Marketing Insights and Analysis
Oprah Winfrey was introduced to Twitter on April 17, 2009 by Evan Williams and Ashton Kutcher. Within moments of her first tweet, thousands of people responded to her.
This could have been a transformative moment in traditional and social media. For a brief moment, television’s arguably most successful personality was engaging directly with her audience in a dynamic conversation on social media’s hottest platform. Instead of engaging in dialogue, Oprah stuck to the traditional broadcast metaphor that built her empire and with which she was comfortable. @Oprah’s limited replies to people, other than celebrities for dinner invitations, demonstrate that she’s using Twitter more as a distribution tool than for conversation.
Twitter is a conversation engine that functions when people post and respond. Twitter can be used for mass distribution of information, but it thrives when the conversation and interaction is diverse.
Oprah hasn’t engaged with Twitter much since her first tweet. In the period since, she’s tweeted 24 times. The most recent tweet at the time of this writing was on May 2, two days ago.
Oprah’s usage isn’t different from many Twitter trial users. Much has been made of Twitter’s low retention rate. The retention rate of users that return the following month has been estimated at 40%. For every 10 people that try the service, 6 people disengage almost immediately (that 40% is listed as a low retention rate from a trial is something we’ll challenge in a different post).
What is different with Oprah’s trial is the audience she could bring. When the press caught wind of Oprah’s disinterest in Twitter, they used the information as an indicator of Twitter’s imminent demise. Twitter had peaked and was unable to reach beyond it’s roots with the digerati. The technology press then rebutted that assumption, stating that Oprah’s non-use of Twitter had limited relevance to the growth of Twitter.
So what does Oprah’s Twitter experience tell us? Let’s first take a look at the data.
Oprah has been a light user of Twitter, at 2.6 tweets per day. A brief scan of her Tweets shows she has responded to 5 different people, but primarily uses the medium to broadcast select messages. @Oprah’s follower count grew to 784k within a month, but has since stagnated to nearly no growth. A Tweep who has an 810,000-to-11 follower/following ratio is using it as a broadcast medium, not surprising given Oprah is one of the most famous television celebrities.
So how does this compare with the other leading daytime television hosts?
Of these daytime hosts, @MarthaStewart has adopted Twitter in a way truest to it’s intent as a conversation engine. Most telling of Martha’s adoption and understanding of Twitter and social media was this post.
While it may not be representative of the overall TV audience, I do believe Oprah.com’s online audience is a fair representation of the potential audience that would use Twitter.
| Oprah.com | Twitter.com | Metric | |
|---|---|---|---|
| Gender | 137 | 103 | female |
| 60 | 96 | male | |
| Age | 9 | 18 | 0-17 |
| 89 | 153 | 18-34 | |
| 143 | 124 | 35-49 | |
| 151 | 88 | 50+ | |
| Ethnicity | 98 | 103 | Caucasian |
| 173 | 101 | African American | |
| 61 | 66 | Hispanic | |
| 60 | 92 | Asian | |
| Household | 115 | 112 | No kids |
| 78 | 82 | Kids | |
| Affluence | 96 | 120 | $0-$30k |
| 101 | 106 | $30-60k | |
| 99 | 83 | $60-$100k | |
| 102 | 97 | $100k+ | |
| Education | 89 | 83 | No college |
| 108 | 114 | College | |
| 106 | 110 | Grad school |
From http://www.quantcast.com/oprah.com and http://www.quantcast.com/twitter.com
An index of 100 would mean a composition that was representative of the US population.
Oprah.com’s visitors skew higher in female, African American, and are slightly more affluent than Twitter. Twitter has a higher Asian composition and a significantly larger number of younger visitors.
To complete the picture, I next conducted a limited, non-scientific poll of 100 random respondents to see what intersection occurred between Oprah viewers and Twitter users.
At first glance, 14% seemed like a high number of intermittent Oprah viewers, but Oprah averages 23 million viewers in the United States each week. That’s approximately 13% of the total US population. Her daily audience is estimated at 7 million viewers. Comparatively, Twitter.com reaches an estimated 15 million people per month (note: this is the web only audience, and is likely well below actual usage numbers).
On the surface, Twitter had a momentous opportunity with the Oprah exposure. Like products that have “made it” by appearing on QVC, an adoption of Twitter by Oprah could lead to the bring millions of new Tweeps.
So what was the size of the opportunity? If we assume the survey results were valid, the size of the total opportunity for Twitter in pure numbers was the 6% of Oprah viewers that weren’t Twitter users. 6% of the her weekly audience is 1.3 million viewers. Twitter’s March 2009 growth rate was 131%. If that growth rate was maintained, Twitter grew by nearly 1.1 million viewers each week following Oprah’s trial (excluding her viewers). Oprah’s non-use of Twitter would not have a material impact on the growth and adoption of Twitter or other social mediums.
