Voice adds a new dimension to how customers interact with your service, brand, or content. A customer’s first interaction with your skill will leave a lasting impression, which is why it’s important to ensure your welcome experience is positive and memorable. Also, providing a guided experience is vital for both new and repeat customers.

Here we show you how a great welcome experience can engage customers from the beginning and how providing helpful prompts reduces friction as they use your skill.

Create a Helpful and Delightful Welcome Experience

Imagine you’re launching a new app for the first time. You’re welcomed with a getting started message, like a slide show or a video, that explains how you can interact with the features of the app. A great example of this is the getting started video that customers can watch when they set up a new Echo device.

The welcome experience is especially important for Alexa skills because it’s all via voice –there are no visual cues, like buttons, to guide the customer. Skills need to welcome the customer and tell them what to do to get started with the skill. In addition to helping new customers have a positive experience, the welcome message is your first opportunity to establish your brand identity on Alexa and create a memorable first impression.

For example, rather than providing an open-ended response after a new customer opens your skill, provide more details about what your skill can do.

Do:

Customer: Alexa, Open MySkill.

Alexa: Welcome to MySkill. Looks like this is your first visit. I’ll explain a couple things you can do and keep in mind, you can ask for ‘help’ any time for assistance. MySkill can provide info on best practices, suggestions on how to do things and much more. What would you like to do?

Don’t:

Customer: Alexa, Open MySkill.

Alexa: What would you like to do?

A welcome prompt (something as simple as “Welcome back to MySkill”) reinforces to the customer that they are in a skill experience, and that they correctly invoked the desired skill. Customers may not realize they invoked a skill and a generic question like “What would you like to do?” could cause confusion.

Do:

Customer: Alexa, Open MySkill.

Alexa: Welcome back to MySkill. It’s good to see you again. What would you like to do this visit?

Don’t:

Customer: Alexa, Open MySkill.

Alexa: What would you like to do?

Customer: Book a Table.

Alexa: Sorry, I can’t do that. Would you like to book a flight to Seattle?

Customer: Huh?

Make sure to welcome new customers by letting them know what actions they can take and always provide a short welcome (back) prompt every time a customer invokes your skill.

Prompt the Customer with Helpful Questions

Customers interact with Alexa without looking at their device. Alexa is a voice-first experience and you can’t rely on a customer looking for a visual indication that the skill is waiting for a response, such as a light ring or a screen display. It’s important to use non-rhetorical questions to indicate to the customer that it’s their turn to reply. If the customer doesn’t reply, the skill will end abruptly and the customer may never try out that skill again.

Imagine calling your mobile (or cable) provider. You hear a welcome message that goes something like, “Thanks for calling MyMobile. Please listen carefully to the menu because our menu options have changed. For new services, say new services, for billing questions, say billing, to change or modify your existing service, say modify, for hardware issues, say hardware, for all other inquires, say operator, to hear these options again, say repeat.”

What should you do now? Is it waiting for you or are there more options? It did say to listen to all options before making a choice… Then as you’re wondering if there are more options, you hear, “Thanks for calling, contact us again soon! Goodbye.”

Adding a simple contextual question, like “What would you like to do?” to the end of the menu options would alleviate all that confusion. For example:

Do:

Alexa: Welcome to MyPizza. What would you like to do?

Customer: Order a pizza.

Alexa: Ok. Pizzas come in 3 sizes, Small, Medium and Large. There are different crusts to choose from, Thin, Hand Tossed, Deep Dish, Gluten Free. What size and crust would you like for your pizza?

Don’t:

Alexa: Welcome to MyPizza. What would you like to do?

Customer: Order a pizza.

Alexa: Ok. Pizzas come in 3 sizes, Small, Medium and Large. There are different crusts to choose from, Thin, Hand Tossed, Deep Dish, Gluten Free.

Handle Errors Gracefully with Reprompting

It’s also important to gracefully handle errors and guide customers back to your skill’s functionality. For example, when a customer asks to do something that your skill doesn’t support yet, acknowledge that in your response. Then reprompt the customer with the last question they were asked. For example:

Do:

Alexa: Welcome to MyPizza. What would you like to do?

Customer: Book a Table.

Alexa: I’m sorry, MyPizza can’t help with that yet. MyPizza skill can place orders for delivery or pick up. Which would you like to do?

Don’t:

Alexa: Welcome to MyPizza. What would you like to do?

Customer: Book a Table.

Alexa: Sorry, I didn’t understand that. What would like to do?

Customer: Book a Table.

Alexa: Sorry, I didn’t understand that. What would like to do?

Customer: Book a Table.

Be sure to also acknowledge customers saying something that isn’t intelligible to the skill by saying something like, “Sorry, I didn’t get that.” Reprompt the customer with the last question they were asked, instead of using a generic reply. In these cases, avoid telling the customer you didn’t hear or didn’t understand them, as this encourages customers to repeat their previous utterance more slowly and/or at a greater volume, rather than rewording the request. For example:

Do:

Alexa: What size would you like?

Customer: Banana.

Alexa: Sorry, I didn’t get that. What size would you like?

Customer: Banana.

Alexa: Sorry, I didn’t get that. We have Small, Medium and Large. What size would you like?

Customer: Small.

Don’t:

Alexa: What size would you like?

Customer: Banana.

Alexa: I didn’t understand you, please say it again.

Customer: Banana.

Alexa: I didn’t understand you, please say it again.

…

It’s also a nice touch to acknowledge the request of something not supported by the skill and to provide options that are. For example:

Do:

Customer: Alexa, I’d like to order a Pizza.

Alexa: Ok, what size?

Customer: Super size.

Alexa: I’m sorry, we don’t have a Super size, our sizes are Small, Medium or Large. Which size would you like?

Customer: Large.

Don’t:

Customer: Alexa, I’d like to order a Pizza.

Alexa: Ok, what size?

Customer: Super size.

Alexa: What size?

Customer: Super size.

Alexa: What size?

…

Last, make sure you provide questions with options the customers can clearly reply back with so they know they need to interact with the skill without any visual cues to do so. Remember to prepare for requests your skill can’t handle by gracefully handling errors and redirecting the customer to what your skill can do. These two methods will make your skill easy to use and provide a consistent experience.

More Resources

• Amazon Alexa Design Guide
• Supportive Prompting Test Cases
• Alexa Skill Certification Checklist
• Alexa Developer Forums

Make Money by Creating Engaging Skills Customers Love

You can make money through Alexa skills using in-skill purchasing or Amazon Pay for Alexa Skills. You can also make money for eligible skills that drive some of the highest customer engagement with Alexa Developer Rewards. Download our guide to learn which product best meets your needs.

Build a New Skill, Earn Developer Perks

Bring your big idea to life with Alexa and earn perks through our tiered rewards system. Publish a new skill in June and earn an AWS IoT button. Add in-skill purchasing to any published skill and earn an Alexa-enabled device for the car. Learn more about our promotion and start building today.