of adoption can be explained by five attributes: (1) relative advantage, (2) compatibility, (3) complexity, (4) trialability, and (5) observability (p. 221`).

Note that this is a list of perceived attributes. Perceived by who? The target adopter.

There are many ways to fail as an innovator, but one of the most common is failing to adopt the perspective of the end user. Rogers begins Chapter 6 with a telling quote: “If men perceive situations as real, they are real in their consequences” (quoting W.I. Thomas Florian Znaniecki, The Polish Peasant in Europe and America 81 (1927)). Adopting the perspective of the end user is an exercise in empathy. This can be very difficult for the innovator, who is often deeply immersed in the technical workings of the project. He or she is at grave risk of falling in love with features that are of little practical value to the target end user. Cf. Curse of Knowledge (cognitive bias that afflicts experts).

Rogers distinguishes between “objective rationality” relied upon by the expert who carefully reviews data and “subjective objectivity as perceived by the individual” (p. 232). The latter is what is relevant to adoption. Most of us try to generalize based on what makes sense to us. Instead, we need to spend all of our time studying someone very different and seeing the world through their eyes. Acquiring this skill set take effort, self-awareness and humility. What you think or I think literally does not matter.

Here is a summary of each perceived attribute.

1. Relative Advantage

Relative advantage is “the degree to which an innovation is perceived as being better than the idea it supersedes” (p. 229). The advantage could take the form of economic benefit, an increase in social status, or both.

It is worth reinforcing the user perspective here. I have seen numerous legal start-ups struggle and fail because the founders were pitching efficiency to law firms. Although clients complain about high legal bills, the law firm that makes a large capital investment in efficiency has a very difficult time capturing a reasonable portion of the value created. See Henderson, The Legal Profession’s ‘Last Mile Problem”. When a salesperson makes the efficiency pitch, they are generalizing from their world, not the world of the prospective law firm adopter. Quality, on the other hand, has a much stronger appeal to lawyers, primarily because it is associated with lower risk. We’ll go deeper on this point in a future post. See also the discussion below regarding trialability and Practical Law Company’s successful entry into the US legal market.

2. Compatibility

Compatability is “the degree to which an innovation is perceived as consistent with the existing values, past experiences, and needs of potential adopters” (p. 240), The phrase “disruptive innovation” undoubtedly helped Clayton Christensen sell hundreds of thousands of copies of his famous book, The Innovator’s Dilemma. However, it not a phrase that will endear you to the vast majority of adopters who have zero interest in having their livelihoods disrupted. The touchstone here is familiarity. The closer we hew to what is known and accepted, the lower the levels of perceived uncertainty. That is a zone where your innovation has a chance of getting adopted.

To illustrate this point, Rogers notes that care should be taken in naming an innovation, as the name often carries influential connotations that can undermine relative advantage (pp. 250-51). Note that compatibility is often treated as an empirical question. “Positioning” research looks for optimal associations with accepted products or services in the adopters’ environment. Likewise, “acceptability” research seeks to identify factors that tend to make or break an adoption decision. Compatibility research is quantifying the emotional, subjective reactions of potential users. The only thing close to this in law are focus groups designed to simulate juror reactions. The best trial lawyers use this methodology in preparation for trial. (E.g., Fred Bartlit once told me he used eight separate mock juries for case he was trying. No surprise — he won.)

3. Complexity

Complexity is “the degree to which an innovation is perceived as relatively difficult to understand and use” (p. 257). Whereas relative advantage and compatibility exert a positive influence on adoption, complexity has a negative effect. The higher the perceived complexity, the lower the rate of adoption. Thus, it is not surprising that successful tech companies obsess over user experience (UX) and user interface (UI). Design thinking often adds value by removing unnecessary and cumbersome complexity. See, e.g., Design Thinking Comes of Age, HBR (Sept 2015). The graphic below illustrates this point. The product on the left was designed for the end user; the product on the right stayed too much within the perspective of the engineer.

For the curious, the iOS Human Interface Guidelines are published online here.

4. Trialability

Trialability is “the degree to which an innovation may be experimented with on a limited basis.” Rogers continues, “New ideas that can be tried on the installment plan are generally adopted more rapidly than innovations that are not divisible” (p. 258).

