Studying the impact of innovation on business and society

The Interest Graph on Twitter is Alive: Studying Starbucks Top Followers

Social media is maturing as are the people embracing its most engaging tools and networks. Perhaps most notably, is the maturation of relationships and how we are expanding our horizons when it comes to connecting to one another. What started as the social graph, the network of people we knew and connected to in social networks, is now spawning new branches that resemble how we interact in real life.

This is the era of the interest graph – the expansion and contraction of social networks around common interests and events. Interest graphs represent a potential goldmine for brands seeking insight and inspiration to design more meaningful products and services as well as new marketing campaigns that better target potential stakeholders.

While many companies are learning to listen to the conversations related to their brands and competitors, many are simply documenting activity and mentions as a reporting function and in some cases, as part of conversational workflow. However, there’s more to Twitter intelligence than tracking conversations.

We’re now looking beyond the social graph as we move into focused networks that share more than just a relationship.

Bringing the Interest Graph to Life

To demonstrate the value of interest graphs, I worked with the team at ReSearch.ly, a unique Twitter search platform that has indexed the last three years of Tweets to instantly provide a real-time and historical analysis of activity around keywords and also the people that Tweet them.

ReSearch.ly visualizes the interest graph, and also provides the ability to search within the search to sort activity by demographics and psychographics, sentiment, bio data, profession, and the list goes on. Essentially, it’s a product that anyone can use to learn about what’s really taking place on Twitter to better understand behavior and earn greater relevance by making more informed decisions.

As an example of audience profiling or competitive intelligence, we used ReSearch.ly to review the followers of @Starbucks, one of the most celebrated brands actively using Twitter today. We started by extracting 1 million follower profiles, sorted by follower count. The results were then further filtered to include only those who published a complete profile. ReSearch.ly provides the option to then organize the resulting information any number of ways, which in this case, we sorted the accounts by bio, location, and gender.

The Interest Graph

While we are what we say in our Tweets, our bios also reveal a telling side of who we really are. In this study we reviewed the complete bios of 50,000 of the top @Starbucks followers to learn a bit more about how they present their life story as well as their interests, opinions, and preferences.

Using the ReSearch.ly Twitter index, we created a word cloud to amplify the most common words used in each of the bios of these connected social consumers. Followers tended to use expressive words that suggest sentiment runs rich in the Starbucks interest graph. Top words include:

1. Love
2. Life
3. Friends
4. Music
5. World

We can also learn a bit more about Starbucks influencers by analyzing what interests them. Looking a bit deeper into the cloud, we can see that not only do emotions rise to the top; other revealing themes also surface:

1. Family
2. People
3. Mom
4. Wife
5. Husband

This is just the beginning. The words associated with the brands demonstrate the emotional and personal connections Starbucks holds with these tastemakers. Campaigns are a direct beneficiary of such data. As we submerge ourselves one level deeper into the study, we find that this information becomes paramount when we link it to individuals through demographics and psychographics. An import footnote is that the word coffee is among the least used words in the bio, but used nonetheless.

Studying Bio’graphy

With a 50,000-person sample in a traditional research survey, it may be difficult to organize individual responses. Here, we further reviewed each of the bios to find the commonalities in how each person presents who they are in a few precious characters.

Of those, we found that…

– 42 percent expressed strong ties to family, religion, and love

– 29 percent boast special interests, which is further discernible

– 22 percent are professionals who state their current place of employment and position

– 7 percent are students

Additionally, we can extract the attributes of @Starbucks followers further to better symbolize their digital persona. Further review highlights that followers…

– Identify themselves as enthusiasts, geeks, addicts, junkies, creatives

– Define the most popular areas of interest as Music, Food, Coffee, and Fashion

– Potentially favor dogs to cats (2 – 1 as per their mentions)

– Work in either Social Media and Marketing (Note: If we were to change the scale of followers, we would open up the sample to a much broader set of professions)

– Also are still studying. Despite the lower percentage, students account for more than any single professional field

Geo Location: Where in the World is @Waldo?

Brands are more than aware that no one marketing strategy reaches and moves everyone in the same way. Beyond demographic marketing, brands must also focus on driving traffic regionally. Having access to location data isn’t new, but using Twitter as a collective stream of intelligence to identify higher and underperforming locales and associative word clouds allow teams to surface the 3 W’s of real-time geo loco marketing:

Where is negative/positive activity taking place?

Why is it leaning in that direction? And,

What can we do about it?

To give us an idea of where the top @Starbucks followers are Tweeting, we zoomed in to their point of reference. We found that top users tend to Tweet from…

1. California
2. New York
3. Texas
4. Florida
5. Washington

Combining London and UK, we find that The United Kingdom would actually join the ranks of the most often cited cities.

Grouping locations provides a holistic view that provides regional marketing metrics and also areas in need of attention.

Here we can see that the top Tweeps are located in…

– US East, 30 percent
– Non US, 27 percent
– US West, 22 percent
– US Midwest, 21 percent

Tweeting from the Gender Lines

Over the years, I’ve studied the gender makeup of social networks and have consistently found that women outnumber men in some of the most popular networks including Twitter and Facebook. On Twitter, women represent the majority share with 57 percent.

