Facebook finds a way to predict your ethnicity
By Mary Ellen Slayter on August 2nd, 2010 | 1082226 comments on this posthttp%3A%2F%2Fsmartblogs.com%2Fsocial-media%2F2010%2F08%2F02%2Ffacebook-finds-a-way-to-predict-your-ethnicity%2FFacebook+finds+a+way+to+predict+your+ethnicity2010-08-02+11%3A31%3A22Mary+Ellen+Slayterhttp%3A%2F%2Fsmartblogs.com%2Fsocialmedia%2F%3Fp%3D10822
This post is by Kaukab Jhumra Smith, a contributing editor at SmartBrief.
A team of researchers from Facebook has come up with a way to predict the ethnicity of Facebook users in the U.S. by using a combination of U.S. Census data and an analysis of users’ first and last names. The team won the award for best paper at a social-media conference organized by the Association for the Advancement of Artificial Intelligence in Washington, D.C.
The Facebook Data team confined its analysis to the country’s four largest ethnic groups: Caucasians, African-Americans, Asian/Pacific Islanders and Hispanics. It used Census records to identify the most likely ethnicity of a user based on his or her last name. For example, said data scientist Jonathan Chang during his presentation, if your last name is Mueller, Census records show you have a 97% chance of being white. Similarly, if your last name is Washington, there is an 89.9% chance you’re black.
The team then refined its predictions by factoring in users’ first names, Chang said. That helped reduce confusion, for example, between Caucasian and African-American users with the same last name. A user with a first name such as LaToya was probably African-American, Chang explained.
By applying their predictions of ethnicity to users’ friend networks, the team found:
- The ethnic makeup of Facebook users has steadily become more diverse and now generally reflects the U.S. population, unlike a few years ago, when Caucasians and Asian/Pacific Islanders were over-represented.
- Users are more likely to be friends with, and communicate most often with, people of the same ethnicity.
- Users are more likely to be in romantic relationships with people of the same ethnicity.
- Particular ethnic groups tend to behave similarly online, with Asian/Pacific Islanders engaging in “unexpectedly high number of wall, video, note, gift, comment and group-sharing actions.”
The team acknowledges a few caveats: Its analysis didn’t account for a lot of factors, such as socio-economic status and education, and it didn’t include smaller ethnic groups in the country. The researchers plan to use more detailed Census data to factor in user locations, professions and other self-disclosed data to “improve the predictive power” of future analyses, the paper concludes.
So what does this mean for businesses looking to refine their marketing on Facebook and other social networks?
Approached after his presentation, Chang said he knew little about business and could not answer how his team’s findings would apply to consumer marketing.
But surely, now that it’s possible to predict ethnicity on Facebook with relative accuracy (accuracy that will only improve with subsequent studies), the next logical step would be to allow marketers to target users based on ethnicity and not just age, location and known preferences.
It would be a simple matter to rotate racially appropriate photos in Facebook ads aimed at 30-year-old women in the Washington, D.C., area, for example, in the hope that they would respond better to someone from their own race.
What do you think — does that kind of marketing cross a line or is it just one more refinement on what advertisers do all the time?
Image credit, Rich Hobson, iStock Photo
Related Posts
- Is the social media universe still expanding?
- How retailers can use social media to attract more customers
- Are video ads good for Facebook?
- 6 ideas to help take advantage of Facebook’s new cover photo rules
- Why Google+ is the better platform for marketers
Twitter
Facebook
Linkedin
RSS





[...] This post was mentioned on Twitter by Jason Murphy, youredgeonline, Michael Garvin, RT17, Social Media Agency and others. Social Media Agency said: #SocialMedia Facebook finds a way to predict your ethnicity – This post is by Kaukab Jhumra Smith, a contributing ed… http://ht.ly/18ouDl [...]
The only thing I could think of while reading this article was the "Donna Chang" episode of Seinfeld. As a marketer, I'm very wary of these predictions.
Some of this is pretty obvious, but there's a limit to predictive power AND ethnic groups are not monoliths! What ethnic group would this program predict for Donovan McNabb?!
An Orangeman of course.
Sorry– a bad, but necessary joke from an alum.
I agree with LilOle – Wow, thinking that people would respond better to ads showcasing some of their own race is just starting to push that "over the edge" button. I actually prefer ads that showcase something that appeals to my want or need.
My recent post Top 10 Things Social Media Isn’t – A List in Progress…
Agree. Something that targets by race will also include and exclude by race, and that puts a very awkward spin on the user experience and often times what people go online for. What users see and are offered is no longer based on real evidence, like user's behavior, listed interests and hobbies, groups, etc. No, this method sneakily makes presumptions about a user's race and adds that with data that (already) overly-generalizes the behavior of ethnic groups. This gets extremely muddy. Who says I want my online experience altered by who you think I am? Understand the lure for this, but it's also an extremely dirty road to go down… and it will get messy.
Steps too close to the edge for me, but I've already got issues with Facebook's privacy policy. It's one thing if an individual's FB activity provides such information, but to go through the back door, so to speak, seems a bit like walking that double-edged sword — one wrong move and OUCH.
My recent post Planning a Hike in the Whites Here’s How
Wow, its perfectly all rigth to implement etnic targeting in Europe. So far all the advertising regulative strictly forbidden to serve ads on languages which are not national, and to lead on landing pages on languages other then official.
