GUEST POST: The Problem with Data-Driven Top 10’s
We seem to get more social networks (Pinterest/Instagram), but are we getting better data? I say no! Are these tools giving business the feedback or the prioritization we need? I say no!
With data I ask what action can I take. Counts alone are not useful. This Four Step Feedback Model from Wired culminates in action. With no action, there can be no reward.
Evidence, Relevance, Consequence & Action is a great mindset to approach your analysis.
Metrics always evoke emotional reactions. Chris Brogan’s post “Stats I Want from Twitter and Other Services” is proof. Chris suggests a Business Intelligence wish-list for Social Networks:
- Top 10 people talking about me.
- Top 10 people talking about my product (or search term).
- Top 10 people responding and engaging with me.
- Top 10 people I talk about.
Chris is not wrong. Data driven Top 10’s are great, but they don’t scrutinize what’s happening inside your data. Chris is right, being able to measure specific behaviours is essential, but is this enough? With a mass of data, we need to be able to visualize and segment the flow of data so we can act appropriately. We need easy access to the movers and shakers!
In my role as co-founder at Listly, I may get 1k unique visitors each day to Listly, but there’s flow – people churn. Who’s new on and influential. Who’s sticky that I have not met? How can I build community around their interest and time investment. I want tools to let me focus and prioritize. Attention has a half-life! I need to act fast.
The problem with Top 10’s is they change slower than people’s behaviour. We don’t just need Top 10’s, we need to track flows in data. We need to see churn! Customers are churning but you can’t see it. New people arrive, some people stay, some people leave. It’s all about people. If you can act on churn data, you can influence behaviour – eg recover a lost customer or welcome a newbie. I’m talking about humanizing your brand with the aid of technology.
Behaviour Flow is highly actionable. This 5 part segmentation lets you surface the movers and shakers for any type of behaviour.
I think of this as the next step in Cohort Analysis. McKinsey call it a MECE segmentation.
- Mutually Exclusive – Everyone is included in one bucket.
- Completely Exhaustive – Everyone only gets counted once.
I learned this trick while consulting alongside McKinsey at Safeway. I also used this extensively while running the Analytic Apps team at Business Objects. Waterfall Charts are great for explaining change. See: “Say it with Charts” via Zelazny. If you have ever be in on a McKinsey engagement you will have seen them.
Here’s some template questions that begin to expose the value of this approach. Imagine you own “X”
- New – Who’s doing “X” for the first time
- More – Who’s doing more of “X”
- Same – Who’s doing “X” about the same
- Less – Who’s doing less “X”
- Absent – Who’s stopped doing “X”
This technique is universally applicable to Business, but how about Social Media?
- Who’s spoken to me for the first time (new)
- Who’s focusing on me and getting my attention (more)
- Who are my regulars (same)
- Who’s fading away (less)
- Who’s missing (absent)
I’d be pumped to have these insights into my Social Media engagement. It’s very actionable. I can approach people in each segment with the same message. That makes it scalable. How about you? What would you like to see?
As a serial entrepreneur, there’s nothing Nick loves more than a new challenge. Nick’s current challenge as co-founder of List.ly, is to put Social Lists on the world’s to-do list! When Nick is not being social or blogging, he’s playing board games, running, wakeboarding or snowboarding. Nick also publishes an award winning board game.