I was at IBM Edge last week and it appeared that one of the big problems companies are dealing with is that suddenly, competition is heating up on how to best understand and respond to customers individually, even though there may be millions of them. The example given was of a three-and-a-half-year-old girl who caused a large grocery chain to rebrand one of its most successful products. This young girl sent in a letter pointing out that the bread looked more like a giraffe than a tiger, and while the store responded thoughtfully, it still resulted in a massive Facebook response. The store now has giraffe bread.
This was a lead-in to the talk on social engagement, and the in-depth example was how social conversations surrounding this conference were spreading. It was a powerful pitch but I caught one important mistake.
Knowing the Customer
This really isn’t about knowing the customer as they exist, but about using analytics to understand the customer as they are being influenced by the information they are seeing, in real time. Customer satisfaction surveys, while still important, simply aren’t ever going to reflect the thoughts of the customer as they currently exist. For instance, in the example above, prior to the Facebook event, surveys would have clearly shown that no one was even interested in a name change for the product. That fact changed in the matter of a few hours. Were this a product satisfaction issue, by the time the company became aware of the problem through surveys, it would likely already know it was in a financial crisis due to lost sales.
When a problem goes viral, the firm may have minutes to craft and execute a response before the problem becomes a catastrophe. What analytics surrounding social network provides, if done correctly, is a real-time showcase of what is going on in terms of perceptions surrounding the company’s brand and products.
The IBM presentation was compelling in that they could show graphically and with great clarity what the opinions were, how the opinions were being generated, and where they were being generated. We could see trend lines and activity in real time; both are critical to figuring out what is needed to either accelerate or mitigate the trend. A marketing manager could, with this information, know almost to the second just what was going on with the brand he or she was responsible for protecting and the solution would certainly mitigate significantly the likelihood of being blindsided by a problem. But IBM left out what is perhaps the most important part of all of this.
I could pick any disaster for this example, but I live in California and here we sweat earthquakes. A number of technologies are being worked on that are supposed to give several minutes of warning. But they could actually make loss of life numbers increase because they could cause premature panic. It turns out that if you don’t know what to do, a warning might actually create a bigger problem because the panic could remove you from a place that is safe to a place that isn’t, or result in a stampede that could result in additional injuries or death. For some screwy reason, we don’t do regular earthquake drills anymore. That means folks generally have no clue what to do except panic. What is needed is not just the alert but a detailed script that provides options and a clear path to safety in order to assure life is protected.
Watson to the Rescue
When you have a PR disaster, you often have minutes to react. Let’s take the Google issue of a few weeks back, for example, when the Google CTO, upon finding that there weren’t enough people in his audience for his satisfaction, screamed “I AM GOOGLE” and stormed out. The right thing would have been to immediately call and apologize, provide a replacement speaker, and maybe have someone do some off-the-record bonding with the event organizer about jerky executives. But Google instead stonewalled. This event went viral to the point where there was an I AM GOOGLE hashtag and a lot of damage to Google’s image, right at a time when several governments are thinking that maybe Google should be regulated.
You see, what you need is not only a heads-up about a PR problem, but real-time advice on what to do about it. Otherwise, folks are likely to either hide under their desks, as Google’s people did, or do something that actually makes the issue worse.
Here, IBM’s Watson could provide part of this solution that few other options could; it could showcase possible responses along with probabilities of their effectiveness based on past events, in real time. This would give the response team both the knowledge of the problem and the ability to make the best decision about how to respond to it in real time.
This was the part IBM left out and, strangely, it is the part it actually has the best solution for, if it were applied.
Wrapping Up: Changing Focus
Years ago, when I worked in IBM Storage, one of the running jokes was that everyone was focused on backup time and no one was focused on recovery. People would buy products that did incredibly quick backups, but when a disaster struck (which was fortunately rare), the recovery of the data, the entire reason for buying the backup product in the first place, wasn’t assured. When it comes to analytics, knowing the information that defines a problem is important but the truly important part is knowing what to do about it. Knowledge without intelligence often not only isn’t helpful, it can actually make things worse.
IBM Edge showcased, albeit accidentally, that once again, the industry is focused on the wrong part of the problem, the analytics, when it should be focused on the quality of the decisions that result.