Example: last few months merely all of us saw ads in Albanian language on Facebook. Which caused interesting discussion – does Facebook recognizes Kosovo as Serbian teritory, does facebook knows something we dont know? btw. Kosovo is part of Serbia with Albanian inhabitants who recently claimed their independence from Serbia Yet Akbanian language was never spoken in Serbia. Also, some Russian or Bulgarian ads are visible on Facebook (for Stella Artois), which is, as Ive said, strictly forbidden for all other ad networks in Europe. If you serve ads to Hungarian people at least landing page has to be in Hungarian. If you are selling something to Austrian people and your company is from Slovakia, its mandatory to have web site in German language etc.
This might solve our European confusion and help advertisers to serve more relevant ads.
[...] have found ways to extract ethnicity from social media profiles. Kaukab Jhumra Smith writes, “Facebook Finds a Way to Predict Your Ethnicity“: The Facebook Data team confined its analysis to the country’s four largest ethnic groups: [...]
[...] have found ways to extract ethnicity from social media profiles. Kaukab Jhumra Smith writes, “Facebook Finds a Way to Predict Your Ethnicity“: The Facebook Data team confined its analysis to the country’s four largest ethnic groups: [...]
I find it unlikely that this is very accurate — as I am Japanese, Native American, Scotch/Irish/English with very non-ethnic names. More and more people are identifying as multiracial and how are you going to target someone like Tiger Woods — black Asian Hawaiian?
Yet another reason NOT to have a Facebook account.
This one's a tricky one. On the one hand I can see why people might be offended at advertising targetted by ethnic origin etc, but on the other hand it's perfectly valid for advertisers to want to ensure their ads reach their target audience and understandably so.
It doesnt worry me that Facebook is able to use data to auto predict ethnicity now, to be honest. With all the data available online and off, it comes as little surprise really.
My recent post 5 Stalker-Like Tendencies Facebook Has Instilled in us ALL!
[...] Knows Your Ethnicity – Facebook has figured out, with remarkable accuracy, what your ethnicity is based on the data you share on the site. A Facebook Data team recently won a team award for best [...]
[...] Knows Your Ethnicity – Facebook has figured out, with remarkable accuracy, what your ethnicity is based on the data you share on the site. A Facebook Data team recently won a team award for best [...]
I wouldn't be offended at being targeted for being white, so why would other people be offended for being targeted for being whatever race they are? I mean, there are sensitivities based on previous exploitation, but…
[...] on ethnicity. Profile pictures can be a dead giveaway in many cases. And several readers of my related post on the SmartBlog on Social Media pointed out that predictions of ethnicity based on names are becoming quite meaningless in an [...]
Thank you for all your comments! I just posted on my own blog about an experience I had with a Facebook ad some weeks ago. The ad, for a freelance writing site, featured a photo of a brown-skinned woman who looked uncannily like my mother. As a freelance journalist and a South Asian woman, I felt compelled to click on it.
It made me wonder if Facebook has already begun to implement some of the findings from this paper in its marketing. It's just a hunch, of course. I have no proof. My last name, "Smith," does not exactly sound South Asian, and my full name gives away no information you couldn't deduce from my Facebook profile picture.
You can find my full blog post here. http://www.kokismith.com/2010/08/creepy-or-inevit….
My recent post Creepy or inevitable
[...] Facebook finds a way to predict your ethnicity [...]
@Lee. I agree. I was born in Philippines with a mix of ethnicies (Spanish/East Indian), my first name doesn't give away my roots and moved to a small town in Newfoundland at an early age. I identified more with the Newfie culture and still do despite living in Toronto for half of my life.
I think this kind of thing would be inaccurate for some. I have a French grandfather and German grandmother about five generations back, and so my last name is French, but I'm Mexican American. Most people identify me as Caucasian with looking at my name only, but then they see what I look like. Stereotyping at its most…not best, not worst, but most.
[...] Is ethnic targeting coming to Facebook? Facebook researchers have figured out a way to guess users’ ethnicity with extraordinary accuracy, simply by cross-referencing first and last names against U.S. Census data. “Surely, now that it’s possible to predict ethnicity on Facebook with relative accuracy … the next logical step would be to allow marketers to target users based on ethnicity and not just age, location and known preferences,” writes Kaukab Jhumra Smith. SmartBrief/SmartBlog on Social Media [...]
This is nothing new. I used to do circulation marketing for a non-profit newspaper catering to a specific ethnic group, and I remember hearing about direct-mail lists for sale that took educated guesses on someone's ethnicity (though I'm not sure we ever bought one). Of course they're not going to be 100-percent accurate, but then what targeted-marketing technique is?
[...] from Facebook.” Sounds like a horror movie. Anyway. The researchers in question have found a way to predict Facebook users’ ethnicity by fiddling with name data and Census data. Among the team’s findings: The ethnic makeup of [...]
It's none of my business how they choose to advertise, but I would be offended if I only see ads from just one ethnic group. It would be like being accused of being racist or something. So I'd just use the firefox addon Ad Block to stop seeing ads all together. Then they will get less money from me.
[...] accurate and privacy-friendly system than what seems to be the current trend where companies use data analysis to uncover this data regardless of what the candidate wants. For example, somehow LinkedIn offers the ability to target [...]