Working with the team here at PeopleBrowsr and ReSearch.ly in conjunction with Klout earlier in 2010, we uncovered en masse, women are more influential than men on Twitter. In fact, the average Klout score within the general Twitter population 34 to 31 in favor of women.

Reviewing Starbucks top followers in ReSearch.ly, it comes as no surprise to see that the women are the predominant source of Tweets, 63 percent women vs. 37 percent men.

The Tweets Have It!: Introducing the Starbucks Brand Graph

The interest graph is defined by connections, but it is brought to life through self-expression. When we combine brand-centric relationships and conversations, the interest graph eventually evolves into what is essentially a brand graph. Within each brand-related graph is a group of highly connected individuals that serve as a company’s network of influence. The ReSearch.ly team extracted 50,000 of the most recent Tweets that included a mention of Starbucks. We then analyzed the connections between people and identified the top 100 individuals and the number of their followers who also mention Starbucks within the 50,000 mentions. We can then bring to light Starbucks influencers as a representation of its brand graph and influential hubs. As we can see, the difference between monitoring and gathering intelligence allows Starbucks to now identify relevant networks and introduce personalized campaigns to further spur advocacy and loyalty.

Here are the top 100 most connected people within the group mentioning Starbucks and the number of their followers also discussing Starbucks:

Accordingly, we can visualize the interest graph as connections, showing how influencers are not only interconnected, but also capable of disseminating relevant information and influencing behavior to varying degrees beyond the traditional reach of Starbucks.  Social consumers and their place within the social consumer hierarchy determine reach and ultimately outcomes. Everything begins however, with recognizing who they are and what inspires or motivates them.

Conclusion

The era of analysis paralysis is officially over. Instead of just listening, companies can now study people and their interests based on what they say and do and also how they color their profiles. This goldmine of insight gives brands the potential to improve marketing, promotional and advertising campaigns to start. What we’re talking about here is the ability to personalize experiences that go beyond demographics and start to employ psychographics and behaviorgraphics – the ability to connect with groups of people by interest and how they interact.

As this practice develops, brands can also gather the intelligence necessary, and widely available, to improve products, services, and spark new waves of tweets gushing with positive sentiment. Doing so over time helps to build the social, and more relevant, business of the future while improving relationships to convert followers into stakeholders.

Brian Solis is the Chief Data Analyst at PeopleBrowsr and ReSearch.ly and author of Engage, the complete guide for businesses to build and measure success in the social web. Follow him on Twitter, @briansolis or read his blog, BrianSolis.com

Connect with Brian Solis on Twitter, LinkedIn, Facebook


___
If you’re looking for a way to FIND answers in social media, consider Engage!: It will help


___
Get The Conversation Prism:


___

101 COMMENTS ON THIS POST To “The Interest Graph on Twitter is Alive: Studying Starbucks Top Followers”

  1. David Perdew says:

    Wow, this opens up a whole new dimension to social media marketing. I can see some laser-targeted campaigns coming out of this data set, if applied correctly. Not only does it help you see what’s being said, but also who’s talking, what they like, and how they feel. I can’t think of a better way to tailor marketing to your business, and reach into the very soul of your audience.

  2. Peter says:

    Hey Brian,

    Thanks for this insightful research. I bet it was fun putting all of it together! I have one question. If students make up 7 percent of your 50,000 sample and they are the largest group, then what is the second largest group?

  3. Peter says:

    Hey Brian,

    Thanks for this insightful research. I bet it was fun putting all of it together! I have one question. If students make up 7 percent of your 50,000 sample and they are the largest group, then what is the second largest group?

  4. Thanks a lot for sharing such an insightful research! Social media marketing has now found a complete new dimension now with this post.

  5. Hi Brian! I thoroughly love this blog article. I’m #53 on the list! That’s pretty fun! I learned a ton of new things about who is creating Starbucks chatter – Well, I didn’t know any of this information before! It seems that the west coast might have a lot of tweets since Starbucks is located right here in Seattle. Thank you for the thorough research and the interesting info.
    ~Melody Overton (@SbuxMel)

  6. liz says:

    It is an interesting case study that was published here. The idea that Twitter and other social sites are helping businesses learn about their customers is not a completely new idea but it was expressed in a more comprehensible and logical matter in this article. People do reveal a lot more about themselves than they realize on these site and that type of information to businesses that are paying attention as Starbucks has, will benefit from this information.

  7. SuperbadIM says:

    That’s an interesting mix of people. It makes sense to look at the followers to gather information. With such difficult topics to analyze (relationships and branding) you came up with some helpful information.

  8. SuperbadIM says:

    That’s an interesting mix of people. It makes sense to look at the followers to gather information. With such difficult topics to analyze (relationships and branding) you came up with some helpful information.

  9. Ricky Yean says:

    Nice information! What are some ways that Starbucks can put this analysis to use?

