Is web marketing heading for a personality crisis?

Mimicking how social networks analyze user content for advertising purposes, a Berkeley-based startup, Five, has launched a web tool that predicts users’ personalities based on their Facebook posts.

Five Labs (labs.five.com) examines the linguistic content of Facebook wall posts, pulling key words to make predictions about individual personalities. Using an artificial intelligence method designed by H. Andrew Schwartz, researcher at the University of Pennsylvania’s Worldwide Well-Being Project, the application’s predictions are based around five personality traits: Extraversion, Openness, Neuroticism, Conscientiousness, and Agreeableness.

The fun part comes when you compare your personality makeup to Facebook friends (including significant others) as well as public figures. Those include many in the tech world (Mark Zuckerberg, Bill Gates, etc.) as well as everyone from Barack Obama to actress Jennifer Lawrence and Mahatma Gandhi. Users can, for instance, see if their personality overall is similar to Lebron James, and gauge whether they’re more or less neurotic than the NBA star or how their other specific personality traits match up.

Nikita Bier, the co-founder of Five, told The New York Times that based on the initial responses to the site on Twitter, "people seem to identify pretty strongly with the personalities we generate." He added, "Only about 10 percent said we were outright wrong about them."

5labs

Quentin Hardy, the Times writer, said the feeling of recognition may be due to the "overall vagueness" of the typecasting, likening it to how people identify with horoscopes.

That view works for Five, which is working to launch an application focused on private social experiences in Fall 2014. On Gigaom.com, Derrick Harris writes that Five "wants to give us a new way of communicating with our friends that’s more akin to a dinner table conversation than it is to shouting to one another across a crowded restaurant."

In its press release, Mr. Bier notes that although Five Labs employs similar techniques as highly public platforms like Facebook, Twitter and Google, Five Labs never stores the information.

"Think of this as a personality snapshot," suggests Mr. Bier. "It’s all for fun." He adds, "But we’re also hoping to educate. People need to ask themselves a profound question: ‘How does my data portray me on public networks — and how might that be used?’"

BrainTrust

Discussion Questions

What are the pros and cons for marketers of matching consumers with personality types online? Would a social network experience based more closely on “private social experiences” work better for both individuals as well as marketers?

Poll

7 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Ian Percy
Ian Percy
9 years ago

Yup, if I were a retailer I’d be training all my staff on how to use this app to figure out Facebook’s analysis of customers’ personalities and then correlate that typecast with specific merchandise which has also been pre-categorized by social media algorithms. The critical key is to correlate the salesperson’s socially-generated personality type with the customer’s so that potential personality conflicts are avoided. Or, you could teach your salespeople how to engage customers in a thoughtful and caring way, find out what they want to buy and sell it to them.

Edward Chenard
Edward Chenard
9 years ago

I’ve used this tool, it’s quite easy to fool. Also, a person’s Facebook self is not their actual self, so you don’t get a good picture of what someone really is like by just looking at Facebook. Even when you pull in the external data that Facebook is collecting, it just doesn’t match because it is completely devoid of intent. Without intent understood, you just have interesting data that doesn’t really give you much to go on.

Lee Kent
Lee Kent
9 years ago

The idea of being able to get down to personas from social activity has a lot of appeal. However, we ain’t there yet! This strategy is indeed too vague and does not provide the level of personality needed to create campaigns.

Keep at it, though. My two cents says we’re headed in the right direction.

Ralph Jacobson
Ralph Jacobson
9 years ago

If the data are trustworthy, then this app could have legs. However, since a significant percentage finds no correlation to their true personality, I would be very wary at this point to leverage any deep insights from the outputs.

There could be great merit in matching personalities for both retailers and CPG brands. The challenge comes when we finally determine that the findings are accurate and then try to assimilate any meaningful actions that may drive profitable growth.

Larry Negrich
Larry Negrich
9 years ago

This might not be the end-all tool, as it makes a lot of conclusions based on very random and sparse data. However, the analysis of public social profiles to improve promotions, engagement, etc. is definitely an area retail marketers should be investigating. The con is that the customer might have some privacy concerns. But as time goes on and individuals continue to (commentary: over) share details of their private lives, that may not be a concern. We’ll see in which direction the winds of sharing vs. privacy blow in the future.

James Tenser
James Tenser
9 years ago

Parsing folks into personality types based on their social media activities—the meta-data or the content or both—will only be useful to retailers to the extent that they are ready to use this segmentation to improve marketing strategy, offer relevance, and merchandising. Ultimately it must deliver ROI in terms of trips, wallet share, and profitable sales.

While personality typing reminds me a little of those fortune-telling booths at the arcade, today, I think it’s actually the first harbinger of a colossal change on the way.

Let’s face it: social, local, mobile, online and search activity today dwarf the amount of in-store interactions that take place. The total volume of data generated in SoLoMoMe makes modern retail enterprise systems seem downright puny by comparison.

So how is a retailer to prepare to serve shoppers who arrive pre-influenced, pre-researched and often pre-decided by their experiences outside the store?

Step number one is learn how to pay active attention. Personality parsing may not be very much yet, but it’s aimed in the right direction, and the accrued insights and disciplines learned will serve the surviving retailers well as they dig into even more responsive ways to serve shoppers.

Christopher P. Ramey
Christopher P. Ramey
9 years ago

Facebook reflects how people want to be perceived; not necessarily who they are or what messaging resonates with them. Although there is a strong argument for email, etc. marketing, I doubt these insights can be executed on the retail floor.