Innovation and market perspectives from leading IOT innovators
The second part of our conversation with Dale Calder highlights how RevTwo is using product context to compress the steps to solve support issues. By collecting data directly from connected products, RevTwo’s in-app support uses AI to solve 75-90% of user problems upfront. Omni-channel support (communication via email, Twitter, chat etc.) does not accelerate case resolution when each interaction starts with “20 questions” to assemble context; “optimal channel support” directs the user to the channel best equipped to resolve issues quickly. Context and connectivity combined with AI have the potential to transform support from a cost center to a revenue driver for businesses.
|
|
Dale Calder |
Dale Calder is a seasoned industry executive and serial entrepreneur, having started and successfully sold multiple companies. He founded, scaled, and exited Axeda Corporation, which was acquired by PTC Corporation for $170 million. Most recently, he founded RevTwo, Inc. a startup focused on bring support to the next generation of smart products, including iOS and Android Apps, IoT devices, Docker and Snap based Microservices, and Windows and Linux based machines and devices.
What are some of the shortcomings in how product support is currently delivered?
When you talk about product support it could be IT support such as a traditional help desk, products that are shipped “in the wild”, or industrial maintenance processes that are in place when there are things running. Smart products don’t necessarily mean smart users. Products will always need support - even the best designed products fall into this category.
We see this domain as very broad. You have big products: industrial, large scale equipment all the way down to consumer products using a smartphone app connected to a home automation solution. All of these problems have similar challenges with operation and use: when something goes wrong, there is typically someone responsible for helping resolve the issue.
Two years ago, we wrote a series of responses for someone asking for help. Even with a gateway product or headless device, the product can ask for help. Help initiation is the first step in product response. If this doesn’t get the job done, we have escalation capabilities to facilitate more expert assistance. In the past consumers, wouldn’t interact in real time with the product at the same time. With RevTwo we condense the process to a couple of steps: Step 1- solve the problem, Step 2- initiate an interactive support response.
Can you characterize the unique challenges to deliver effective support?
There are two things to be leveraged: context and connectivity. What shocks me has been how context is ignored. Chatbots are largely trying to replicate what a human would do and there are a lot of good people making bots trying to solve solutions. This can be technically challenging problem through, using queries to collect the data to gain context. When you are asking “20 questions” with humans or bots, the whole activity is about trying to get enough context to get to a person’s issue. Almost always you burn time and money. The whole support industry has been built on lack of context – it’s like The Emperor’s New Clothes. Omnichannel support is the hot trend in the industry, where I can ask a question on email or Twitter and someone will answer. There is no context - this is a “hello” relationship.
RevTwo has something called “optimal channel support”. Instead of having something that’s designed obtusely that customers have to go to twitter for help, you want to design support so customers will go to where there is the most context to get the best experience. You want to be avoiding answering all of the same questions when you’re going through different levels of support. This is inefficient and it’s the wrong set of solutions.
Now when we have smartphones that know who you are and what you are doing, they can provide the context and enough information to avoid the “20 Questions” debacle. When there is context, an AI [agent] can do the job better than a person. The industry ignoring the whole piece of the puzzle – it’s like going to step 6 while ignoring steps 1 to 5. If you knew these things up front you could do support differently.
At Axeda we were able to see the benefit of insight powered by data, as we were so immersed in the culture of the edge. We were leveraging the power of information from remote products, and this helped us look at the problem in a different way. The RevTwo approach gives end uses a running start on support.
Why haven’t others thought to take a similar approach to support?
Support is a lot like putting out fires, and when the fire hose is operating at full bore, you are just worried about emptying the water from the room. People have a firefighter mentality playing Whac-a-Mole. People never stop to ask themselves if should they be playing Whac-a-Mole: if they even stop to ask the question, they can lose their job. If you can take the size of the pipe down by 70%, you can really think about how you want your resources to interact with customers. Companies like Guess are able to use their support to provide style advice.
How will AI impact the jobs of all the people working in call centers?
Most people worry that AI will take jobs. What AI is taking away is tasks that aren’t worthy of a human’s time, and this actually increases the value of a human being. The bank ATM industry is a great example: they made remote branches cheaper, and when costs went down, tellers weren’t just counting money, they could become an extension of the marketing organization.
The opportunity is not just Whac-a-Mole, saving money on the cost of help desk tickets - now you can redefine how you deploy your resources with things that only humans can do. This increases brand value and customer loyalty – and takes support from a cost center to a place that generates real revenues. That’s how we think about it and that’s where the big opportunity lies.
Share it
