Skip to content

How to Master Customer Experience Optimization: An Interview with Dean Shaw (3 of 3)

Share this article: Facebooktwitterredditlinkedinmail

Throughout our interview series with Dean Shaw, Global Program Manager at SAS, we’ve covered a diversity of topics including live chat customer engagement and live chat lead generationIn the third and final installment of our series, we discuss customer experience optimization and the importance of focusing on ‘Little Data’ before going big.

RB.AI: Are there common pitfalls that you’ve noticed in your experience with live chat implementation, or omnichannel engagement?

DS: Our omnichannel capabilities are a challenge as we have disparate systems that don’t speak to each other. Mingling with others in our profession has told me that this is pretty common as many companies don’t have a single all-encompassing solution that provides chat, social, phone, email, and lead tracking etc. So, like us, they do the best they can.

As for chat, it isn’t rocket science as far as implementation. Just a few lines a javascript and away you go. The key is a commitment to the program and ongoing training, optimization, and monitoring. The pitfalls are usually an uninspired approach to chat where the code is thrown up but then it’s not properly staffed, live chat agents aren’t trained, and unrealistic, misguided goals are set. The other more obvious pitfall is having no chat program at all. Based on our experience, it’s baffling to me that all larger companies don’t have it.  As shown above, its about half of our inbound engagement volume, while phone hovers around 10%. The sands have shifted and the 1-800 number approach to customer support is dying. Email is not much better -from a customer perspective the message goes into a black hole somewhere. Chat and social offer immediacy, which is ground zero for a customer experience optimization. Regardless of channel, response time is critical.

Across channels, however, our response performance levels are pretty incredible. Regardless of what channel the visitor chooses, we want that experience to be as positive as possible.

Chat: 3s

Email: 57m

Phone: 17s

Social: 2h 8 m

If you put a customer on hold for 17 minutes you start out on a bad foot no matter how well the call goes. Contact deflection only works if self-service is simple and clear – phone trees need to be phone shrubs. Deflection solely to minimize support costs is short-term thinking that won’t get you anywhere.  Customer experience optimization is everything from customer service and support to site and retail experience.

RB.AI: In what ways have you used big data to understand a customer base and communicate more effectively? What opportunities does data analytics present now to brands that weren’t possible five years ago?

DS: Big Data offers a lot of great opportunities to drive customer experience and by extension revenue. But so many of us are still so bad at the little things, and we need to focus on these issues before going big. Customer journey, bounce rates, conversion rates, funnel analysis, A/B testing -these are all simple, little data pursuits that we routinely get wrong on a basic level. As a guy at a business analytics company that offers very robust solutions, I’ll admit, Big Data can have enormous value for understanding the customer experience, but it’s also very complex. As mentioned above, the absence of consolidated systems that speak to each other seamlessly in near real-time is a huge challenge that Big Data almost requires as a starting point. So my advice to organizations is, yes, pursue Big Data initiatives that include AI, IoT, personalization, and all those other buzzwords, but make sure you are doing the little things right. When I say little things, I mean offer easy availability – like chat – to hold the customer’s hand. Be the customer. Use transcript analysis to find weaknesses in the customer journey, and then by all means, fix those things.

Here’s a recent example of what not to do from a company through which I recently purchased cologne: FragranceNet. I bought one item and was subjected to three emails daily ad infinitum.

Interestingly, they have chat support on their site but I have no clue where it is.

My point is there are a lot more FragranceNets out there than companies that have it together enough to engage in Big Data initiatives. We are making rookie mistakes with very simple things. Whether its Big Data or small data we need to focus on the customer and do the simple things right. Make sure you’re really good at Little Data before succumbing to the lure of Big Data.

Dean Shaw, Global Chat Program Manager, SAS

As global chat program manager for SAS, Dean leads SAS’ Chat Program that, in addition to maintaining ridiculously high CSAT scores, also acts as a lead generation and chat analytics machine. Dean began his digital marketing story with at a time when Herman Miller chairs and foosball tables defined you as viable venture capital target. He continued on to the mobile industry with Motricity when the StarTAC was the coolest mobile device on the planet. Following that, acted as a digital analyst at SAS capturing the impact of web, mobile, and social media on the digital experience. He also measures everything as though his life depended on it.

Follow us:Facebooktwitterlinkedinrssyoutube
Meredith Lackey

About Meredith Lackey

Meredith Lackey is a content producer with expertise in computer networks and systems. She received her B.A. in Philosophy from Hampshire College, and her M.F.A in Moving Image from the University of Illinois Chicago. Her output spans long form writing, B2B and B2C content, and branded video content. Meredith has worked with Tony on creative projects and at Boopsie since 2014. She loves gritty travel experiences where she can learn about diverse people and new trends.

Scroll To Top