Whether or not a company is born digital, delivering a quality digital customer experience has emerged as a key performance indicator for technology leaders.
So say the CTO at Kayak, the CIO at DBS Bank, and the CIO at Adobe Systems Inc., who expounded on this idea during a panel discussion at the recent MIT Sloan CIO Symposium. Simply put: Customer satisfaction equates to company success, and technology such as artificial intelligence is the link between the two.
The three technology leaders are aggressively helping build a digital customer experience strategy that benefits both customers and the company. Doing this requires collecting data on how customers interact with the company and then finding ways to make those interactions more efficient -- and more intelligent. Here is what each had to say about using advanced technology to monitor, enhance and capitalize on customer experience.
The error budget
One of the most practical pieces of advice on creating an effective digital customer experience strategy came from David Gledhill, group CIO and head of group technology and operations at DBS Bank in Singapore. He encouraged the audience to follow his lead and steal Google's concept of an "error budget," which can help companies strike a balance between moving fast and keeping customer service top of mind.
The error budget, a concept that's evolving at DBS Bank, is a joint key performance indicator between technology and operations "to gauge and monitor customer experience" on digital platforms, according to Gledhill. "Every time a customer gets a performance degradation or [experiences] a struggle, it counts against the platform," he said. Whatever those strikes are -- be they performance issues or incomplete transactions -- the company should determine a threshold and "round everything up to a single number," Gledhill said.
Once the strikes against the digital platform hit the error threshold, developers have to stop and "refocus their efforts on solving those customer pain point interactions," Gledhill said. He pointed audience members to Google's book Site Reliability Engineering: How Google Runs Production Systems for more information.
Mapping the 'customer journey'
Cynthia Stoddard, senior vice president and CIO at Adobe, said AI and machine learning have always been a part of the software company's products. "We refer to it as the Adobe magic," she said.
But what the company is attempting to do now is to use those tools to improve the customer's experience with Adobe products -- especially with its Creative Cloud. "What we want to be able to do with it is really unleash the power and let our customer have access to it so that we can remove the mundane and let people focus on the creativity," Stoddard said.
Cynthia StoddardSVP and CIO, Adobe
Part of Adobe's digital customer experience strategy is to map a customer's "journey" across its product set, which can help illuminate both customer friction points as well as repetitive activity that might be ripe for automation. "Our view of the world with AI, from a product perspective, is more of a Harry Potter view of the world," she said. "We want to do good things and help people do their tasks quicker."
Stoddard said she uses an "outside-in" approach to understand the customer's perspective by "looking at their journey points and ensuring that we remove all friction points," she said. But she also said it's important to look at the world from an inside-out perspective, which focuses on designing for enterprise scale and efficiency.
When the two perspectives conflict, "the customer comes first," she said.
A hybrid approach
At Kayak, the digital customer experience strategy is the strategy for the company, according to Zacharia. "The dominant metric is completion of transaction -- has the user found what they're looking for?" he said.
But as Kayak's developers experiment with how to better serve their users, their ideas can sometimes produce an undesirable result. "If you change the user experience way too much, the users might be taken aback," Zacharia said. "And it takes time to retrain them."
This happened recently when Kayak developers implemented a machine learning algorithm for sorting flights. Rather than sorting by price, the algorithm sorted by likelihood that a customer would complete a transaction. "For some users, the snackers, we call them, those who run a search to see what the current prices are, they were taken aback that they didn't see the cheapest price on top," he said.
Zacharia and his team addressed the issue with a hybrid approach -- the cheapest fare is on top and the rest of the results are sorted by likelihood of conversion. "It works for the user -- for now," he said.