So, you’ve decided to bring some analytics hotshots onto your marketing team. What should you do first? Whatever you do, don’t let them near the data!
Not yet. It’s too dangerous. More peril than you could shake a stick at.
Don’t get me wrong. I’m a huge fan of analytics to improve marketing decision making. I spent four years leading a team in CapitalOne’s marketing and analysis group. I have a deep appreciation for organizations that have built competitive advantage on analytics. In fact, one of my “must reads” for marketers is Davenport’s Competing on Analytics (good introductory excerpt here).
But what I have also observed with high-powered analytics is real-world myopia.
You see it in many spots—quant jocks working on Wall Street build models and run analytics with little regard to what’s going on underneath in the real world. Those kinds of antics landed us in the middle of an economic calamity. Fun little read in the WSJ last week on this very topic (and by “fun”, I of course mean horrifying).
There’s a parallel, if less apocalyptic, danger in a marketing environment where information richness grows by the day. As we marketers strive to capitalize on expansive datasets, we too risk real-world myopia. Our teams and agency partners will optimize click-thrus and friend counts and traffic in a way that’s disconnected from how our customers’ attitudes and behaviors actually change.
So, instead of letting those quants near the data, your first move should be to get them into the field with your customers. You can’t do this too quickly. Get them to watch customers using products in your category. Test them by asking what your customers are doing 3 minutes before and 3 minutes after using your product. Get them direct exposure, in real life, to the way customers make decisions and form opinions about your products.
Okay, now you can let them near the data.
But keep the heat on. Push your team to connect assumptions and conclusions to real-world phenomena. Try Avinash Kaushik’s blog as a tonic for real world myopia. Above all, get your quant folks regular, direct exposure to the real world of your customers.
MLC Members: for additional resources, see the Council’s resources on CRM and data collection.
Related posts:
on 27 January 10
Respond
I wanted to let you know that Tom Davenport, Jeanne Harris, and Robert Morison are publishing the follow-up to Competing on Analytics next month. It’s called Analytics at Work: Smarter Decisions, Better Results. The book is a more practical, how-to look at analytics and outlines a five-step model for deploying and succeeding with analytical initiatives.
If you are interested in receiving a review copy before the book publishes next month, please let me know!
on 28 January 10
Respond
Very poignant article in a time of ever increasing automated online analytics being deployed to “learn” what customers like and teasing out “insights” with no real basis in the realities of the actual business quite often. It’s not just the quant jocks (of which I am one) but the Behavioral Targeting tools, online analytics software, and black box next most likely product engines that need to be given a dose of reality as well.
on 28 January 10
Respond
Couldn’t agree more. Most really good customer profiles and predictive models start with insights gained from first-hand, direct knowledge of the customer, not just the data. But just don’t have them watch customers. Have the Analytics People listen to hours of taped customer service phone conversations or weblogs. Have them talk to the people in your organization who deal with the customers every day, accumulating weeks/months/years worth of insight about who the customers are (and aren’t) and what they want. How many fancy analytic exercises end with someone on the “front line” saying “I could have told you that!”
on 28 January 10
Respond
This is so far from reality, it’s not even wrong. Patrick has created a (burning) straw man.
Firstly, the quants WSJ are talking about are very different – and I’d dispute the article’s claims – at the end of the day the crash was about senior management buying “pigs in a poke” – items whose risk and price they couldn’t calculate. Australia has plenty of quants – and we still have an economy. It was just that our banks were better regulated and didn’t buy anything they didn’t understand – unlike a lot of the US banks.
And I’ve been working as a marketing analyst too for years – and I’ve seldom seen a senior analyst who worked without understanding the business process, context and talking to people as close to the front line as possible.
It’s an unfair caricture to say analysts aren’t interested in the front line – as much as a caricture as the senior marketer who spends millions on the basis of a “gut feeling”.
on 31 January 10
Respond
Thanks for the thoughtful comments.
Ken, I’d like to respond to your comment specifically. I should have stated that the antics undertaken by quants helped land us in the economic calamity we are in. To be sure, the causes underlying the meltdown are a many-headed hydra. I agree with your point that lax regulation was also a contributor. You see it in Canada, as well, whose banks are the better off for smarter regulation.
My larger point, though, is that individuals with advanced analytics capabilities often rely too much on finding patterns in numbers, and drawing conclusions as a result, without the grounding of what’s going on in the real world to support those patterns. As I’ve worked with marketers across various industries in the last 10 years, I’ve encountered it over and over again.
In my experience, its not that statisticians and math PhDs aren’t interested in what’s going on with customer behavior. Rather, its that its too easy to stay cooped up in front of SAS, sifting through the datamart to find patterns that could deliver outsize profits (or clicks or friends), without venturing out to form hypotheses based on real world phenoma. Jim’s comment about behavioral targeting tools and automated analytics engines, I believe, supports my point.
Thanks for reading Wide Angle. I look forward to continuing the discussion.