Let’s cut to the chase folks… chatbots have struggled to gain much of a foothold in the corporate tech innovation stack. However, our analyst Melissa O’Brien has spent the last few years studying how these engagement technologies are evolving deeper into the enterprise where they can truly augment staff and reduce a significant amount of their time, while driving a whole new digital way of engaging for both employees and customers… from the back office right through to the front. So let’s take a look at which services firms are getting good at creating these digital associate workers for enterprises…
Melissa, we’ve been observing the evolution of “conversation AI” for a good decade-plus now, so what’s new? has the pandemic driven more uptake?
The demand for conversational AI has exploded over the last year. The automation tools we call digital associates were one of the digital superheroes of the pandemic, as conversational tools picked up the slack in handling volumes of interactions when human associates were not available due to a lack of work-from-home preparedness. Reduction in staff coupled with spike in volumes of interactions in many industries such as ecommerce created a burning platform. For many companies, this rapid and massive disruption resulted in accelerating digital initiatives already in play and pushing many lingering POCs into production mode. Industries like travel and hospitality, which had a brief period of an incredible uptick in customer interaction volumes a year ago, have now developed conversational tools poised to address inevitable future surges driven by pent-up demand.
Now that acceptance and adoption has increased significantly, the imperative is to move beyond the low-hanging fruit, the really simple and repetitive stuff, to really see what these tools are capable of. The next evolution of DA’s is a lot more about using ML and AI for conversational complexity, having a good design that fits into an overall experience strategy, and integration with enterprise systems in order to deliver greater value. Many companies which hadn’t dipped a toe into this space pre-pandemic are struggling a bit with the learning curve now, where others that had a foundation built already are expanding their use cases. Enterprises are now really looking at digital associates from an employee and customer experience perspective rather than only call deflection, efficiencies, and cost savings. In many cases, we’ve seen them as effective tools for sales conversions and building brand loyalty. So there are a lot more potential outcomes that companies are now looking to conversational AI to help achieve.
Why do you think this space has struggled to attract the feverish hype of RPA and “dumb” back office automation? What needs to happen to get the digital associate value proposition weaved into the whole intelligent automation and AI narrative?
I think part of the trouble is this market landscape is vast, the tools poorly defined and buyers are confused. This is part of the reason we’ve taken to calling the tools “digital associates,” to convey that they are digital workers built to provide a tangible outcome. Digital associates are not distinct segments like AI, automation, and analytics – in fact, they are tools that if designed properly use elements of all three to maximize their power. The face of automation has been largely tied to RPA, a singular tool where a lot of the use cases are straightforward – the quick time to value in many of these tools makes it a fairly fast and easy win for certain processes. With conversational AI, there’s a lot more nuance to think through employee and customer journeys, the complexity that is human language and interaction, and the connections that need to be made in disparate back end systems for them to operate in a meaningful way at the engagement layer.
Digital associates run the gamut of maturity, but often get associated with their simplest cousin, the chatbot. Chatbot fatigue has been a real roadblock; 10+ years ago the fad was to slap an FAQ bot on every website, expecting it to deflect calls and create efficiencies. More often they provided little help outside of website navigation, and created customer frustration and degraded the experience in the process. So people who hearken back to those experiences will have an aversion to trying out conversational tools or be hesitant about their value. This is obviously changing now as the real digital associates are starting to prove their worth.
I think this is where the OneOffice helps people conceptualize the impact digital associates can have on an organization. Firstly, start with an EX or CX focused outcome objective and design the DA backwards from there. As you’ve said yourself Phil, digital associates are augmenters of the human experience. They start to add more value as they become more sophisticated and embedded in our processes. And thinking in terms of the Triple-A Trifecta, the digital associates’ AI-powered brain is complemented by automation arms and legs to retrieve, report, update and transact. Data is the fuel the associate ingests to produce insights (analytics) and achieve optimal performance, including the personalization users are expecting.
And which suppliers are currently winning in this market, Melissa? Both in terms of services and products?
Accenture, IBM and Cognizant took the top three spots respectively in our recent services Top Ten. These providers demonstrated an impressive breadth of capabilities across industries and enterprise functions, both IT and business, and a depth of design, process and technology prowess. While we didn’t include advisory firms in our assessment, focusing instead on BPO and IT services implementations, firms like KPMG and Deloitte have significant depth of expertise in developing solutions for clients in this area also.
On the products side, the number of vendors having some level of capability in this space is astounding. You’ve got the cloud providers which all have their platforms and developer suites, namely Google Dialog Flow & CCAI, AWS Lex & Connect , Microsoft Luis & Power VA. And then your established OOB software firms like IPsoft and Kore.AI, and including some of the RPA firms such as Pega and NICE dabble in various forms of digital associates. And we’ve been watching some interesting up and comers like last year’s “hottest hot vendor” Techforce.ai, and most recently we’ve been briefed on XpertRule’s Viabl.ai platform which comes in with an interesting value proposition focused on decisioning.