  10. Nice analysis. I wonder how much of this all is significantly different from e.g. Coca-cola followers, Skittle, Oreo or Red Bull – and would love to know the percentage of tweeps giving away their location

    Currently analysing brand interactivity on Twitter of 10 brands myself, one thing is for sure: it varies highly, from old-fashioned broadcast to chatter

  11. BasIl Farano @freshbas says:

    This is really interesting information. Great story. Shows the true potential of “targeted” messaging and the best way to have influencers “spread” a message.

  12. This is a really good start. I can’t wait to see how we’ll be able to build onto that model and make the data even more granular in 6, 12, 18 months. I love it. Thanks for sharing, Brian. Good stuff.

    Cheers.

    • briansolis says:

      Merci Olivier. Granularity is already here. I could have written an entire book of they psychology and connectedness of Starbucks followers and followers of followers at every level…bound by interests, what moves them, what else they enjoy, the ties that bind…point is, this is how brands need to look at social media in order to invest in long-term value.

    • iconic88 says:

      I wish I knew how to say…Right on Brian!! This is spot on.

      This is a great model for many to use as a guideline. Now they know “who” to probably connect with, the edge will be “how” and “why” to connect with these people. How to engage, how to create and build a legion of fans (like Man United fans ;)), how to deliver value over a long period of time, how be of service when you’re not in service, and why? why are we doing this? what’s our purpose?

      Just like you said mate, identify “what moves them, what else they enjoy, the ties that bind, preferences etc etc”…..

      THIS IS FUN FUN FUN!!

    • I like the idea of continually updating this list and to see how it is changing 6 months to a year out. Some names may easily drop off. I think there is a difference between a person with a massive following tweeting once or twice about Starbucks, and then getting a ton of RTs because of his or her following size, and the person who is a little smaller, consistently tweeting about Starbucks. The large profile seems like it would be volatile on the list. Of course the large profile, if tweeting consistently, would be potentially amazing in its profile. Basically, Brian Solis had me eating out of his hand with this blog post.

  13. NEENZ says:

    Step 2 of my methodology to build a sustainable and engaged community is Listen. Listening is more than just monitoring the discussion about your brand. What Brian has introduced here, the Interest Graph, is digital active listening — the most effective type of listening.

  14. NEENZ says:

    Step 2 of my methodology to build a sustainable and engaged community is Listen. Listening is more than just monitoring the discussion about your brand. What Brian has introduced here, the Interest Graph, is digital active listening — the most effective type of listening.

  15. NEENZ says:

    Step 2 of my methodology to build a sustainable and engaged community is Listen. Listening is more than just monitoring the discussion about your brand. What Brian has introduced here, the Interest Graph, is digital active listening — the most effective type of listening.

  16. Max A. says:

    Thanks for sharing this insight. It’s interesting to learn about the methodology used for analyzing target groups.

    Why did you use a pie chart at the bio-intrest breakdown – isn’t it possible that someone is a student with special intrest who has strong ties to his family?

  17. Max A. says:

    Thanks for sharing this insight. It’s interesting to learn about the methodology used for analyzing target groups.

    Why did you use a pie chart at the bio-intrest breakdown – isn’t it possible that someone is a student with special intrest who has strong ties to his family?

  18. rick wion says:

    Tremendous stuff. Great insights and the infographics make it quickly understandable.

  19. Heather R says:

    Brian – I think it’s really interesting that a snapshot like this can produce meaningful metrics, but how do you model watching it continually, and charging the changes? Twitter moves minute by minute, hour by hour. I imagine this sampling and analysis would take days to do …. It would be more useful to be able to see this daily, not just every 6 months.

  20. Neil says:

    i think chris brogan gets paid from Starbucks to talk and tweet about the company

  21. LoriMoreno says:

    This def. makes me really want a cup of coffee!

  22. Paul says:

    Great. The ability to know a follower to this extent is truly awesome. But its not eavesdropping or intrusive..they have participated in the “brand graph” – they want to build a closer relationship with the brand. We have indeed come a long way in a short space of time..generic marketing broadcasts to engagement through brand graphs.

  23. Thank you for creating this report on our audience. Very interesting to see! Stephen Gillett -CIO, EVP Digital @ Starbucks @stephengillett

  24. JO says:

    Good stuff – are there any free online tools available out there than can help do the work (by scanning bio’s etc)

  25. valkpim says:

    this moment for start…

  26. Raffy Pekson II says:

    All I can say is, “Just Awesome!!!”

  27. I was very glad to find this site on google.I wished to say many thanks
    to you with regard to this wonderful post!! I definitelyloved every
    little bit of it and I’ve you bookmarked to check out new stuff you
    post.

  28. Hi! I just wanna
    give you a huge thumbs up for the informative info you have here on this
    post. I will be coming back to your blog for more soon.

  29. So where is the actual interest graph here? It’s a social graph there, but no actual interest connection :/

  30. Mattocks says:

    Hi Brian – I am interested in how you extracted the Twitter bio information for Starbucks.  I am a business school student at Kellogg and we are trying to perform a similar analysis and I am trying to determine what utilities/tools are necessary to extract, categorize and analyze the data.  Any guidance or help would be amazing.  

    Best,Mattocks

Leave a Reply

Your email address will not be published. Required fields are marked *

Join Our Mailing List

You have Successfully Subscribed!

Stay Connected