A detailed look at how to mine and then leverage deep and raw social data about potential influencers.
Since Google has officially declared war on linking, brands and agencies are turning to alternative ways to drive organic traffic, reach, and visibility. One of the tactics of choice today is Influencer Outreach.
For those among you who are unfamiliar with Influencer Outreach, it's a fairly straightforward process. You identify people inside the social networks that are influential in your vertical/category, connect with them, and get them to engage with your brand in their own channels where they have a sizable following. This could be accomplished by sending them a product sample which they would then review or by providing those with some other form of valuable content like downloads, coupon codes, or invites. While some of the "influence" happens organically, some sort of value exchange is often being provided.
Social media is one of the greatest, most open and honest channels out there. People openly discuss their likes, dislikes, favorites, and goals, and some might become influencers within their own spheres. Simple numerical key performance indicators (KPIs) such as followers and fans identify their influence, and metrics such as likes, favorites, shares, and retweets highlight their connection and relevance to a specific topic or product. Social media can give us true insights into the behavior of the world, and while we had this with search data on a macro level, social media data allows us to create these insights on a one-on-one basis.
With the fast-changing and constantly evolving social ecosytem, I have been spending a lot of time testing and evaluating some of the influencer identification tools out there. To be honest - and given the large amount of social data available - I have been very unimpressed with the functionality and flexibility of these tools. Most of them simply take the output from the social media platforms and apply some simple sorting like most tweeted, liked, etc. They do not offer any form of advanced modeling, rule-based engines, advanced targeting functionality, or integration with paid social channels.
Today I would like to share with you my approach to identifying influencers, measuring their influence, and some of the insights it can drive.
Anybody who has read anything I have ever written knows that I tend to solve things through a very data-centric lens. This is no different with social media. So here are some tips to get you started quickly finding your brand's influencers and drawing out the most insights possible from social media data.
Let's assume we are the social media agency or department for an upcoming TV show in the horror/zombie genre scheduled to launch this fall. We need to start identifying our potential influencers and create engagement strategies around them so that we can activate them when needed. For this hypothetical exercise we have identified The Walking Dead as our top competitor.
In order to identify the influencers in this space, we need to collect as much information about them as possible. This can be done in a few different ways. If you have development manpower to build API adapters and custom databases, I would recommend connecting directly to the social media platforms. Most of them provide very easy-to-integrate APIs that allow you to extract a wealth of information. If you don't have the luxury of a development team, we will provide an alternative solution that allows you to get that rich data without getting too "geeky."
Let's assume we created an audience survey and identified that the highest activity around this genre is happening on Facebook and Twitter. (Both Twitter and Facebook return the data in JSON format - to learn more about JSON I recommend checking out its Wikipedia page.)
Here is a quick way to get started on both of those data sources:
Facebook allows access to their data via the Graph API. While the Facebook Graph API deserves an article on its own, simply put, it allows you to query Facebook's database and extract large amounts of valuable data about anything Facebook. If you are interested in learning more about the Facebook API, they provide a nice getting started guide in their developer section.
So as an example, in order to extract all the activity on The Walking Dead Facebook page, you would call up the following URL:
As you can see, the query gives us VERY granular information about all the activity on The Walking Dead Facebook page. Aside from the actual conversations, it highlights who is sharing, commenting, etc. This data can be extracted and analyzed in anything from Excel to mode advanced analytics tools like Tableau or Spotfire.
Twitter is just as simple, but yields larger sets of information about the conversations themselves. The simplest way to get data from Twitter is to use the search API for conversation data and the user API to get information around the users.
As you can see in the example responses, the URLs provide a wealth of information regarding the topic, the content, the participants, the location, interactions, and truthfully, more than I could list here.
One other option I would recommend, especially if you don't have the tech resources, is the use of a social data aggregator such as DataSift. DataSift allows you to easily identify a source such as Twitter, Tumblr, or Facebook, create a filter (search term, age, page, etc.), and start recording the data stream immediately. Then DataSift will monitor all the conversations that match your filters, record its results, and export that rich data into your format of choice.
Let's get back to the original mission:
The next section shows you some dashboards and apps I have built on top of the social data we collected above. (The creation of dashboards and models is not part of this article, but if you reach out via Twitter I will be more than happy to share some samples and ideas.)
While all sources mentioned above provide the complete name of each user, I shortened them to four characters and appended a ... in order to preserve privacy. However, the data returned from these APIs contains complete names.
This is the screen the NSA is jealous of and where things get pretty deep. It is an interactive dashboard that allows you to select an influencer that engaged with The Walking Dead on Twitter. Once you select that person (top left), the other screens will show the topics this person engaged with (top right), the hours this user engages with the show (bottom left), the days (bottom middle), and the devices and platforms used to communicate (bottom right).
Now that we identified who we want to talk to, we need to work on what to say to them. This screen interacts with the first one and allows you to dig a little bit deeper into the relationships between the users and the individual content pieces. Once you have selected a group of influencers (the target), this screen will show you the most liked and shared topics by these users (top left); you'll also see what type of content works best for that sub group (top right).
Our quest to win over our influencers goes a step further by analyzing engagement patterns. This analysis shows you any outliers in terms of behavior. For example, it appears that there are some topics that get relationally many more comments than likes (bottom right). Again, this will give you a deeper insight into what type of content your target influencer group likes to share and engage with. This is an important distinction; as an example, you can easily get two rival groups (Android vs. iPhone fans) to generate an endless string of comments, but with just comments and no shares you will not increase your reach.
OK, now that we know what to say to our target influencers, when do we say it? The screens below will help you to identify historic trends (top left), activity by hour (bottom left) and the same view broken out by gender (right). These types of insights will make sure you are putting your message in front of the right people at the right time (and makes data lovers like me very happy).
As you can see, there are many ways to mine and then leverage this type of deep and raw social data. I hope this article sparked some ideas and I would recommend you monitor this space and data available closely; with more and more connected devices, I am sure the number of data points will dramatically increase over the next years. Please feel free to reach me via Twitter with any questions you might have.
On the heels of a fantastic event in New York City, ClickZ Live is taking the fun and learning to Toronto, June 23-25. With over 15 years' experience delivering industry-leading events, ClickZ Live offers an action-packed, educationally-focused agenda covering all aspects of digital marketing. Register today!
Benjamin Spiegel is the managing partner, vice president of innovation at Catalyst, a GroupM agency. He is a digital advertising veteran with more than 14 years of experience in media, data, and technology. For the past three years, he led the search practice across the GroupM Network; today, Ben leads the innovation practice at Catalyst providing thought leadership and innovations for its Fortune 500 client brands.
Hong Kong, May 5-6, 2015
Gartner Magic Quadrant for Digital Commerce
This Magic Quadrant examines leading digital commerce platforms that enable organizations to build digital commerce sites. These commerce platforms facilitate purchasing transactions over the Web, and support the creation and continuing development of an online relationship with a consumer.
Paid Search in the Mobile Era
Google reports that paid search ads are currently driving 40+ million calls per month. Cost per click is increasing, paid search budgets are growing, and mobile continues to dominate. It's time to revamp old search strategies, reimagine stale best practices, and add new layers data to your analytics.
May 6, 2015
12:00pm ET/9:00am PT