Today’s interview is with Michael Ramsey, VP of Customer Workflow Products at ServiceNow, a cloud computing platform that helps companies manage digital workflows for enterprise operations. Michael joins me today to revisit the idea of the messy middle of customer experience, what’s happened since we last spoke about this around 2 1/2 years ago, what companies are doing to clean up their messy middles, the role AI and process mining plays in that, how generative AI is helping with this sort of endeavour and who’s doing it well.

This interview follows on from my recent interview – The pragmatic approach to transforming customer service – Interview with Mike Upton of First Tech Credit Union – and is number 472 in the series of interviews with authors and business leaders that are doing great things, providing valuable insights, helping businesses innovate and delivering great service and experience to both their customers and their employees.

NOTE: A big thank you goes out to the folks at TextExpander for sponsoring this episode of my podcast.

TextExpander is an auto-complete tool that allows your team to eliminate repetitive typing and stay on the same page with just a few keystrokes allowing you to delight more customers in less time.

Click here or on the banner image above to find out more and to get a 20% discount for the first 12 months of TextExpander if you use the code: SWINSCOE

Here are the highlights of my chat with Mike:

• Michael first appeared on the podcast talking about the messy middle of customer service back in February 2021.
• The messy middle is still a thing but the idea is certainly resonating with more and more people and many customers are evolving their approach and making some progress on it.
• Another way to think about the messy middle: What are the operations or the customer operations behind your products and services that you deliver to your customers? And, what work needs to be performed on behalf of a customer?
• Another way that we’re talking about this is that the messy middle is essentially glued together by humans. It requires human middleware, if you like, to get this work done.
• Over the last couple of years, customers have become much more cost conscious and they don’t want to make an investment lightly. But, there is optimism and a conscious view about using resources efficiently to drive growth and better customer outcomes.
• One of our Telco customers, Rogers up in Canada, have fuelled their growth, like many in this space, through acquisition. This created an even more complicated, messy middle than other organizations with many customers having multiple services. But, internally to Rogers, those customers were served by different customer operations teams. We helped them centralize those so they had common processes and common teams even though historically these were separate teams. They had some really interesting outcomes just in terms of how they’ve enabled customer self-service and were able to achieve a 41% decrease in the number of customer cases that had to be worked on.
• Process mining and what we call process optimization allows us to put the scaffolding in place for digitizing a process.
• Even if humans are having to perform a lot of tasks or actions within that process, we can start mining the data from that process so we know who it gets assigned to, if it gets assigned from one person or one team to another, when tasks get performed or completed and others get kicked off. We have all that data within our system and we can start service insights based on that to see, for example, if that work item is getting bounced back and forth internally and why.
• Gen AI is really compelling new technology and I think it is going to have an immediate impact on some areas.
• For example, for ServiceNow, it helps us solve a problem that was very difficult for us to solve with previous A I. One of our challenges was that for something like a bot or a virtual agent it took a lot of training data and highly skilled people to deploy a virtualization or a bot with natural language understanding.
• We can overcome that very quickly now and can deploy self service experiences that used to take months in weeks now and, in some cases, even days.
• However, right now you still need to do work to give the customer what they want.
• An interesting way to think about it is tool time vs task time.
• Example of application of Gen AI with Starbucks, the coffee company, to help in situations where work has to be done across the company to correct an issue. For example, if a customer orders a coffee for pick up from the app but they ordered it be picked up at the wrong location. Leveraging gen AI we were able to create a solution that either refunded this order or transfer it from one store to the other without needing a lot of human intervention and approvals.
• We should be empowering employees to use their own discretion. But, it’s also important to make sure they have all the information they need so that they can actually use their discretion properly.
• The reality is that technology and digital services are helping automate and make self service more and more accessible to more people so the likelihood is that when you contact a brand about something, it’s more likely to be a more complex kind of problem. Therefore, it’s increasingly likely that they will end up touching those messy middles. That makes it a very present and a very real problem.
• Just because it’s complex, doesn’t mean you shouldn’t tackle it.
• To go and find some of these messy middles, look to your data and the top and most frequent requests you get from customers….those are your low-hanging fruit. Then think about the outcomes and experiences that you want your customers to have and work backwards from there.
• Michael’s best advice: Do the discovery on how your customers interact with you and follow it to end. Understand all of your partners, your employees, the different personas that are gonna either interact directly with that customer or if they’re going to do work on behalf of that customer and understand what tasks they need to perform, what context they need to perform those tasks in and what they need to know to be successful.
• Do that regularly because things change and you hire more people, you roll out new products, you acquire or get new customers, customers change as do their needs. This is not a one and done thing.
• Michael’s Punk XL brand(s): Xerox, Starbucks

About Michael

Michael Ramsey is VP of product management for ServiceNow’s Customer Workflow products, which enable organizations to create seamless customer experiences and drive fierce customer loyalty. He’s responsible for strategy and execution throughout the product lifecycle, including managing strategic partnerships, investments, and mergers and acquisitions.

With more than 20 years of experience as a product manager, solution architect, and software consultant, Michael brings with him deep expertise of the SaaS business model and enterprise software, as well as in-depth knowledge of cloud and mobile software development with hands-on experience running billion-dollar product lines.

You can find out more at ServiceNow’s website, say Hi to them on Twitter @servicenow and connect with Michael on LinkedIn here.

Credit: Photo by Khara Woods on Unsplash