Messaging: Myth vs Reality (pt2)
Over the years we’ve seen operational metrics that show various forms of digital engagement, whether web chat or messaging, provide operational efficiency benefits. On the surface when you analyse the data, this is actually what it does show and many chat vendor ROI based business cases have been built on this premise.
It’s only when we start applying some rigour that we start to see the flaw in the argument. If I’m grumpy with a brand, I’ll call. If the matter is sensitive, I’ll call. If it is really complex, I’ll call. However, if I just want to know something simple, I’ll message.
Can you see where we’re going with this? Not all customer conversations are born equal but contact centre operational metrics tend to use high level averages to monitor workload and roll up the large number of quick turn round calls with the longer and more complex. Saying that digital engagement is more cost effective than voice channels because I can do 3 chats concurrently doesn’t really explain the situation properly.
Now I’m not saying web chat and messaging can’t provide fantastic customer experiences, just that we need to be careful when investing funds and truly understand the efficiency implications.
The light bulb moment! Unlike voice, where dialects and language reduce the effectiveness of AI understanding, digital engagement (where we’re typing) has the opportunity for a much greater level of accuracy. This means we can now combine a loved channel (Messaging) with a new technology (AI) to offload a large percentage of our consumer conversations from our operational staff. Our customers are happy as they don’t need to wait and our staff are happy as they don’t have to read your returns address for the 50th time that day. They can spend their time truly helping customers with their sales and service needs.
As we consider our AI and automation strategies there are 3 main areas to focus on. Each has their own complexities, costs and limitations. Automation to support:
· Your customers
· Your team
· Your legacy processes
The customer point has been made many times. It’s about delivering their desired outcome, with minimal effort, to provide a perception of excellent service.
The impact is your operational team will now get the more difficult conversations as the customer facing AI runs out of ideas or where the intent is out of scope. To contain the time spent on each customer they need tools that either provide them the right knowledge without having to search, or be able to offload parts of a back end process to automation.
Robotic Process Automation (RPA) has been seen as a standalone technology in our infrastructure landscape for a number of years with the majority of projects delivering back office finance or logistics process automation where the inputs and outputs are well defined. The real value however is when you map your business based on customer journeys, outcomes and intents. Used in this context, and as part of your overarching AI / Automation strategy, RPA has the opportunity to provide much greater value across the whole business. By delivering end to end automation for customer intents and offloading workload from your contact centre staff you are now affecting the efficiency of thousands of employees rather than the hundreds of today’s RPA projects.
Consider this as a panacea - A messaging bot that supports your consumer with sales support (sizing / availability / delivery times), that then provides customer service (‘where’s my order’ / ‘manage my return’), that can then process an application for a store card using legacy back end systems with call outs to credit check agencies.
This all sounds like a wonderful application of messaging and automation technology, and may be the end state but we need to manage expectations and ambition into separate phases to ensure success. In our example let’s break down the components and consider an approach.
We start with data. First, we map out our customer Intents so we can determine which are the best automation candidates by customer value, business value, frequency, solution cost and complexity.
We now know where to start. The FAQ type intents will create the highest volume of contacts and take a very significant proportion of our operational budget even though we’re not adding a great deal of value to our business. At some point we will make a commercial decision to stop automating further intents as they occur so infrequently there is no value. For other intents it’s better for our customers that the team manage the conversation than bake more automation into the system.
We then need to consider the Simple Transactional intents where today our operational teams act as the glue between our disparate systems. There will be clusters of customer intents that will require a particular back end system to be used, so when prioritising which of these to automate map the project cost and delivery time against the cluster that will provide greatest business benefit. Integrating these systems using bot capability to orchestrate a business process will provide a true end to end value out of any RPA investment.
For the complex transactional intents there may be a decision to not automate these at all. A decision to elevate the capability of the customer facing team using messaging and AI is more effective than spending on ever more complex business application and state management code. For example, a bot that understands the context and content of a conversation, automatically providing more information or sending a message to a subject matter expert in the background, will improve the customer experience more than a poor customer facing bot.
Key to addressing this level of automation comes in the form of low-code / no-code application development and integration. Being agile in the build phase and reduced maintenance cost in the production cycle. There are a number of these platforms available but as a minimum they need to provide built in RPA capability, be AI platform agnostic, have out of the box integrations for your key applications, have the ability to build specialist connectors, and be SaaS based to ensure your business is future proofed.