December 1, 2017
When it comes to data analysis in online marketing, I sometimes get the feeling that we go a bit too far. Call it data-trigger-happy; but, in one way or another, we have come to the point where we feel the need to work through the entirety of data from campaigns, channels, and reports in order to have a complete picture of our audience.
This means we listen to all discussions surrounding certain sentences, monitor keywords or brands across all possible channels. We read articles in the press, forums, blogs, social media, etc. And what does this large mass of data teach us? Of course what the general trends are, but not much more. No viral or strong campaign has ever arisen from basing strategy off general trends. After all, you only bring a slight variation to the existing major trends. From ‘data to insight’ is therefore not an analysis of the mass, but of the specific influential groups.
First, let’s see why data analysis has become so important in the online marketing world. Overall, this is a result of our short attention spans. We scroll like crazy through Facebook, Instagram, Twitter, Snapchat or search results and not much information is actually absorbed. This is expected considering the abundance of channels and content that comes our way.
The only way these platforms deal with this dilemma to by developing a thick information filter. Based on your previously ‘used’ information and behaviour, algorithms determine which content is presented to you. The biggest issue for today’s online marketers is learning how to reach your target market and capture their attention despite these filters. What’s the solution? Data. But if you’re not careful, data can move you further along the path of irrelevance.
Social media filter feeds or search engine results, which also have different values, often provide tunnel vision. A good example of this is the discussion surrounding the role algorithms played in provision of information during the US presidential elections. Another example is Juan Buis’ (an editor of the Next Web) argument about Spotify.
There is a drawback to this fully automated curation, for example, on Spotify. The algorithm is based on the mass and knows it too well, meaning you will never be surprised or continue to struggle to find songs and albums that are similar. Once you belong to a certain group, you will be targeted as that group and algorithms will ensure you are pushed deeper and deeper into that filter bubble until there is almost no questions or surprises. And that is the agreement for data-based campaigns of this magnitude: too much knowledge about too large of a group.
It’s tricky to analyze too much data from a gigantic target group. The larger the group you are analyzing, the greater the chance you’ll end up with a predictable and generic segmentation. Even further, you can miss an important opportunity to truly surprise the target group.
The result often consists of fairly general insight into trends. Something you can often conclude on your own. To stay with the metaphor of Juan Buis, if you like Beyoncé, then chances are that Rihanna’s songs will also appeal to you.
You get stronger results and original insights more quickly by focusing on a small influential part of the whole group. Data driven campaigns revolve around identifying the right audience. The first question to ask yourself as a marketer: who can I reach first to best leverage influence over the masses?
The 1:9:90 model is a strategic approach to segment the audience of a campaign.
The ‘top’ 1% of your audience are influencers who develop content themselves, shape the market and drive conversations about a topic. When they talk, the rest listen. Historically, this group mainly consisted of traditional media. But as Marketingfacts Special makes abundantly clear, influencer marketing groups now consist of journalists, bloggers, entrepreneurs, experts and almost every profile with a large online audience.
This group acts as curators, providing a lot of content with context that makes its way through the reports of influencers. Generally, they share their own opinions. This group makes recommendations, shares relevant content, subscribes to newsletters, podcasts, and other series, posts comments and updates peers with new knowledge.
The mass reads, listens, and watches until they weigh in. They are actually quite satisfied with this reasonably ‘passive’ attitude in which they read content that appeals to them. This group decides how convincing the 1 and 9 percent actually are when it comes to telling a brand story.
The advantage of this is that you can then start working with data from a smaller group. This prevents you from having to dig through a lot of data of mass size. And thus, you can make decisions more quickly regarding messaging, channels, etc.
A good example comes from music, even though I know that real creative musicians do not get pushed onto stages. Imagine you want to climb the charts in the coming months. You can start by looking at what is now popular in the top 40, make a comparable song and hope that your song becomes a hit. I think we all understand this can only be achieved through a very large marketing budget that ‘buys’ success and attention. Sounds familiar when you consider the mechanism of social advertising, right?
An alternative to this approach is the 1:9:90 model. What groups influence the charts? DJs, festival programmers and editors of services like Spotify can be influenced by experts and active communities. You can get usable data from there. In a recent interview, 3FM producer Vera Siemons herself indicates how she discovers new music:
“I am still a fervent music blog visitor. For example, I find the interface of Indieshuffle very pleasant, so I keep visiting that site weekly. In addition, I download a list of new tracks from other music blogs. I find last.fm and her statistics very nice to keep track of.”
Data becomes useful insights when you segment to better align your choices. If you want to surprise music connoisseurs, DJs, or the public, look for ways in which your numbers stand out for the 1 and 9 percent.
For example, for marketing and communication campaigns, this means that you can focus on analysts, CIOs and CMOs of Fortune 1000 companies or consumers who are interested in certain technical gadgets. Or maybe you focus on HR professionals who want to gain insight into new IT tools in the field of work or diversity.
By sticking to the influential 1 or 9 percent within the large target group, you arrive at the insights that matter. Which gadgets are being picked up slowly? Do we see a change in profiles that follow CIOs on LinkedIn? Which podcasts are recorded, listened to and shared by HR professionals? With that knowledge you build up your storyline, make choices about the channels where you want to reach your target group, and you have the knowledge to find the balance between paid, earned and owned media.
It’s a way of avoiding the dozens of general, predictable marketing choices and leaning into specific influential groups.
Need help with data analysis and gleaning insights? We can help.