DataRobot Humans and AI Blog image BG v1.0

Humans and AI: AI, Marketing, and Behavioral Economics

July 6, 2021
by
· 5 min read

Who loves paying taxes? No one? I thought so. Yet it’s possible to nudge people to more willingly pay their taxes on time. 

The Behavioural Insights Team, also known unofficially as the “Nudge Unit,” was founded by the UK government in 2010 to use behavioral science to make public policies and services more effective. One of its more successful experiments is the use of peer pressure to improve tax collection. 

Ordinarily, HM Revenue and Customs, the department responsible for tax collection, sends a reminder notice to those who haven’t paid their taxes on time. Only 33% of those who receive the reminders respond by paying their taxes.

dreamstime m 154297721

The Behavioural Insights Team ran an experiment, testing four variations to the language used in reminder letters on a trial group. 

  1. “Nine out of ten people pay their tax on time. You are in the minority that does not pay their tax on time.”
  2. “Nine out of ten people in your local area pay their tax on time. You are in the minority…”
  3. “Nine out of ten people with a debt like yours pay their tax on time. You are in the minority…”
  4. “Nine out of ten people with a debt like yours, in your area, pay their tax on time. You are in the minority…

Although each version of the reminder used peer pressure to nudge behavior, the fourth variation, the one with the greatest specificity, was the most effective. Tax payment responses went from 33% to 39%. Although 6% might not seem like much, that’s 18% more people paying their taxes.

Wouldn’t it be awesome if you increased sales by 18%?

Old-Fashioned Service

When I was a child in the 1970s, shopping was a different experience. Shopping malls were rare, and most stores were not part of a chain or franchise. My family bought meat from a butcher shop that was owned by someone who lived around the corner from us. One of the employees knew us by name. He remembered our usual orders and favorite cuts of meat. Shopping was a social experience. Stopping to chat was an important part of customer service.

Contrast that experience with modern shopping. Retailers have become more efficient by embracing economies of scale. Malls are commonplace. We buy prepackaged meat from a supermarket. In this time of social distancing, we order supplies on the internet and have them delivered to our homes.

Humans evolved to live in tribes and small communities, and it shows in the cognitive limits of our ability to connect with people as individuals. In 1992, the Journal of Human Evolution published “Neocortex size as a constraint on group size in primates,” which measured the correlation between neocortical volume and typical social group size in primates and human communities. The paper concluded that there is a limit imposed by neocortical processing capacity that appears to define the maximum number of individuals with whom it is possible to maintain stable interpersonal relationships. In the 1998 paper “The Social Brain Hypothesis,” the same author concluded that the size of the human brain’s neocortex “represents a biological constraint on social interaction that limits humans’ social network size to between 100 and 200 individuals.” 
The middle of this estimated range (that is, a value of 150) is referred to as Dunbar’s Number, named for the British anthropologist Robin Dunbar, who discovered this limit. More recent research, “Modeling Users’ Activity on Twitter Networks: Validation of Dunbar’s Number,” has

confirmed that this cognitive limit on social relationships extends to social media. Our technology has enabled more connection than ever, but our neurocognition still experiences a fundamental constraint.

Dunbar’s Number also places a limit on the scalability of human-driven sales processes. We’ve traded old-fashioned, personalized service for scale and convenience.

AI, Sales, and Marketing

One of the comparative advantages that computers have over humans is the ability to handle vast amounts of data and operate at scale. Without computers, banks couldn’t handle trillions of dollars of transactions each year. And without computers, businesses wouldn’t be able to support millions of customers.

In the twenty-first century, we’ve seen the rise of the internet and AI. Recently published research shows that AIs can match the performance of top human sales staff, achieving four times more sales than your inexperienced sales staff. So, it’s no wonder that some of the most popular AI use cases are in sales and marketing. AIs can deliver hyper-personalization at scale for sales and marketing use cases such as lead scoring, upselling, cross-selling, next-best-offer, and next-best-action.

AI is not held back by Dunbar’s Number. It can deliver old-fashioned personalized service at scale. But can it do even better than that?

AI Can Use Behavioral Economics Too

Behavioral economics can use peer pressure to nudge people into better behaviors. The same approach works in marketing. The greater the similarity and specificity, the more effective the nudge to purchase.

Paradoxically, the trick to applying peer pressure in marketing is to apply hyper-personalization, and that’s something AIs are great at.

Imagine a bank that is planning to cross-sell personal loans to its customer base. The bank would typically use AI-powered lead scoring to build a shortlist of the customers most likely to be interested in the personal loan offer. Behavioral economics tells us that we should also find one or two attributes of each customer that makes them likely to respond positively to the marketing offer, and then share a statistic about how peers with those characteristics are more likely to use this product. This is a more personalized form of peer pressure.

pasted image 0 36

In this example, the AI system powering the lead scoring and customizing the email explained that David was included in the marketing campaign because of his age and home address. Peers with similar characteristics are 1.9 times more likely than the average bank customer to take a MegaLoan.
Using AI-driven marketing, lead scoring plus behavioral economics, you can select the best prospects and nudge them so that they are even more likely to purchase.

Humans and AI Best Practices

Humans are complex. Although we want to be treated as individuals, we also want to conform to peer behaviors. AI-driven hyper-personalization offers a way to treat your customers as individuals and as members of a community. Research shows us that best practice is to tell your leads and customers why they are likely to purchase and how they are similar to people who purchase.

You’ll need best-practice AI to take the step up to best-practice marketing. Use advanced machine learning algorithms to deeply understand customer purchasing behavior. Hyper-personalize your communications with leads and customers. Use prediction explanations to understand why they are likely to purchase, and as the intelligent source of targeted peer comparisons.

Demo
See DataRobot in Action
Request a demo
About the author
Colin Priest
Colin Priest

VP, AI Strategy, DataRobot

Colin Priest is the VP of AI Strategy for DataRobot, where he advises businesses on how to build business cases and successfully manage data science projects. Colin has held a number of CEO and general management roles, where he has championed data science initiatives in financial services, healthcare, security, oil and gas, government and marketing. Colin is a firm believer in data-based decision making and applying automation to improve customer experience. He is passionate about the science of healthcare and does pro-bono work to support cancer research.

Meet Colin Priest
  • Listen to the blog
     
  • Share this post
    Subscribe to DataRobot Blog
    Newsletter Subscription
    Subscribe to our Blog