Published in The Holmes Report October 26, 2014
We recently held our first Social Intelligence Summit in EMEA at London’s City Hall. This will be the first of a series of global summits that we hold in the future, all relating to the spirit of our annual PreCommerce Summit that is held right before SXSW.
Paul Holmes asked if I would write a column based on the presentation, which has appeared in The Holmes Report. I thought I would also share it with our readers of Common Sense. All the best, Bob
Column — The Rising Importance of Small Data
The real shift that matters in big data, for communicators, is the shift to small data excellence. Big data is actually clumsy. It’s like being at a cocktail party where you can see hundreds of people, but you can’t hear well enough to speak with any of them.
Meanwhile, our brains are actually sorting through the avalanche of online data for us in ways that we don’t quite realize, since it is subconscious. We repeat behaviors, ask common questions and seek out a finite set of content, despite our belief that we are “wide-open” to change. We put limits on ourselves, based on what we see our peers doing, to make life easier for us.
If we know the patterns of the crowd, we can understand where to focus.
This leads to a world where we are acting more like detectives to find behavioral clues that unlock insights that really matter. It’s what we call “forensic analytics”.
Let’s think about our world. For any brand or topic, there are no more than 50 influencers who drive the majority of share of conversation in a given country. There are approximately 15 keywords that lead most people to the content they desire via search. We can’t think of more than 200-400 questions, on average about any major topic. In a crisis, we ask less than 80 questions over and over.
We think in finite terms.
Because we are so predictable, the next generation of analytics will be centered on “which filter” we need to get the view that is relevant to our needs. A typical brand may have 30-40 filters. It’s like having a camera with the most amazing set of lenses you’ve ever seen.
Patterns lead to clues and clues lead to actionable insights.
Now, let’s go crazy and imagine that an NGO decided that it didn’t like a specific company’s actions. You can now identify every member of that NGO online, segment them by what they care about and use that as a filter vs. the world’s media and know exactly what they are saying and doing online down to the nth detail. The brands that understand this are now becoming smarter than the market. My hope is that it leads to better dialogue between adversaries, since knowing each other’s viewpoints often leads to breakthroughs.
In the old days, meaning 2012 and 2013, we would create brand taxonomies. In 2014, we create market-driven taxonomies. We no longer wing it. Today, the conversation and the behaviors of the market dictate what the architecture of a taxonomy actually consists of, leading us to a place where we can see what cardiologists are saying in Dallas about heart surgery vs. cardiologists in Chicago or in the UK.
The right taxonomy acts as as a truth filter for today’s listening platform.
If we take it a step further, we know that a relatively small set of customers define a given market. For example, there are 300-400k resellers in technology. There are 600k physicians in the U.S. of which 60k set the tone. There are 2,000 CIOs who drive the IT industry. This is a small data challenge and, today, you can identify exactly who these folks are, place them in a custom search engine to track what they say, who they are influenced by, what time of day they are most active and what type of media they prefer and greatly improve how you spend your time in earned, owned, shared and paid media.
Our ability to optimize resources for the entire company has never been higher.
The ability to understand small data is also leading to a shift from paid to smart media. Earned media is really our customers consuming, talking and sharing. If we are looking at their conversations and actions correctly, ESO (earned, shared and owned) will define what we should do for paid media.
A truly customer-driven media strategy lets ESO define the P.
It’s time for earned media experts to rise to the challenge and redefine how we build media planning models. For example, we have done a pilot that we call, TinderBox, looking at all content consumed in Texas related to politics. We index about 6MM articles every time we refresh and figure out which content matters at the town level. It turns out that the DMA system we all know and which paid media is based on is off its game. We can see the 130 towns in Texas that matter by topic and influencer. In the future, we want to know which “25 towns” matter for immigration reform vs. “20” towns for medicare. We won’t blindly put advertising against population alone.
The DMA structure and how we view word of mouth is about to change.
These trends are part of a larger shift in communications and marketing that we refer to as Storytizing. In the future, we won’t be starting with the idea that we’ll advertise to a market. Our customers will show us the blueprint for action via their conversation and behavior and then we’ll empower them to share our story and amplify it with earned, shared, owned and paid media ideas.
We won’t advertise and we won’t count impressions. We won’t care which part of PESO we use. All we’ll care about is the market’s desire to share our brand’s story.