Your Genesys Blog Subscription has been confirmed!
Please add [email protected] to your safe sender list to ensure you receive the weekly blog notifications.
Don't Show This Again.
I talk with a lot of customers about their bots – chatbot and voicebot deployments – even artificial intelligence (AI) overall. And I’m continually surprised by what I learn. Every business struggles with AI transformation a little differently in terms of defining goals, how to approach AI and understanding what such a transformation entails. And yet there are some universal challenges and concerns I hear over and over.
Torsten Moritz knows these challenges well. He’s responsible for service owner communication systems at Bosch Service Solutions, a division of Bosch. Bosch Service Solutions is a leading international supplier of business process outsourcing solutions for complex, technology-driven services, which includes internal and external clients.
I spoke with Torsten about his team’s process and approach to a large-scale bot deployment as well as what he learned along the way.
Can you tell us about the main challenges in your division?
Torsten Moritz: We support a variety of customers and divisions that represent a huge diversity in services, including support for 35 languages. We run the Genesys platform in 15 locations globally.
The complexity, with all the different requirements, has a big impact on how we build up our platform, how we adapt and push new technologies, and our integration with different bot platforms and CRM systems. In addition, we’re often asked by our customers to integrate not only bots, but other AI technologies too — sometimes after they’ve already chosen a solution vendor.
What were your priorities when you developed your bot strategy?
Moritz: We were most interested in ease of integration and agent handover capabilities, rather than bot capabilities. For many businesses, it’s the other way around. But we see bots as an extension of the service we provide to our customers. Our overriding goal was never about containment within the bot. It was to figure out how a customer could be transferred to an agent without a break in the interaction.
I once asked a customer whose bot implementation failed, ‘How easy was it for the customer to reach the agent?’ I was totally surprised when he said there was no agent there at all! They had no fallback; it was all about 100% containment of the customer within the bot. To me, that failure was inevitable. It might have been a result of not properly managing expectations.
How is that related to expectations?
Moritz: Designing the bot in the first iteration should focus on ensuring that delivery of customer experience is not broken by the bot technology – that is already a good achievement. Trying to achieve 100% will not work because, for some customers, you will never meet the goal. And it’s not even a goal worth achieving. Plus, being iterative is the benefit of AI. Getting better means that it happens over time.
You mentioned that you support 35 languages. How does that influence your decisions around bots?
Moritz: We don’t need to support all these languages now. But being able to do so is important when evaluating bots. It’s all about the flexibility to use best-of-breed technology. Our customers use a variety of tools that are good at solving their unique problems and desired outcomes. We can’t have a situation where we have a bot in English or have no bot at all. We have to be AI-agnostic.
There are other advantages as well. Having independence from a specific bot means we don’t have to redesign anything in the business logic or rethink the platform. We can mix bots based on routing decisions. For example, we might not want to use a bot for a VIP customer because they should be routed to a live agent right away. But, in other situations, or for other customers, we want to keep engagement with the bot — regardless of the bot vendor. We can use any bot knowing that our business rules will still stay in place.
How does this impact your costs?
Moritz: It actually saves us time and money when you consider the cost of maintenance. When something does change, like a regulatory requirement, we would have had to make those changes in all these disparate bot solutions. The chance of missing one or more could cause a lot of problems, including fines for being non-compliant.
Our customers expect that we’re looking at the big picture and reporting on what’s going on with their bots, including overall impact on their contact centers. We can do this better when we’re AI-agnostic. And the result is that we’re delivering seamless customer experiences.
Hear more from Torsten Moritz on bot lessons learned at Bosch. View our on-demand webinar, “5 Things to Consider when Building Bots.”
Subscribe to our free newsletter and get the Genesys blog updates in your inbox.