What was missed in the press coverage was the opportunity of Twitter to expand into new growth markets. Oprah’s audience skews very high in the African American and over 50 age segments, areas where Twitter is weak. Twitter trails Oprah by nearly 90 points in the African American segment alone. While Twitter did not need the sheer number of visitors offered through Oprah, the influx of these unique segments would potentially help diversify the conversations taking place on Twitter and help entrench it as the place for conversations.
Oprah doesn’t need Twitter to continue to be successful. Syndicated shows like Regis & Kelly haven’t changed their format in years and retain a strong, if aging, population. Oprah could profitably ride the format of her show for years and retain a healthy audience.
A closer look at Oprah’s demographics show two troubling trends. First, her audience is aging. The majority of viewers are over 35 with a large number over 50. Secondly, her audience skews low with Asian and Hispanic demographics, and high in African American viewers. Hispanic and Asian populations are increasing (as a percentage of the population), while African American’s are decreasing. While Twitter is weak in its’ Hispanic audience, it has strong numbers in each of the other segments.
If we assume equal effectiveness at conversion, Oprah would derive as much or more value from engaging in Twitter and capturing these new segments, than Twitter would from converted Oprah viewers.
Oprah does not seem content to rest on her laurels, and her career points to a continual evolution in how she engages the community. As her competitors in daytime talk continue to effectively leverage Twitter as a way to interact with their audience, it will be interesting to watch if she re-engages with Twitter in the future.
You must be logged in to post a comment.
Information is essential in generating good insights, but it cannot be a crutch in making decisions. The posts here are intended to explore and are not perfect, but that's part of the point.
Craig
May 5th, 2009 at 12:10 pm
I disagree that "Oprah.com’s online audience is a fair representation of the potential audience that would use Twitter." The main overlap you have is gender, but apart from that demographic, nearly every other attribute of Twitter users is contrary to Oprah fans: Twitterers tend to be young, upwardly mobile, employed, and socially active. Oprah's audience is dominated by middle-aged, stay-at-home moms, who aren't nearly as enamored with computers as those 20-somethings that have to stare at one all day. Is there some overlap? Yes, but not as much as you think. Plus, as Oprah herself ages, her audience will tend to age with her (this is a common phenomenon in persistent television personalities), moving her further away from Twitter's true growth engine: college students & young professionals.
Plus, there's the issue with your convenience sample, which isn't very representative of the larger population, so extrapolating your findings from it to the Internet at large is unlikely to be very accurate.
Sorry to be a nattering nabob…I appreciate the effort and like most of what's in the post…just those two things didn't sit right with me.
Doug Schumacher
May 5th, 2009 at 3:28 pm
Really good post, Rob. I've actually been looking at some Twitter figures myself lately for a possible article — comparing how celebs use it vs new media people.
I like how you brought in the Quantcast figures to project potential future growth, especially regarding Oprah and her current viewership demos.
Of your survey (understandably non-scientific), curious if that was centered towards new media people, as the 84% using Twitter regularly seems quite high.
Really liked the post. (preferred your previous design template, though
doug
ericdionbaker
May 5th, 2009 at 7:31 pm
Rob – Great post! My response would not fit here, so I posted it on my blog instead:
http://reconstructingedb.wordpress.com/2009/05/05...
Great post, bro!!
robsaker
May 5th, 2009 at 8:58 pm
Oprah.com\\'s online audience is a fair representation of her potential that would use Twitter. That\\'s not a stretch. A user that visits Oprah.com is likely to be a viewer of the show, but has a base technical ability required to leverage Twitter. Are there Oprah viewers that don\\'t visit Oprah.com? Absolutely, but we have no logical reason to believe the differ from the audience that does visit Oprah.com. Using Oprah.com\\'s numbers are also an acceptable assumption in they be more similar to the actual segments of her audience that would be willing to try Twitter.
I would challenge the argument that stay-at-home mom\\'s aren\\'t enamored with computers. They may not be as engaged as 20-somethings, but experience working with clients in this segment shows this group consistently demonstrates interest and a stronger need to socialize, two key drivers for the adoption of social media.
Regardless how you read the tea leaves, the differences you mention in segments were in fact one of the main reasons for the post. Oprah\\'s audience is aging and she is distancing herself from new viewers. My argument is that Oprah, more than Twitter, would have benefited from the ability to reach and engage new viewers.
robsaker
May 6th, 2009 at 7:47 pm
Thanks Doug. Yep, the survey was focused towards the digerati. I was primarily interested in the intersection between Oprah and Twitter users, so the key data point there was overlap. There is bias there, but I don't think it's material (would've needed that group to provide fact).