In the user experience, what data matters to a business? How can organizations meaningfully measure and improve it, end to end?
Riverbed was a sponsor for Simpler Media Group’s Digital Experience Summit summer event, which took place online on August 17-18. SMG met with Jon Hodgson, VP of product at Riverbed Technology, to explore questions like these, inspired by his session titled, “Managing Customer Experience in an Omnichannel World.”
Hodgson started as a systems administrator and then evolved into sales engineering before embarking on his current multi-disciplinary perspective as VP of product. He spent many years solving problems for businesses and customers and eventually came to realize that none of the behind-the-scenes stuff showed what the human encountered.
“What I was focusing on was only a piece of the problem — the clicks that were failing, the interactions that didn’t work — and only a secondary piece," Hodgson said. "What did someone do? Did it work? How well did it work? And, why? Those questions should always be the starting point.”
Ultimately he came to see that user experience monitoring and back-end monitoring are really two sides of the same coin, and when we can look at them in a unified way, we can begin to solve the real underlying problems.
Simpler Media Group caught up with Hodgson for a Q&A:
Wrong Dataset, Wrong ExperienceSimpler Media Group: In what way do you feel your experience gives you unique insight on the omnichannel world?
Jon Hodgson: What’s commonly missed is the ability to differentiate between a symptom and a root cause, and that’s often a problem of not having the right dataset. To ensure the user experience when we don’t always have control over their device or how they’re using it, what we are able to control is the underlying technology. In the monitoring world, we need precision and granularity and breadth of data to understand the experience of that technology from front to back. Anytime there is sampling involved, things can slip between the cracks. So you ask yourself, did you miss something because you didn’t capture it, or because it didn’t happen?
Lower quality data isn’t just imprecise, sometimes it’s simply wrong. Sampling is an issue many industries need to contend with. For example, in online shopping, a thumbnail of a pinstriped shirt can look solid if this isn't taken into consideration. Audio recordings, weather radar, ultrasounds, movie making — all need to be aware of this issue. It frustrates me that in our space people think we don’t have this problem. Monitoring every second versus every minute yields different results. When we accept bad data we go chasing the wrong root cause. A lot of data used in experience monitoring is not sufficient, and that is not OK!
Experience problems are micro problems that happen in micro moments. The moment something bad happens is the moment employees think about quitting and consumers leave. We need to not get distracted by the good moments and the average moments, and instead focus on those bad moments. Until we get good data about those problematic moments and unify the findings and the insight from multiple perspectives, we won’t get a unified benefit.
Simpler Media Group: What are some ways organizations are failing their customers (and “customers-by-proxy”)? How can omnichannel help?
Hodgson: Richard Branson said if you take care of your employees, they’ll take care of your business. The reality is there are so many functions in a business, and all of them are necessary for a customer to positively interact with the business. To think that something way back down in the weeds of a business won’t impact the customer experience is myopic.
For example, I live in the country on 10 acres with a 1,000 foot long driveway. Halfway down the driveway is the designated package dropoff to keep the delivery traffic away from my kids and pets. But Amazon’s app for home delivery doesn’t issue the delivery notes until the drivers are within a stone’s throw of the house, and every one of them passes the package box and doesn’t realize it until they get to my front door. It’s always a different driver so I can’t train them to do it better, and I can’t call Amazon to fix this problem. A development team that doesn’t understand rural life is impacting my customer experience.
The butterfly effect means that things happening in the business all around the world and on every level impact the customer experience; it’s all tied together. This goes back to what Richard Branson says, and the point is, don’t be myopic! There may be something so far in the weeds that you think it doesn’t matter, but it matters! Don’t forget the flaw of averages. One tweet can destroy a company. Focus on the outliers — the three red dots in an ocean of green.
Human Opinion Trumps All
Simpler Media Group: What are the challenges of balancing rigid technical telemetry with subjective human experience? How might those challenges be addressed?
Hodgson: As I described in my session, there are three approaches to monitoring the user experience. We can learn things about the user experience through rigid technical telemetry. For example, if your CPU is at 100% and Outlook is crashing a lot you’re probably unhappy. That’s experience by proxy. And then there’s objective monitoring: what did you do and how long did it take? Most importantly, we have to consider the subjective human experience: how did you feel about that experience?
Remember, even if your status lights are all green, they’re wrong if no one is happy. Call center associates are measured aggressively on their performance, so they can feel the milliseconds of performance lag. So ask them how they feel — the human experience — and then peel it back to understand and validate the particular operation behind their general sentiment. Track the complaint, figure out the symptom, and then you’ll be able to identify the root cause. Human opinion trumps all in this.
Simpler Media Group: As companies continue to adopt a unified mindset for the customer experience, what is some insight you can share on the short term and long term ROI?
Hodgson: The bottom line is there are different ways you can look at this. The customer experience has some very obvious measures, such as conversion rates. Do customers purchase stuff they’ve put in their shopping cart? When the system lags, earnings go down (the checkout isn’t completed). It’s easy to measure the cost of checkouts that didn’t finish, so you can also ask, if we improve X, what’s the earning improvement?
Internal system issues are harder to track, but usually we look in terms of productivity. For example, you may find that in the journey from answering the phone to hanging up, most associates take four steps, but some take eight steps to do the same thing. This might simply be a training issue, so you can teach the associates to work more efficiently and shave a lot of time off that way.
It’s important to map technical geeky delays to business impacts like productivity. Time is money, and almost everything can be tracked that way. If you convert hours into FTEs, can you show that fixing X will allow you to hire N additional FTEs? Figure out what an hour is worth to the business, and go from there. As I like to say, “Of all the things we could do, what are the things we should do, first?”
Watch this DX Summit session on demand here.