Read This Before You Add Interactive / Conversational AI to Your CX
Technology for improving customer experience
Companies are looking more to interactive (including conversational) AI—in the forms of voice applications, chatbots, and multimode / multichannel interactions—as a way to provide better and more powerful experiences to customers. With the rise of those technologies and smart devices, the ways that people interact with digital products is again changing on a large scale. Digital services on our mobile phones and in our cars and homes are undergoing a massive transition to allow a greater range of input and output that goes beyond screens and buttons.
This expansion is not merely about adding technical capabilities that enable natural language processing. Companies need to rethink existing interactions to account for customer choices that might include moving between new and incumbent channels. Because, as always, customers are focused on their needs and goals, they expect the company—yours—to provide experiences that are easy to understand and navigate especially with new technology.
The experience someone currently has with your digital product is probably the result of a well-thought-out approach. You’ve likely focused on best practices for web and mobile in the context of your business. That’s great and puts you in a good position to start looking at how interactive AI can be another channel to care for your customers well. You are probably also well-positioned to see how current interfaces could support and adapt to conversation-style interactions. If you are deciding or have decided to build your interactive AI solution, we’ve created a list of useful steps to guide your journey from where you are now to a new and improved level of customer experience.
Start with why
Adding interactive AI is a significant undertaking. There is a direct correlation between your customers’ motivation levels to use AI and the success you will have getting a return on your investment. They have to want to use it, and you can’t make them. Since that’s the case, you need to make sure that your new technology will meet or exceed their expectations. Trust is hard to earn, but easy to lose. If your customers have a poor experience with your AI, they will default to trying to get to a human agent, so it only makes sense for your team to get it right from the get-go.
The most common motivations are that the new interactions will make it easier, faster, or more enjoyable to do what they want. Better yet, you want a combination of those. Another common reason is that the changes will enable them to do something that they can’t do yet. When your customers interact with the updates for the first time, you want their reaction to be, “I love this because I can get what I need in a better way!”
Secret sauce: User Experience and Market Researchers are the best people to help you figure out why your customers will be motivated to use interactive AI that you offer.
Skills and roles needed for success
Working with new and more complex technologies like AI means you need to uplevel the skills on your team through education and possibly hiring. In addition to your standard software team capabilities, make sure you have talent in the following areas:
Planning interactive AI projects
User research for interactive AI
Design for interactive AI (We offer a workshop for this)
Data science and machine learning
Development for interactive AI
Testing for interactive AI
Business analysis for interactive applications
Secret sauce: Even with the advances in AI tech, great design is the most critical element to success. Don’t skimp there.
Goals and strategy
As with any major project, it’s important to conceive a plan that includes both vision and execution. The first step is to establish a vision of the desired outcome and goals for it that are appropriate, measurable, and understood by everyone involved. A clear vision and the right goals will motivate the team, guide prioritization and decision-making, enable accurate assessment of progress, and most of all gauge success. Creating a vision and corresponding goals is never as easy as it might sound, so we like frameworks that help like the North Star Playbook, the Project Goal Statement, and a Statement of Design Intent. Involve your team and stakeholders in these exercises. We can help. Collaboration produces better ideas and stronger cooperation.
Using the vision and goals, you will want to create:
Risk assessment and mitigation
Timelines for skill/role acquisition
Priorities, sequences, and execution plans
Process identification and education
Data modeling and metric capture
Go-to-market strategy including an introduction to customers
Secret sauce: Jumping into a big project with new tech and roles comes with risk. Reduce it by starting small, learning well, iterating to proficiency, and then scaling to the bigger vision.
Interactive AI projects use the standard Software Development Lifecycle with additions for specific tasks and elements. This list expresses that:
Research users with learnings from the current offerings as the foundation
Set goals and strategy (above)
Determine required platforms and infrastructure
Document functionality and requirements from goals, strategy, and technical feasibility
Explore, prototype, and iterate possible experiences across modes and channels
Producing the right designs
Modifying and building experiences in code
Verify experience, functionality, and metrics capture
Get feedback from actual users
Iterate for launch-readiness
Launch, monitor, and optimize
Secret sauce: It’s likely that most or all of the team and stakeholders will be new to working with AI. It’s important to acknowledge that uncertainty and problems will be part of getting the work done, and do regular check-ins.
Building well is most, but not all, of the work to reach your project goals. After the initial launch, it’s likely that metrics will show that the goals have not been achieved. Performance optimization is the way we improve user experience and reach overall success. Don’t be surprised or intimidated by issues. Virtually every ambitious project encounters them. The analysis and action of optimization is a necessity given the current state of technology and understanding of human behavior. Our performance optimization playbook and workshop goes into detail about how to:
Capture the right data and information
Analyze performance against goals
Identify problem and opportunity areas
Devise solutions across UI and AI
Design, develop, and test
Launch and iterate
Secret sauce: With interactive AI, it’s important to examine closely when, where, and how people might run into difficulty, even when the probability is low or there are elements you don’t have direct control of.
Alright. Now you’re better equipped to take on a new interactive / conversational AI project. You’ll find it challenging and rewarding, like many of the best things in life. Good luck and let us know how we can help further!
Looking to dive deeper with your team on optimizing voice, chat, or multimodal performance? CCAI offers a 3.5 hour workshop for teams looking to get more from conversational AI.