When it comes to designing great IVR and conversational AI user experiences, understanding and designing for context of use is critical. Context Gathering is core to the VoxGen user-centred design process, and for very good reason.
Before we even start thinking about design solutions, we need to understand the specific user requirements and context in which the product will be used.
Take an IVR for example. We need to identify the user tasks that need to be supported, the information needed by the user, where and when users will be using the IVR, triggers for calls and how these all fit within the wider customer journey. Our Context Gathering process allows us to conduct the early research that enables us to understand these user requirements early on and feed that insight into an iterative design and evaluation process.
Great IVR and conversational AI designs are about more than just users
Understanding user needs, behaviours and expectations is core to great IVR and conversation AI design and a key focus of our Context Gathering research. But we also need to understand how user requirements align with business goals.
Our process also focuses on understanding business processes so that we can uncover areas where user and business requirements may not align. For example, one client we were working with had a business process requirement that all callers must go through ID&V (identification and verification) before their query could be answered.
Through call listening and user interviews, it was soon clear that callers had a very different expectation. They expected to be asked for their account number when calling about something related to their account, but not for more general queries. If we had designed the IVR around a business process that didn’t align with customer expectations, IVR performance and usability would have been compromised. Instead we were able to work with the client to design the IVR around caller expectation and support callers who didn’t have an account number to hand.
Is the optimal IVR and conversational AI user experience technically feasible?
Steve Jobs has been famously quoted as saying “You’ve got to start with the customer experience and work back toward the technology – not the other way around.” Here at VoxGen, you’ll hear and see many people quoting this, from our Sales team to our UX team. It’s core to our approach.
However, we sometimes need to be pragmatic in terms of what is technically feasible. The reality is that technology can impact the feasibility of the implementing the optimal IVR and conversational AI user experience.
Context Gathering allows us to identify the optimal user experience and work back from there. While we always have the end goal in sight, it may be necessary to work in iterations or phases in order to get there. Understanding the technical landscape that we’re working within means we can identify where legacy systems may not be up to the job yet, but help clients put a roadmap in place to work towards the optimal experiences for their customers.
The IVR and conversational AI Context Gathering process
So, what does Context Gathering research typically involve? As with any user research approach, the methodologies we select are dependent on the project, timescales, available budget and research questions we need to investigate. However, there are some core methodologies that we include in our process that give some of the biggest insights when designing conversational interfaces like IVR, SMS and chatbots.
A great starting point is to talk to stakeholders across the business. These are the people that know their business inside and out. They understand some of the current pain points and can offer valuable insight into known issues, strategy and business priorities. The essential component of successful stakeholder interviews is to talk to stakeholders from different parts of the business. Customer Service stakeholders are likely to have different needs and priorities to Marketing or Technical stakeholders for example. Interviewing stakeholders from as many business functions as possible helps to build a full picture of the business needs, expectations and goals and enable you to build better conversational AI experiences for your customers
Our projects typically involve companies with a contact centre. Agents are at the heart of interactions with customers. They know the typical call types, triggers for calls and even reasons for customer frustration. Agents provide great insight into how calls are handled. When interviewing agents, it’s important to speak agents from different skill groups. Technical support agents will handle different types of calls to service agents who may handle queries like payment and order status. Understanding the specifics of these different calls is critical designing great IVR, SMS and conversational AI chatbot user experiences.
Without doubt, call listening is one of the research methods that can produce the most valuable insight, particularly if we’re designing a new IVR. Listening to calls of agents interacting with real customers highlights specifics around reason for call, customer pain points, triggers for calling, and whether callers have tried to use other channels first. It also provides some insight into the callers’ state of mind when making the call. For example, someone calling to report a stolen phone or someone who can’t make a payment by a due date will be in a very different psychological state to someone who is calling to get a balance. It’s the intricacies of this kind of insight that can turn a good IVR into a great IVR. If we can design for empathy and emotional engagement in the right context, callers will have a much better experience and brand perception is more positive.
Call listening is also important to inform the design. Understanding the dialog between the agent and the caller can help inform the task flow and dialog within an IVR. Identifying the language and terminology used by the caller means that we can ensure we use user-friendly language and avoid the use of jargon.
Evaluating the existing experience
To improve an experience, we need to understand and evaluate the current experience. This can include things like an expert review of an existing IVR, analysing current performance data like task completion, call volumes and user behaviours within the existing IVR (hang ups, errors etc.). We may also conduct baseline usability evaluations or run surveys to get feedback on the current experience. This not only helps identify requirements and opportunities for optimisation for our redesign, but also provides important input into benchmark criteria that a redesign can be compared against to ensure experience and performance is being improved.
Finally, clients are always interested to know how they compare to their competitors and to understand what can they do to stand apart from the competition. At VoxGen, we work with clients across many sectors (including retail, telco, health, insurance, finance and utilities) which enables us to bring insight from our own experiences in terms of what makes a great conversational and conversational AI experience. Competitor reviews are also an important insight into functionality and experiences offered by direct competitors for the clients we’re working with.
Context Gathering is a critical part of our user-centred design process here at VoxGen. Early research and evidence-based design not only helps us to produce better designs and great user experiences, but also means we can ensure business goals and user goals are aligned, and highlight to our clients when that may not be the case early in the design process. But Context Gathering is also only one part of our user-centred design process. You can find out more about the VoxGen user-centred design process and our research approach from our Design Cloud.
Design IVR Conversations not Call Flows
Planning any type of automated customer communication can be tricky, but in the voice channel it’s...
Designing conversational AI chatbots and what we can learn from IVR
Conversational AI chatbots are a hot topic. Looking at our Twitter and LinkedIn feeds, the interest...
The what, why and how of great conversation & conversational AI design
It wasn’t too long ago that the only way customers could have a ‘conversation’ with a company was...