Several years ago, the original US sales team of Practical Law Company (PLC) shared with me how they successfully established their US operations. PLC sells annotated forms and practice guides for sophisticated corporate work. Although PLC had a complete lock on the UK market, they had no US customers when they landed in New York in 2007. Through trial and error, they soon discovered that the single best way to overcome the skepticism of US lawyers was to put them in front of a computer and let them use the PLC product. After experiencing the product’s immense utility, subscriptions were relatively easy to close. By the time PLC sold to Thomas Reuters in 2013 (for a price between $300-450 million), PLC had 700 legal departments and 86 percent of the AmLaw 200 as customers. See Thomson Reuters to Acquire Practical Law Company.

Trialability was certainly relevant to PLC’s rate of adoption. However, the PLC product line also had a huge relative advantage over the incomplete, out-of-date, and unannotated internal forms they were replacing. Trialability enabled perspective adopters to experience the quality difference. To enable high quality decision-making, it is important to keep analytically distinct each of the five perceived attributes of an innovation. Trialability is different than overall relative advantage, though both levers are important.

5. Observability

Observability is “the degree to which the results of an innovation are visible to others.” Observability is very much related to relative advantage and trialabilty. If an innovation is trialable for early adopters, its relative advantage can be more easily observed by other parts of the social system. See foundational posts 004 and 007.

The importance of observability is documented in an early and influential diffusion study that focused on adoption of hybrid seed corn in two communities in Iowa. See Ryan and Gross (1943). What drove adoption for the vast majority of farmers was not the technical sales pitch made by college-educated agronomists. Rather, it was the observably better corn growing on their neighbor’s land. The technical pitch was primarily relevant to the innovators and early adopters in the social system, who set the adoption cycle in motion. The average time between “knowledge awareness” and the “adoption decision” (technical terms of art in diffusion research) was a fairly lengthy six years. See chart below.

I believe the above chart is very relevant to all the hype regarding how artificial intelligence is going to revolutionize the legal field. AI does not have a relative advantage that is easy to observe. Mere efficiency (an obvious and potentially observable advantage) is not good enough for many lawyer-adopters, as efficiency currently creates collateral business problems that most clients fail to acknowledge. See Henderson, The Legal Profession’s ‘Last Mile Problem”. AI is also very complex. These perceived attributes are going to impede AI’s rate of adoption in law. Many smart people in legal start-ups are trying to use design principles to solve or mitigate these issues. Yet, the best of them know they are climbing a very steep mountain.

Summary of perceived attributes

As noted earlier, the five factors above explain 50% or move of the variance in adoption rates. Stated another way, if you have an innovation you would like others to adopt, focus your attention on these five factors. This simple, empirically derived piece of guidance is one of the reasons that applied research can be so powerful.

Four other categories of variables influence the rate of innovation adoption (II to V in the graphic above). Most of them cannot be significantly influenced by the efforts of innovators, though they are highly relevant because they enable an innovator or early adopter to handicap the odds of market acceptance. In other words, they bear on practical questions like, “should I put more money in?”; “should I sell now?”; “should I fold the business?”; “how long is adoption likely to take compared to other business contexts?” Thus, let’s finish the model with a eye toward how it applies to the legal industry.

II. Type of Innovation Decision

At some point after a potential adopter becomes aware of an innovation and weighs its relative advantages, a decision will be made to accept or reject. There are three types of innovation decisions.

Optional. Basically everyone in the social system is free to decide for themselves. This is market-based. E.g., smartphones, healthier foods, Facebook.
Collective. Through agreement or strong cultural norms, adoption requires a consensus of the entire group. This mechanism has the most negative impact on rate of adoption. It is also the mechanism that best describes the typical law firm partnership.
Authority. One decision-maker makes the decision for the entire social system. E.g., corporate executive; government official. Although authority innovation-decisions are generally the fastest, they run the risk of being “circumvented by members of a system during their implementation” (p. 29).

