Now that we’ve covered some of the basics behind why voice is taking off (Part 1), some key terms and design heuristics to know (Part 2), and how to demonstrate return on investment (Part 3) – we’ll finally turn our attention to the critical steps to envision and execute an Alexa skill that meets your customer’s needs. Remember, it’s not just about having a Skill in the marketplace. It’s about having a Skill that positively impacts the customer experience.
While Amazon provides prescriptive instructions on the steps needed to bring a Skill to market, we’ve outlined four steps to ensure that your skill is customer-focused and ready to deploy to the Amazon Alexa Ecosystem.
Building an Alexa skill, like all customer experience projects, starts with truly understanding your customer and their needs. It’s important to comprehend how your customer’s want to and will use your skill. Doing so will have a direct impact on the brand voice and role for Alexa. Here are some questions you should look to answer as part of the Discovery Phase of the project:
After your customer needs have been determined, creating use cases for how your Alexa skill can fulfill those determined needs or pain points is the next step. Use cases should be clear and concise. From your use cases, determine what questions a user might have and how Alexa can provide suggestions, guidance, or more information – and how to position the communication prompts so that Alexa can answer those questions.
As you build out your use cases and scenarios, we suggest you start by defining your MVP (minimum viable product). This is the minimum viable product you can take to market and provide value to the customer. Before creating a huge, multi-faceted skill – build your MVP skill. Using a lean methodology-based approach minimizes risk and investment while allowing your organization to learn and tweak the product in future iterations. A few points to keep in mind:
Remember the principles we discussed in our second VUI blog entry? It all comes down to not overwhelming the user and adhering to tried and true design principles. One of the important considerations with voice design is to be sure to account for user intent and NLU (natural language understanding). This is how the computer processes recognized words and extracts the user’s intent or meaning from them2. The tricky part with VUI design is that there are often many different ways a user can verbalize the same intent. For example, purchasing an item, a user can say “add X to my cart”, “I need to buy X”, “I’m out of X and need more”. These intents need to be clearly thought out and depicted in the dialogue flow diagram.
Rather than creating wireframes and visual design mock-ups, VUI designers create flow diagrams – or sample dialogue pathways that can occur between a user and the system to achieve the goal set out in the use case. At Paragon, we call these Interaction Models. Below are a few tips for building Interaction Models that are inclusive and will serve as your guide for development:
Don’t forget that typical GUI design tips are still relevant: Make it easy. Make it clear. Don’t assume.
Take a human-centered approach in all of your design work – including voice. And that means soliciting and incorporating user feedback throughout the design and build process. Iterative, design sprint techniques are helpful to gather user-feedback early and often throughout the process, allowing the flexibility to incorporate such feedback during design while minimizing risk.
It’s best to observe users interacting with the Alexa skill, how they complete tasks, any pitfalls, and ask follow-up questions along the way. Usability testing the proof of concept via design sprints will save time and money in the long run. Once you’re in production, don’t forget to check your logs regularly to see how you can improve your skill to best meet your customer’s needs.
Interested in learning how to integrate Alexa into your marketing plan? Let’s chat.