In 2012, I started a tiny Meetup group in Washington, D.C., to compare notes with other applied behavioral scientists in the corporate world. There were lots of people who were interested in the field but almost no one who had done any field experiments. The next year, as I wrote an early book on the topic, I again looked for examples of behavioral science in companies. My peers and I found only a handful, scattered around the country.
That picture has completely changed: a wide array of companies and nonprofits are now applying behavioral science to their work. In partnership with the Behavioral Science Policy Association (BSPA) and Action Design Network (ADN), I recently conducted an international survey of behavioral science teams. We found that three-quarters of current corporate and nonprofit teams have started up since September 2014. During that time, their geographic diversity has grown rapidly as well, with new teams starting in countries like the Philippines, Peru, Vietnam, Jordan, and Switzerland.
That survey of behavioral science teams provides a unique look at the extent of our field. Here, I’d like to share some of the key results with you.
About the survey
Last year, I started updating that early book, Designing for Behavior Change. I began by interviewing companies in the field who were applying behavioral science. I quickly realized that the field had grown so rapidly that individual interviews wouldn’t suffice. In partnership with the BSPA and ADN, we developed and fielded a detailed survey of how applied teams use behavioral science in their work, with questions covering the teams’ position and purpose within their organization, the particular techniques they use, and the struggles they’ve faced. The survey focused on the people doing the work—the behavioral teams themselves, not the client companies or governments that they worked with.
Three-quarters of current corporate and nonprofit teams have started up since September 2014. During that time, their geographic diversity has grown rapidly as well.
To date, respondents from 324 distinct organizations have completed they survey. To benchmark that figure, we compiled a list of all known organizations with behavioral science teams—from existing databases, internet sleuthing, and outreach—and created a master list of 597 organizations. While we can’t know for sure how many others are out there, the master list appears to be quite comprehensive (the list of organizations that agreed to have their information published can be found in the ADN Behavioral Teams directory). The survey thus gathered detailed response from 54 percent of this known universe of teams. (More information about the survey methodology, data cleaning process, and analysis can be found in the full report.)
What did we find?
The survey results and master list of organizations are rich in detail, and we continue to mine them for insights. However, here are a few key findings thus far:
Most behavioral teams are in for-profit companies
Of the teams surveyed, 385, or 65 percent, are within for-profit companies (N=592 organizations with known type in the master list). Ninety different academic institutions with explicitly behavioral science groups have been identified thus far, as well as 63 government institutions and 54 nonprofit organizations. Academic institutions and government agencies often have a different focus—both in why and how they apply behavioral science—than for-profits and nonprofit organizations. Thus, to focus the rest of the discussion, we’ll look specifically at for-profits and nonprofits.
Indeed, the range within business and nonprofits is vast. Organizations as varied as large multinational firms (Walmart, Coca-Cola), aid organizations (Save the Children, International Rescue Committee), consultancies (Ipsos, Mad*Pow), and finance firms (NASDAQ, Morningstar) all have behavioral teams.
Our field is rapidly growing, but still quite small
While many of these organizations are large, the respondents indicated that the footprint of behavioral science was actually quite small. The median team size was four people; within companies and nonprofits, the largest behavioral science teams employed around 200 people. In total, the respondents represented corporate or nonprofit teams with 1,925 members and indicated that another 1,424 individuals applied behavioral science on other teams within their companies.
Combining these figures, and assuming that the survey represented just over half of the worldwide total (see above), we can very roughly estimate the total worldwide employment within companies and nonprofits. These teams appear to employ around 6,319 people.
Six thousand people employed in applied behavioral science in nonprofits and companies is notable, but not large. To put that into perspective, there are an estimated 200,000 psychologists in the United States alone.
Consultants are common; otherwise teams are embedded in existing departments
Over half of all companies and nonprofits are consulting or contracting firms (55 percent of survey respondents and 53 percent of such organizations on the master list). These range from major firms like BVA to small behavioral science–focused consultancies that have popped up over the years, like The Behaviorist in Canada and Habittude in Spain. Putting aside those who are in external consulting, the most common placement of these teams was within data science (32 percent) departments, followed by product (30 percent), marketing (25 percent), and design (20 percent). Teams within human resources, while an exciting new area of behavioral science, are relatively rare (5 percent), as are standalone teams within organizations.
