Step 5 of 5: AI Self-Service Without Compromise – Human-Centric Design from a Team of CX Experts
This is the fifth of a five-part blog series that outlines the Five Best Practices for AI Self-Service Without Compromise. Use this guide to automate your contact center and Customer Experience (CX) with AI self-service in voice, chat, and text.
Best Practice #5: Human-Centric Design from a team of CX Experts.
Steps 1-4 in the Best Practices for AI Self-Service Without Compromise lay out the blueprint for automating more without sacrificing an ounce of CX. Step 5 discusses the kind of CX team needed to execute the vision. While conversational AI technology is amazing and transformative for the contact center, it is only a toolset. In order to automate more without sacrificing even a drop of Customer Experience, you need a team that is experienced in different CX disciplines from pre-production to post-production.
What does a best-of-breed CX team look like?
Conversational AI self-service technology requires a CX team across seven functional disciplines. This blog will serve as a guide to all those disciplines and why they are required to deliver CX without compromise. Those functional areas are:
- CX Consultant – maps the business need to a business case for conversational AI
- Solutions Expert – architects the solution roadmap and ROI
- Project Manager – aligns all resources (CX Design, Engineering, IT, QA) to deploy solutions quickly
- CX Designer – acts as the human experience activist to design effortless conversation flows
- Quality Assurance (QA) – exhaustively tests each application for optimal user experience
- CX Insights Manager – performs ongoing monitoring and analysis of data and recordings/transcripts
- Success Manager – ensures the solution meets expected goals and provides regular reporting
Let’s explore each individual role more deeply to better understand the importance of each and how they work together to create the optimal CX.
The CX Consultant brings together key stakeholders to understand the business case – the problem that needs to be solved and what it would mean to the business to solve it. This person must have a deep understanding of your business, conversational AI automation, and overall contact center technology & infrastructure because they are responsible for outlining the use cases to be automated. The CX Consultant identifies not just where to start but what the roadmap should look like for additional contact center automation – what the vision for the project should be. This vision needs to be coupled with a business case that can be presented to upper management/C-suite for buy-in and signoff. It is the CX Consultant’s job to identify what’s needed to make the business case compelling, honest, and low-risk.
The CX Consultant works hand-in-hand with the Solutions Expert and the Success Manager, but also must be an excellent collaborator across your entire CX team, contact center, and business. It is their job to outline the business goals that need to be achieved, and generate excitement about contact center transformation via AI-powered virtual agents.
Solutions Experts are the ‘right hand’ of the CX Consultant because they are responsible for architecting the preliminary outline of what a call center virtual assistant solution might look like. To do this, Solutions Consultants spend their time digging into data – both quantitative and qualitative. First, they review contact center data: call volume, chat volume, average handle times, customer satisfaction scores, etc. If the data is broken out by each individual use case, that is even better because they will be more equipped to identify the perfect call types and chats for automation based on contact center deflection and ROI expectations.
The Solutions Expert gathers info from a litany of qualitative sources, including live agent interviews, training manuals, call recordings, agent desktop, and anything more you have. They literally sit with your live agents to monitor workflows, processes, and data sources to see how well that fits within conversational AI capabilities.
This helps them fill in the gaps left by the quantitative data. For example, the quantitative data might indicate that a particular call type has a long average handle time (AHT), which could mean that the process is rather complex. But after a live agent interview in which the live agent walks the Solutions Expert through their process, it may become clear that the agent process is extremely manual: moving between multiple systems, authentications, and sometimes even copy-pasting data. With API integrations to these multiple systems, virtual agents can handle these processes faster and easier than a live agent, which will free up the live agent from tedious tasks. What you’ll find is that the more inputs Solutions Expert receives, the better he or she can be. This is why data and documentation in your contact center are so important.
Solutions Experts are responsible for a number of key deliverables. First, they work closely with the CX Consultant to identify the right call types and chats in your contact center that are perfect for automation. (Importantly, Solutions Experts are well-trained on what not to automate, as that is equally as important as determining where to start.) Then, they design a solution roadmap tailored to your specific needs. They also use your contact center data to develop detailed ROI calculations that incorporate projected call center deflection across various call types and chats. Lastly, Solutions Experts become very important to the actual build and implementation process because they represent you in describing the solutions and projected outcomes to Project Managers, CX Designers, Engineering, and Success Managers.
