It’s well known that Walmart keeps a massive amount of data on customer purchases. Walmart is also a leader in analyzing that data. Before Hurricane Ivan hit in 2004, intuition might have suggested that Walmart’s Florida stores should stock up on generators and bottled water. But the IT gurus crunched the numbers and discovered that…Pop-Tarts were most likely to sell out. Some of the same techniques used by Walmart have bottom-line benefits for customer-agent interactions.
In recent years, the collective wisdom that’s stored in millions of on-line transactions has been data mined by businesses to derive clusters of customer attributes that match up with better sales opportunities.
Say you call up to complain about your uncomfortable seat on a red-eye flight from San Francisco to New York. The agent knows (from a data mining alert) that this type of customer—seat complainer on long flight—is likely to be interested in vacation packages to Hawaii. The agent then makes a successful up-sell to this weary but delighted customer.
With storage technology capable of holding terabytes of data and plenty of data mining tools available, we’re in a golden age of number crunching. In fact, the title for this blog post was run against a data mining tool used by publishers to predict bestsellers (I was edged out slightly by Zack’s better-selling blog title!).
Sounds like it should be possible to mine contact center interactions…
And it is being done. The key innovation has been to employ speech recognition to search in real-time for trigger words and phrases. Called speech analytics, this technology is capable of mining contact center conversations to learn the real reason or root cause for a customer query.
Scenario: I call customer support to complain about my slow browser. But through analyzing similar conversations across many customers, it turns out that my browser is likely not the problem. The real culprit is my low-level of RAM.
As in the case of the Pop-Tarts, there are sure to be counter-intuitive surprises hidden in these customer-agent conversations that are just waiting to be discovered and successfully exploited. And if you’re a contact center manager who has already found interesting connections using speech analytics or data mining, send me your story.