Team composition and origin
There appears to be no single path by which behavioral teams are started within an organization. Our respondents described a mix of bottom-up and top-down approaches—from starting a small new company specifically geared toward behavioral science (32 percent) to a CEO or department head driving it (20 percent, 16 percent) to individual contributors making it part of their work and growing from there (18 percent). What was uncommon, however, was someone outside the company convincing the company to start one (2 percent); that is, behavioral (organization) change comes from within. As an aside, that is how it has happened in my behavioral roles at Morningstar and HelloWallet: I was already an employee of each company and started our teams from within.
When it comes to the individuals on the team, 52 percent said that they had a formal degree in behavioral science. Among the others, 83 percent learned through books, 77 percent on the job, or through formal coursework (42 percent) or informal online classes (52 percent) that did not result in a degree in the field.
Are they “real”?
One of the most common refrains I’ve heard among academically trained behavioral scientists is that the numerous companies out there simply aren’t doing “real behavioral science.” The survey results do not support that cynicism. We find that teams are using well-established techniques such as social influence (83 percent; e.g., social norms and social proof), manipulation of the choice set and of attention (each 74 percent), and habit formation (58 percent). Perhaps more importantly, 69 percent of respondents said their teams measured their success in terms of A/B tests or other forms of RCTs. We should be cautious though: the median number of experiments the teams conducted in the past twelve months was only five.
One can readily find examples of companies that appear to be using behavioral science in name only. But in our analysis at least, that appears to be the exception rather than the rule—organizations are truly seeking to apply behavioral science and do so with some level of rigor.
One can readily find examples of companies that appear to be using behavioral science in name only—both through a quick search on the internet and among survey respondents. This is especially true in the realm of consulting, where companies big and small have jumped on the behavioral science bandwagon and are likely to move on to whatever is in vogue soon enough. But in our analysis at least, that appears to be the exception rather than the rule—organizations are truly seeking to apply behavioral science and do so with some level of rigor.
Two types of challenges were identified in the survey. The first were operational. The single biggest challenge that teams faced was getting their interventions implemented in practice (42 percent) or measuring their impact (41 percent); generating ideas for interventions was not commonly a problem (14 percent). In the comments and subsequent interviews, respondents similarly mentioned that some of their key challenges were implementation and impact measurements. This was especially the case among consulting and contracting shops where the client does not recognize the need for that work (they simply “want the solution”).
The second challenge is the ethical application of behavioral science. While that wasn’t the focus of the survey, we could see warning signs on the margins. The most common purpose to which behavioral science was applied is product use (62 percent) and sales (52 percent). On their own, this is understandable and likely without problem. However, in most cases the target audience for these interventions did not know about them—41 percent of respondents said that virtually no users know about the organization’s use of the behavioral interventions; 18 percent said a few do; and only 22 percent said that most people did or everyone did. With high-profile examples of behavioral science being used in hidden and harmful ways in companies from Uber to ThredUp, we can reasonably guess that other less-public examples exist as well.
Overall, we found a certain level of optimism within the teams—although it’s worth noting the majority of respondents completed the survey before COVID-19 hit. Over 50 percent of applied behavioral science teams expected to grow within the following 12 months. The median behavioral team expected to expand by 25 percent (average increase: 53 percent), which if it held true would entail a growth of 1,580 (3,349) roles over the course of the year.
Similarly, we find an openness to collaboration. Sixty-five percent of teams within for-profit companies reported that they were or previously had collaborated with academic partners in applying behavioral science. Forty percent reported they collaborated with other companies, such as clients and vendors.
Both of these are promising, and hint at further growth and cross-pollination in the years to come.
Where to learn more
Our field is growing, diverse, and fascinating. While I could only cover a subset of the survey results here, you can receive the complete results at BehavioralTeams.com. Anyone who enters their email address will get a copy of the report. If you yourself work on an applied behavioral science team, and haven’t already done so, please do complete the survey as well; it’s because of input from hundreds of people around the globe that we’ve been able to gather this understanding, and we’d appreciate your help continuing to expand that understanding and sharing the results with the broader community.