The Project Manager aligns all resources and keeps everyone on task to meet timelines. They serve as the liaison between all CX Team members, IT, and engineering to ensure goals are being accomplished. Simply, the Project Manager ensures goals are understood, deliverables are clearly outlined, and timelines are met. One person has to be responsible for coordinating the below:
- Detailing solution specs to the engineering team for any required application customization;
- Coordinating IT and engineering resources on telephony integrations;
- Driving webservices development for integration to data sources and other systems;
- Obtaining QA resources for testing and analysis;
- Tracking changes to avoid scope creep;
- Pushing solutions to “go-live” (i.e. virtual agents begin taking live calls);
- Analyzing and communicating initial results from CX Insights Managers;
- Working with Success Managers to meet overall goals for automation.
The Project Manager has only one thing in mind: to get customer service automation solutions “live” as quickly and accurately as possible. Their role is extremely important because the faster you can stand up an automated call center solution, the sooner you can begin gathering data about how customers interact with it. The sooner you have that data in the form of live call results and recordings, the more adeptly you can make changes to improve success rates. And of course, strong success rates deliver ROI.
The CX Designer usually joins the discussion once you’ve decided which call types or chats you’re going to automate. The primary function of this role is as a behavioral scientist – a human experience activist who represents your customer by designing conversations that are simple, fast, and effortless. One of the biggest misconceptions about virtual agent automation is that the AI development regarding natural language understanding (NLU) is the hardest piece to get right. However, you could have the greatest conversational AI on the planet but completely whiff on the CX. Using AI to create a smooth, effortless customer experience requires deep understanding of conversational AI capabilities and how to map out that human-to-machine experience that is as good or better than speaking with a live agent.
As much as contact center leaders may have expertise in processes and call scripts for live agents, it is very different when architecting for a human-to-machine interface. The CX Designer leverages all of their experience designing self-service to develop optimal conversation flows, “guardrails,” and “business rules” for containment and transfers. In doing so, the CX Designer’s ultimate goal is to make the virtual agent “invisible” to the user; in other words, the conversation is so natural and easy, the customer forgets that they’re speaking to an AI-powered virtual agent.
This team will typically start by digging into the call types to automate by listening to call recordings of live agents, reviewing training manuals or process flow documents from your contact center, and looking at call scripts. This gives them a sense for how the call type is handled today. Then, in consideration of the capabilities of AI-powered virtual agents, they begin defining the “happy path” for that call type. Think of the “happy path” as the perfect conversation between caller and virtual agent: the caller knows what he/she wants, the virtual agent captures all intents correctly, all data is available, and the caller accomplishes the goal.
Once the “happy path” is designed and well-understood, the team begins building in the common exceptions. Process flow diagrams (e.g. Visio diagrams) demonstrate all the different “forks in the road,” intents and outcomes, and how the virtual agent will handle each one. CX Designers take careful measure to put themselves in the shoes of the customer, which helps them design with empathy and insight into how actual callers will feel and act when they find themselves on the phone with a virtual agent. Most every person that dials customer service is feeling some level of tension and it takes an experienced designer to ease that tension as quickly and seamlessly as possible.
You might think the job of the CX Designer is architecting the initial build and implementation, and then they’re done. However, that couldn’t be further from the truth. The CX Designer is always involved in some capacity. During the initial rollout, CX Insight Managers monitor data and customer calls closely to dissect the CX and every possible reason that customers might hang-up or be transferred. This is fed back to the CX Design team to tweak conversation flows and/or “business rules,” then A/B test those changes. This is what triggers success rates and customer satisfaction to improve initially. However, this process between CX Insight Managers and CX Designers continues on an ongoing basis, which causes success rates, CSAT, and containment to continue to improve over time.
Quality Assurance (QA)
Your customers are going to take calls seriously, and so should you. Your QA team does just that. Their role is to test solutions exhaustively for every possible flaw. They pinpoint problem areas, identify potentially confusing prompts, test every possible intent and outcome, and make suggestions for improvement. Naturally, QA is a technical role; but in testing virtual agent solutions, you want your QA team to be CX-obsessed – just like you. With CX as the new battleground, QA is on the frontlines of defining what yours will look like.