We’re really seeing this as an ecosystem play right now, as most of the successful implementations we’ve seen have required fairly complex customizations and integration from the service providers. While the products players tend to focus on selling licenses, the service providers provide with industry knowledge, deep understanding of clients’ processes and design expertise to help companies think through employee or customer touchpoints and really map out what they want to accomplish. And most of the service providers are completely technology agnostic, so they can come in and leverage an existing vendor relationship or preference or help evaluate vendors and select based on their requirements.
How are the call center firms approaching digital associates – are they reacting to customer demand, or still fighting to protect the legacy “butts-on-seats” model? Which ones are unafraid to disrupt the model and push conversation AI aggressively?
Many legacy contact centers still operate on FTE volumes and they’re safe for now because there’s still plenty of demand for labor-focused services (which is being driven quite a bit by growth in digital native companies whose rapid growth requires help to scale operations.) But most of the contact center leaders have a balanced approach which strategizes with clients on digital-first customer journeys which include a blend of automation and agent support. Concentrix is one of the most aggressive we’ve seen in this market, which landed at #9 in our recent report, the top ranking of any of the contact center pure plays. The challenge that the CX players continue to have is that of brand perception. They know that with the increase in self-service and automation that in many client engagements, if they don’t cannibalize their business another provider or vendor will. So they’ve built solid AI and automation capabilities for those clients with the appetite to work with them in that regard. The trouble is that some enterprise tech leaders won’t consider a contact center to work with on emerging tech, so in order for them to really gain traction in this space, there’s a mindshare gap to bridge, especially with IT decision-makers.
And finally, Melissa, how do you expect digital associates to evolve as we get used to new ways of working? are they going to branch way beyond customer-facing solutions?
Customer service examples are the most ubiquitous, but digital associates have already spread well beyond that realm. In the top ten, we saw a plethora of use cases in procurement, HR, IT helpdesk, finance, accounting, and much more. One of the interesting surprises we found was the prevalence of recruitment and HR bots. We’ve seen conversational AI tools which can help shepherd a candidate through the application and interview process, and also offer employee support once onboarded, helping staff access, understand, and process important information around benefits and compensation, performance, and time reporting. This is really a result of how accustomed we’ve become to automation in our consumer lives, that we expect to interact with digital ‘colleagues’ at work. And while chat solutions are still very commonplace, I do think the future of this market lies in voice. Not phone-based or intelligent IVR’s, but similar to virtual assistants like Alexa and Siri we’ve become so accustomed to in our consumer lives. Voice is opening so many doors right now to create new ways for people to engage either with our work as employees or with companies to buy services and products.
Many of the examples we saw were also quite industry-specific solutions, and I think we’ll continue to see greater depth in those capabilities too. Personalized menu customization for restaurant chain ordering, retail loan processing, and approval, program enrollment and Rx management for healthcare payers, and retail order status digital associates were just a few of the highly verticalized solutions we’ve seen.
One of the important topics I expect to see the industry tackle next is how these tools are supporting diversity and inclusion efforts. For example, service provider TTEC has developed a bot as part of a broader D&I offering which allows employees to ‘practice’ difficult conversations, around prejudices regarding race or gender for example, with the conversational AI tool. The platform then provides feedback and advice on how to handle the conversation. Another big one to unpack is how these tools play a role in influencing and in many cases reinforcing gender bias (did you know that “female” digital associates get greater adoption levels than male ones?). We recently saw Accenture coming out with a non-binary offering as an option to address this. I think we’ll see a lot coming from the top service providers which frame digital associates in the light of “ethical AI.” How to ensure an ethical approach an important conversation to have, because emerging and increasingly sophisticated conversational tools will help shape our future reality, from the fun and entertaining, to potentially serious and unnerving. The conversation is an extremely complex AI problem to solve, but with all the opportunities we’re seeing emerge, it feels like this is just the beginning of exploring the possibilities.
HFS premium subscribers can access the Digital Associates Top Ten Services report by clicking here
Posted in : Customer-Engagement, customer-experience-management, Digital Transformation, OneOffice
Chatbot fatigue is a result of either poorly designed conversational flows / stories or a poorly performing intent classification model, in my opinion. I hope level 4 and 5 chatbots will be the norm soon and we at NeuralSpace have all the capabilities to do so. Conversations with chatbots and humans will almost not be distinguishable.
Level 4 and 5 chatbots: https://blog.rasa.com/5-levels-of-conversational-ai-2020-update/