The type of innovation decision is very relevant to the legal industry. Back in 2015, I organized a panel of legal innovators for the ABA Center on Professional Responsibility. One of the panelists was an venture capitalist who was an investor in Modria, an online dispute resolution service that uses an automated dispute resolution methodology similar to those used by eBay and PayPal. As a former associate at a prominent Silicon Valley law firm, the VC helped pioneer some of the early investment in legal tech, albeit not all investments worked out well. In front of an audience of 300 law firm lawyers, the VC stated that he would never again invest in a technology that was designed to be sold to law firms because “law firms don’t made decisions like rational businesses.”

Placed into the Rogers decision framework, the VC was frustrated by the collective decision-making process of law firm partnerships. From far away, it looks irrational. Up close, however, it’s justified as culture.

That said, it is very easy to confuse the long sales cycle in law with the more fundamental issue of relative advantage. For example, many partners hear their clients clamoring for greater efficiency, and hence are willing to listen to sales pitches. Yet, the partners don’t know to how to honor the clients’ wish because it requires to them to simultaneously (a) pay for, and learn how to use, expensive, complex innovations, and (b) endure a loss in revenues because the clients insist on using hourly production to measure value. Insistence on hourly billing, or shadow billing of AFAs, is a great example of a compatibility restraint that impedes innovation. The legal profession has a very serious last mile problem.

I am confident that the rise of the legal operations role within legal departments is substantially due to the authority innovation-decision advantages of having a single general counsel who possesses traditional executive perogatives. That authority is increasingly being delegated to legal ops professionals who have a clear directive for better, faster, less expensive. See Post 005 (discussing CLOC and the rise of the Type 6 client).

Yet, in the best of circumstances, change management in legal departments is no cakewalk. My friend Jeff Carr, formerly GC of FMC and now at Univar, acknowledged the challenge of MPR, or “massive passive resistance”, in implementing necessary change. Having achieved remarkable financial results through his ACES model, Jeff became a fierce proponent of general counsel as leader, a discipline and topic completely foreign to most lawyers.

If you ask Jeff about the key to successful implementation of change — e.g., requiring every in-house lawyer in his department to regularly score outside counsel using a standard grading rubric — he is likely to point to his face: “See this look. This is the look of me not caring. These metrics are necessary for the functioning of the company. Please do your job.” Another prominent general counsel who successfully transitioned a large legal department away for the billable hour, and has served as an influential advisor to many general counsel, acknowledged to me that such a transition could easily entail the resignation or dismissal of roughly 30% of the department — that was the volume of turnover in his department and other successful legal department transitions he has observed. Change is hard, even for highly educated professionals.

Suffice it to say, whether its collective innovation-decisions, or the reluctance of lawyer-leaders to stay the course because we have little training or experience as managers or leaders, the legal industry presents special challenges for innovation adoption and diffusion.

III. Communication Channels

The rate of innovation is positively influenced by the number and quality of communication channels. This is true in two ways. First, early adopters may become aware of an innovation through a new communication channel (e.g., the trade press or an industry conference). Second, more and better communication channels make innovations more observable to the rest of the social system. Not only does this facilitate economically driven adoption decisions based on relative advantage, it also works to set and reinforce group norms. Thus, a subset of adoption decisions will be socially driven by a desire to fit in or avoid feeling left behind or out of date. Again, diffusion of innovations is a social process; incentives are present, but they are often more social than economic.

Not surprisingly, the advent of new communication channels like print journalism, radio, television, and the Internet have all increased the pace of innovation adoption. The rise of mass media is one of the most important areas of study in diffusion research. Following the publication of the first edition of Diffusion of Innovations in 1962, Everett Rogers, who was a sociologist by training, joined faculty of Department of Communications at Michigan State University. At the time, MSU was the leading institution in this fledgling academic discipline.

Communication channels are important to innovation because they increase the flow of information. Yet, factors that influence total flow are different than the factors that influence the persuasiveness of the information content. For the latter, relative advantage, compatibility, complexity, trailabilty, and observability remain the touchstones.

Legal Evolution is designed to be a new communication channel that will help accelerate the pace of legal industry innovation. As noted in Post 001, this publication is an experiment in applied research. To be successful, I need the readership of legal innovators and early adopters — the light blue portion of the curve. I hope this elite readership enjoys Legal Evolution’s clean layout and the absence of banner ads. If you have the stamina to read a 3,500 word foundational post, these niceties are the least I can do.