QA is a two-part role. First, they need to perform real-life tests to evaluate virtual agent performance. When they find bugs, they discuss with Engineering and the Project Manager to get them fixed. Second, once solutions “go live,” they work closely with the CX Insights Manager to listen to live call recordings and determine why certain things are happening. For example, a CX Insight Manager might see that 25% of callers are being transferred after a virtual agent asks a particular question, so the QA team will listen to 100 of those calls to find out why. The QA team will then make recommendations about what to change in order to reduce that transfer rate and improve success.
QA is a valuable resource to the CX team because they check the work of everyone else. The CX comes first and QA guarantees that remains true.
CX Insights Manager
With many AI self-service implementations, the work ends following a short tuning phase after it is deployed and starts taking live customer calls and chats. That approach is ultimately the “Achilles heel” of most automation projects, and a surefire way to frustrate your customers. This is why CX Insights Managers are so important. One person must be dedicated to monitoring the solution on an ongoing basis, which means pouring over data and examining containment rates against outcomes. This is a job that is never “complete” because it is an endless pursuit of perfection to improve the CX and containment rates over time.
CX Insights Managers get involved as soon as the solution begins taking live calls. As mentioned, their responsibilities never end as long as the solution is running. During the first week or so, CX Insights Managers are evaluating all aspects of the virtual agent’s performance, including such granular details like how performance is being tracked. Are you capturing the right information? When a call is marked “successful,” was it really successful? Where are customers having the most trouble? What can we do to fix it? What is the number one reason for live agent transfers and what can be done about it?As described above, CX Insights Managers rely on QA to listen to actual call recordings when the data appears actionable, which helps them find fast solutions to any early speedbumps.
After initial tuning, most solutions smooth out and reach the desired success rates outlined by the Solutions Expert. However, that’s not enough for the CX-obsessed contact center. You want a CX Insights Manager who will continue to dig into the data and continue to find ways to improve the CX. On an ongoing basis, CX Insights Managers work closely with Success Managers to build data-driven reports that help show the continued value of automation. These reports are valuable to share with executive stakeholders in your organization to prove out success and continued support for future AI automation projects.
A CX Insights Manager’s goal is to work on behalf of the contact center to identify every possible way to improve the CX until reaching perfection.
Last but not least is the Success Manager. This team is business-oriented and their mantra is to get things done. While Project Managers coordinate resources to build and deploy solutions, Success Managers make sure that those solutions meet the goals of the organization from a bottom line and CX standpoint. Thanks to CX Insights Managers, they take a data-driven approach to review how solutions are performing and connect that performance to the initial business case for automation. Success Managers provide regular reporting to showcase the value of every conversational AI solution; they know which strings to pull, so that your virtual agent solution will continue to deliver results. Collaborating across the entire CX team and Engineering, your Success Manager ensures the first phase of the solutions roadmap meets expectations post-deployment.
Once the Success Manager demonstrates success in the first phase of the solutions roadmap, they take the lead on the next phase for further automation. With intimate knowledge of your business and goals, they ensure the next phase of the solutions roadmap contains the right call types. They also shape which voice-based solutions will scale well to chat or text for a seamless omnichannel experience. Success Managers drive the omnichannel vision of your organization with executive stakeholders and bring in the right resources on your behalf to make sure it gets delivered.
Why the CX Team is Important
Without a team of CX experts across all these functional disciplines, your conversational AI implementation is doomed to fail. You may have seen it happen yourself. Conversational AI is amazing and transformative technology for the contact center, but it is only a toolset.
For the most human-centric CX possible, SmartAction bundles its proprietary conversational AI technology with a team of experts across each of these CX disciplines. SmartAction understands just how hard it is to architect AI automation and perfect it over time. That’s why SmartAction has curated and developed the necessary CX talent, enabling organizations to simply outsource their automation needs instead of doing it themselves. As hard as self-service automation can be, nothing makes life less hard than that. Meet the SmartAction CX Team>>