Using Big Data to Improve Retail Marketing
This morning, I stumbled on a great presentation on how Target uses Big Data for personalized marketing. The presenter, Andrew Pole, runs Target’s Guest Data and Analytical Services group. He describes how Target brings disparate data sources to build out a “guest portrait” for each customer and walks through several examples of how this data is used to drive personalized marketing actions. If you are at all interested in data-driven retail marketing, I’d encourage you to check it out (by the way, Mr. Pole’s work was described in a recent New York Times Magazine article on “How Companies Learn Your Secrets”. The article has since been widely-discussed and become quite controversial).
I enjoyed the presentation for many reasons.
- Their approach to working with customer data and using it to drive marketing actions has a lot of similarities to how we (CQuotient) do things.
- While there is much talk out there on the promise of Big Data, there’s not much on the specific actions taken by retailers with the help of Big Data. Mr. Pole’s presentation is a welcome exception.
- Mr. Pole keeps jargon to a minimum and delivers a very clear and pragmatic account of how it is done at Target.
It is clear from the presentation that Mr. Pole and his group are delivering significant business value to Target. Any retailer would benefit from having such a group.
But what might it cost?
From the presentation, there are 50 people in the group, 25 in the US and 25 in India. My guess is that there are a fair number of statisticians and data scientists in the group. A back-of-the-envelope estimate suggests a $4m annual salary for just the US-based staff (fully-loaded $150,000 * 25 = $3.75m). Even ignoring the other costs – the cost of the offshore team, database platforms, tools such as SAS and Unica – this $4m annual expense for an analytics group is something that only a very small number of retailers can afford.
Apart from the cost, there’s also the issue of how to attract and retain the right talent. To be perfectly candid, the kinds of people you want – data scientists, for one – aren’t likely to want to work at a retailer. Further, the ability to ask the right questions of the data, bring the right algorithms to bear, ensure the validity of the results etc. requires a level of “analytic seasoning” that’s very hard to find. Hiring freshly-minted data scientists isn’t going to give you that.
So what should all but the very biggest retailers to do? Watch the likes of Amazon, Wal-Mart, and Target in envy and do nothing?
There’s a way.
Emerging customer-analytic platforms (like CQuotient’s) and supporting services can help retailers leverage Big Data for marketing better, faster and cheaper than what an internal Target-like group can do.
We have cracked the code on teasing out detailed product and other tastes at the individual consumer level and are working with a select handful of retailers on putting this to use. Happily, early results are very strong.
So … dear retailer, worry not, there’s a way :-)