Consider: Who do you want to take care of first and in the most boot-lick fashion? The person who personally spends $100 thousand a year with your store or someone whose friends and followers spend a total of $25 million? Is someone with a huge number of Twitter followers, Facebook friends or a popular shopping blog more worthy of the royal treatment than someone who personally spends a lot? Granted, there are shopping carts full of complexities and nuances in analyzing such profiles—generating a meaningful influence rate, if you will—but that's where the fun comes in.
Even worse, beyond those analytical complexities, there's the issue of how to get this data in front of the eyes of store associates and also how to do it in a timely matter. Of course, to start things off, there needs to be a reliable way to identify these influencers as they enter your store.
The practical matter of this issue is that it isn't very practical today. But as retailers get more creative about how to leverage the tons of superb social data out there, ways will be found to access and make use of it. For starters, let's craft the laboratory theoretical data model perfection scenario.
Ideally, the influence would take into account four things: number of followers/friends; amount spent in total by followers/friends; how well the influencer's advice is followed; and the uniqueness of key followers.
Otherwise, Macy's North America might label someone with one million followers a VIP, only to discover that most of those followers are seismic engineers throughout Australia, South Africa and the Middle East.
Even a combination of average and total spend may not be that meaningful unless there is also data about whether a high percentage of those sales are made a by a relatively small number of friends/followers.
Consider the extremes: Do 80 percent of the followers react within two days to any suggestions made by the influencer or do 95 percent of the influencer's suggestions get ignored by 99 percent of the followers?
Influencers who are followed by groups of hard-to-reach prospects (such as older high-net-worth customers, perhaps, or intense fans of a particular type of product) may merit a lot of points.An awful lot of data exists in those bullets that is simply not easily accessed today by retailers. But even if all of those data points were suddenly made available tonight, there is a shameful lack of mechanisms to make them useful to associates.
For this to work, associates need to have a way to identify these influencers as they approach and to be able to understand the type of influencer each is. (Arguably, the simple directive "This person is really important. Go into full boot-lick mode" would probably do a lot of good on its own, but as long as we're dreaming up a retail nirvana, let's do better.)
As for the identification of these influencers, that could be done directly or indirectly. A few years ago, Home Depot experimented with a special RFID-equipped loyalty card that was only given to the very top customers of the stores involved in the trial. The customers knew they were given a special card for discounts and other goodies, but they were not told about the RFID chip—a chip that was identified every time they entered the store. Once they were inside, a bulletin was blasted to all associates that this particular customer had just entered Door #2 and was then near Aisle 7. A short summary of the highlights of his/her purchase history was then displayed.
The ultimate in sneaky, though, would be to use security cameras leveraging facial recognition. (A Carnegie-Mellon University study this month proved that it could work and work quite well.)
There's also the Las Vegas approach, in the "What Happens Online Stays Online" sense. Should this influencer ranking entitle the visitor to special E-Commerce treatment? Differentiated pricing? Higher level instant-chat help? Real-time page tracking and assistance?
What about call centers? Can caller-ID make the backend of this system the easiest yet? No hold for high-influencers? Automatic authorization for refunds, free shipping or anything else, without supervisor permission?
John Bastone, a customer intelligence global product marketing manager for SAS, has been arguing for this type of an influence index for some time and suggests that it's little more than a programmable equivalent of what retailers have been doing for decades.
"We've all long known that word-of-mouth recommendations have a lot more clout in influencing purchase decisions than any kind of marketing message," Bastone said. "All we're trying to [do is add to] what we already intuitively know as marketers."
There's a very cold-hearted—but realistic—way to look at this. It's a fact that all major retailers are going to screw up customer interactions from time to time. One key benefit of an influencer VIP system is to increase the chances that the screwed customers are not the ones with two million Twitter consumer followers.
The very idea in retail—and other verticals—of VIP customers has been tweaked in recent years as more sophisticated analytical systems quickly replaced the highest revenue customers with the most-profitable customers. Wal-Mart, for example, is the highest revenue distributor for a huge number of consumer good manufacturers. But with Bentonville's legendary and well-earned reputation as a ruthless negotiator, the Wal-Mart-related profits are not nearly as glorious.
Banks have discovered that the customers who have the largest deposit amounts demand the most service, attention and perks, whereas much lower income customers expect little, accept long waiting times and pay tons of check charges, interest, penalties and ATM fees.What social networks are doing is delivering a way to quantify the huge buying power of a customer's friends and associates. As writer Carl Reiner once quipped: "You can choose your friends, but you can't choose your relatives. So choose friends with a lot of money."
The influence factor is also based on a lesson so many of us learned in high school. When choosing who to interact with—especially in a non-pleasant manner—the nature of the individual is just one factor. (For example, if making a speech about how the student body's intelligence seems to be plummeting by the hour, citing the school's muscle-bound star quarterback is probably a bad choice.)
But a far greater consideration is the student's associations. It was quickly discovered, for example, that members of the marching band were a very large and tight-knit group. The bandies had a Warsaw Pact perspective that attacking any one of them was an attack on all of them.
That philosophy very closely parallels Facebook friends and Twitter followers. If people think a retailer has ripped them off and they blast a note to that effect to followers, those followers tend to immediately see the rip-off personally as the group dynamics kick in.
(At my school, members of the chess club—and yes, I was a member—were the opposite of the marching band members and were seen as ideal torment targets. If a chess club member was attacked, the other members would most likely scurry away and theorize the most likely scenarios of where the conflict will go and the rough probabilities. It was less Warsaw Pact/NATO and more Keystone Kops with slide rules.)
SAS's Bastone also makes a good point that a true influence yardstick wouldn't even stop with the analysis of a customer's followers and would have to extend—with some reasonable limit—to the followers of the customer's followers.
"The re-tweet or mention of a post is an incredibly valuable data point, as it simultaneously acknowledges that the influencer's post was in fact read and it serves to amplify the reach of that message to people who aren't necessarily following that influencer," Bastone said. "In other words, it is both a Nielsen rating and an indication of the relevance of the message."
Admittedly, very little of this data gathering and analysis—let alone a practical means to deliver that information to associates in a timely and convenient way—is doable today. But the only way it's ever going to be doable is for retailers to think about it today and decide what they would like to eventually see—and to then start changing processes to enable that vision.
The hurdles are mostly logistical and process-oriented, rather than technological. The raw data is freely available today, and the analytics needed are truly not that much more complex than what is currently being used. The biggest hurdle is getting retail execs' minds to internalize a new way of thinking. If you think consumers are resistant to change, take an honest look at your own senior team.