By the way, what is the likelihood I could adequately reach my target readership if I published this analysis in a traditional law review?

IV. Nature of the Social System

In Rogers’ model, the nature of the social system is the fourth category of variables that can impact the rate of adoption of an innovation.

For the legal industry, the nature of the social system generally impedes innovation adoption. The most established, influential, and prestigious portions of the legal profession — large law firms, the federal judiciary, legal academia, and the ABA — tend to be traditional bound and skeptical of change that does not initiate with them.

Part of this conservative ethos may be the product of Rule 5.4, which has been adopted in some form by every state. Rule 5.4 prohibits lawyers from co-venturing with other professionals in any business that involves the practice of law. If lawyers can’t be business partners with accountants, engineers, software developers, process experts, and data scientists, etc., that’s going to cut down on the opportunities to learn from them. This makes our social system much more isolated from other innovative parts of modern information economy.

Enough said about that.

V. Efforts of Change Agents

Chapter 9 of Diffusion of Innovations is focused on the change agent. It begins with the following quote:

One of the greatest pains to human nature is the pain of a new idea. It … makes you think that after all, your favorite notions may be wrong, your firmest beliefs ill-founded. … Naturally, therefore, common men hate a new idea, and are disposed more or less to ill-treat the original man who brings it (p. 365, quoting Walter Bagehot, Physics and Politics 169 (1873)).

This is harsh but also has a ring of truth to it. To avoid a hostile reception, effective change agents seek out individuals more disposed toward their message, a group disproportionately comprised of innovators and early adopters. After the change agent assists this group in obtaining a large advantage that others can observe, the change agent will become more accepted within the broader social system. But probably not until then.

Change agents can be university field specialists trying to disseminate agricultural best practices for the good of the state economy. They might also be public health professionals seeking to curb a longstanding but harmful cultural practice that is increasing the spread of disease. The biggest challenge facing change agents tends to be “hetereophily” — i.e., they are often conspicuously different than members of the social system in terms of background and technical expertise. Hence, they struggle to communicate effectively with prospective adopters. Successful change agents find ways to overcome this hurdle. Rogers writes, “As a bridge between two different systems, the change agent is a marginal figure with one foot in each of two worlds” (p. 368).

In the legal industry, change agents are most likely to take the form of technical sales people who are trying to get onto your calendar to sell you a new technology or service. At industry events, these folks are typically called “vendors.” The connotation associated with vendors is often negative. In my opinion, this is a parochial way of viewing the world that cannot be squared with our poor record on client service, innovation, and access to justice. In light of these issues, perhaps we should be more gracious and openminded to those offering tools for improvement.

That said, change agents also exist in established law firms and legal departments — they are quixotic lawyers and other professionals convinced there has to be a better way. In turn, they forge ahead without an empirically grounded theory to guide their actions. As a result, their courage and good intentions are too often wasted.

As editor of Legal Evolution, I’ll acknowledge my own desire to serve as a change agent. After 15 years of study, it is clear to me that traditional methods of legal problem-solving are underserving clients and broader society. See Post 001 (explaining problem of stagnant legal productivity); Post 006 (connecting the breakdown in judicial system with declining legal job market and declining legal enrollment). This systemic breakdown can only be shored up through innovations that improve legal productivity — i.e., combining lawyerly judgment with better people systems, process, data, and technology. Higher productivity will enable more legal output to be afforded by more people and businesses. I realize this entails a value judgment on my part — I generally favor the innovators. But it is also a judgment informed by a lot of data and field research. I am also motivated by the longterm welfare of my students at Indiana Law. I need to be part of a system that works for them and their clients.

My change agent role at Legal Evolution has a very simple formula. After explaining the basics of diffusion theory — through these foundational posts — I’ll present finely drawn examples of innovations that appear to be working in the field. In each case, I’ll provide as much context as possible, as the goal is to enable the success of legal innovators and early adopters.

Post 008 will be the longest foundational post by a wide margin. But this is the heart of diffusion theory and how it can be used as a tool of applied research.