My local supermarket has a satisfaction target of 85, and hovers around 80-81. According to research conducted by The Dunvegan Group in 2011, those customers giving an OSAT (or Overall SATisfaction) rating of 10, are no less likely to change allegiance than those giving a rating of 8. This is particularly true in the geographically sensitive supermarket business: I go to this particular supermarket because it is convenient to do so.
What’s more, I often go to the slightly less local Organic store before (or after) to pick up those items that I can’t get from the nearby supermarket, or which I don’t like at the local supermarket. Each time when I reach the register, they ask me “did you manage to find what you were looking for”. The answer is always “no”, though they never follow-up with any action. The organic store doesn’t have a bunch of things that I want, and they’re also 30% more expensive than my local supermarket.
It amazes me that whilst I provide both stores with enough information to understand me as a customer, and tailor the customer experience to increase my loyalty and business, that neither are using their big data to do so.
What big data do supermarkets have at their fingertips?
In most cases I use the same credit card each time I visit, so every item I buy can be tagged against me, including time of day, weekday, item and quantity. They know my name so can infer my sex, and can check this against local records for address information.
They could integrate this with temperature and rain/snow weather information, as well as the precise location where I shopped. In many localities this can be combined with demographic information like average salary, political leaning or joblessness.
In addition, it would be easy to incentivize me to use a store card, so they can track more customer information. By encouraging me to fill in a profile, they could get me to provide even more information like my Twitter or Facebook details, which then allows them to scrape my social media information in real time.
In fact, my local supermarket just changed their pricing strategy to give discounts even if you don’t have a store card, so I don’t use the store card any more. Even before then, they didn’t encourage you to sign into the store card online.
What would this mean for the supermarkets?
The first thing they would be able to do would be to accurately predict what I’m going to buy and when I’m going to buy it. This allows them, especially for high value or perishable items, to balance stock-outs with wastage. I happen to like a particular brand of bread, and it’s often out of stock, and I don’t buy bread at all in that case.
The next thing they could to is to do market basket analysis against what I buy, as compares to what people like me buy. People like you that buy this brand of oil also like this brand of pickles. Now, the supermarket can place the goods next to each other, and offer a small discount if you buy both.
And then they could send me personalised offers, to my home. You have to be very careful with this, as Target found out. People can find data mining very creepy, and what Target do is to mix in personalized advertisements with random items, to make you feel at ease.
What’s more, it would allow store manager Tom (who is a nice guy, as it happens), to be a bit more personal about what brings me satisfaction than by writing “All 10’s Please customer service is very important to use” (sic) on a white board. He could find out via online surveys what he would need to do to make me happier.
As it turns out, I’m easy to please. I’m frustrated by stock-outs of the same items every day, and the lack of open checkouts which mean I have to stand in line. Did they really need big data to figure that out?