5 psychological rules you should follow when building a bot

July 24, 2016

This post first appeared in VentureBeat.

B.J. Fogg is a behavioral scientist who has written extensively about how computer products influence people.

He coined the term “Captology,” which is an acronym for computers as persuasive technologies. When people view a product as having a life, the computer system inside the product can leverage the principles of social influence to motivate and persuade.

In his book Persuasive Technology: Using Computers to Change What We Think and Do, Fogg proposes five types of social cues that can cause people to make inferences about the living presence in a computing product.

My advice? Consider these social cues when designing and building your bot. Here they are, with my comments.

1. Physical rule

When we find someone or something attractive, the “halo effect” kicks in. We view the attractiveness of the entity as having other positive traits. Thus the face or look of your bot should be well researched during the design phase. Is a robot face really resonating with your target demographic? Does it have some appeal?

The concept of attractiveness extends to voice, too. Think about what Apple did with Siri. Users can choose the accent which they find most appealing. We are also more forgiving to entities which we interpret as more attractive. So as you work through the bugs within your bot, make sure your target demographic finds it attractive.

2. Psychological rule

Robert “The Botfather” Hoffer says we should inject attitude into chatbots as a way to establish a genuine sense of connection with the user. Cues hinting to your bot’s emotion can cause a user to subconsciously feel the personality of your bot. This is best achieved through well-researched conversational copy. It should seem genuine and real.

Think through the range of the conversational experience. How does your bot respond when a user expresses a positive sentiment? How do they respond when a user needs help? These are all great opportunities to inject a flavor of branding. You can even add a little moxie to your bot by adding a few Easter eggs. Bots designed with a well-personified brand are more likely to influence decisions.

#Bots designed with a well-personified brand are more likely to influence decisions. Click To Tweet

3. Language rules

The language of your bot should never be neutral. Words rule everything in the world of conversational software and messaging design. They are the only asset you have to establish the voice and tone of your bot. The importance of investing in well-written and well-researched conversation copy should never be overlooked.

If your bot is built to create strong engagement over time, consider offering periodic praise. According to Fogg, this is one of the most powerful persuasive tactics that can ultimately lead the user towards a desired action.

4. Social dynamics rule

We are more easily persuaded by those who we view as similar to us. This is another powerful persuasion principle highlighted by Fogg.

For your bot to be an effective persuader, the user should feel as if the bot is similar to them. A great way to determine which conversational traits of your bot work best is to leverage multivariate testing through Msg.ai. However, at the time of this writing, this tool is only available to enterprise companies and major brands.

5. Social roles rule

Bots can also assume roles of authority. Fogg says that “for computers that play social roles to be effective in motivating or persuading, it’s important to choose the role model carefully or it will be counterproductive.” Your bot doesn’t need to necessarily pass the Turing Test, but the sooner it can establish credibility, the sooner its authority and subsequent ability to persuade will grow.

As bots continue to make the leap outside of the Silicon Valley bubble and into the minds of the mainstream, I see a future where well-designed bots full of personality and interactivity will become a persistent force influencing our decisions via persuasion.