Bridging the Uncanny Valley: How to Efficiently Humanize Data-Driven CRM

0
133

Share on LinkedIn

In the CRM business, knowing your customer has always been the Cardinal Rule for success. The advent of Big Data has accelerated the process, allowing us to amassing hordes of data about what, when and how much customers buy, errors they encounter, pages they visit or avoid, and which social networks and other sites they frequent. This intimate knowledge of our customers—even beyond their behaviors on our own site and interactions with our own product—has become a tremendous asset, not only for partnership and monetization purposes, but also for the value it provides in improving customer relationships.

Much like a personal shopper, many companies now tap into this data to provide personalized recommendations and individualized—sometimes proactive—service, actually offering assistance before the customer even asks, based on this extensive, real-time data.

It’s a model we all know quite well from the bricks-and-mortar world: who doesn’t love it when the server at your favorite restaurant remembers your “usual” without having to ask? Or, the rock-star feeling of walking into the corner coffee shop and having your grande, half-caf, nonfat, mocha waiting for you at the checkout counter? Just like Norm from “Cheers,” we all want to go where everybody knows our name, and treats us like a regular, valued customer.

But, what happens when that level of intimate knowledge crosses the line?

More Human than a Human

The term uncanny valley describes a phenomenon in which near-human appearance and movement in non-human subjects cause real humans to feel uncomfortable or even repulsed. For example, rather than feel familiar and humanistic, robots and 3D computer animation that so closely resemble natural human beings can actually have the opposite effect, causing us to instead feel strong revulsion toward these androids that seem almost “too” human.

This uncanny valley also exists in the data-driven world of online commerce and customer support, as detailed data about customer behavior and history is made available to agents, it almost feels as if they know too much. Of course, we want to leverage as much detail as possible to provide better service—to solve problems more quickly and be proactive with tips and recommendations. But, at some point, it can cross a line and make the customer feel a little like Big Brother is watching and, well, a little creeped out that the agent knows so much. Privacy becomes a concern.

Most Internet users are well aware that throngs of data are collected about them every time they hop online. And, that data is—for the most part—anonymous: cookies routinely track the sites they visit to serve up relevant ads and other content, but none of their personal identification information is attached to that data. Google’s Gmail service stepped into the slippery slope of the digital uncanny valley when it began serving up ads based on the content of users’ emails. Some users were repulsed and closed their accounts. Others accepted the algorithmic snooping as the price to pay for a free service and continued. In Google’s case, it was clear that a computer was doing the snooping—a non-judgmental, non-feeling machine that doesn’t care what sites you surf or what you send or receive via email. But, when a human has access to that kind of data, the tables quickly turn.

Bridging the Gap

The key is to find a happy medium—a way to make customers feel like a VIP who is privileged to exceptional attention because they’re a valued customer, but not like they’re entering a naked body scanner every time they use your app or call customer support.

The solution is a combination of technology, attitude and training. When it comes to customer data, with great power comes great responsibility. Using technology that enables you to mine that data, surface relevant details and provide exceptional support is a valuable opportunity. It can come as a pleasant surprise to customers who enjoy getting the red carpet treatment, which goes a long way toward increasing sales, fostering loyalty and encouraging word-of-mouth.

But, companies must be sensitive to the level of data they allow agents to access, and they must train agents to use the data judiciously, on an as-needed basis, and never as a way to show off their intimate knowledge.

The fine line between being creepy and helpful varies from business to business and the type of data in question. For example, a website hosting provider that proactively corrects a technical issue for a customer is helpful. But companies that handle more sensitive data—like financial services, health care and other secure entities—need to be more reserved in their use of customer data. Each company must determine where the line is for their unique business and tailor a plan that is beneficial for both them and their customers.

Making it clear that computers—not people—are collecting all of this data can also help to ease customers’ concerns over privacy and the creepy factor. If we’ve learned anything from the recent NSA reports, it’s that people can’t be trusted. But, when it’s an algorithm that’s surfacing a limited amount of data to a support agent, the process becomes less intrusive. After all, computers can’t judge our purchase and other behaviors the way some may feel a support agent would.

As the level of data available to agents increases over time, training agents to interact with customers in a helpful, yet sensitive, way will become even more important. Growing privacy concerns and the uncanny valley are forcing companies to strike a balance of using that data judiciously to solve problems quickly, without making the customer feel as though they’re being watched, scrutinized and exposed. The right technology and proper training can help any data-driven CRM effort go from creepy to caring.

Christopher Gooley
Christopher Gooley, CEO and co-founder of Preact, has architected, developed and launched multiple B2B and B2C online services. Christopher builds data-driven tools to help companies better understand their customer's actions to enable support and account management teams to proactively solve problems, increase engagement and multiply per customer revenue.

ADD YOUR COMMENT

Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here