Over the last few years, third-party skill builders have turned to In-Skill Purchasing (ISP) to generate revenue and build sustainable businesses with Alexa. ISP allows developers to sell premium content for a fee in custom skills. Customers pay for premium content by using the secure payment options associated with their Amazon account.
When Todd Moore developed the White Noise Alexa Skill, the rapid surge in the skill's popularity resulted in a sharp increase in hosting and maintenance fees. Moore turned to ISP to cover these increased costs and support the development of new features. As a result, White Noise continues to bring in more users, generate revenue, and provide an even better Alexa experience to its customers.
Moore has been coding since he was 10 years old. He holds a computer science degree and began coding smartphone apps in 2008. His app White Noise quickly reached the number one spot on Apple's App Store and remains popular to this day.
Following this early success, Moore founded TMSOFT to work full-time on software development. He was an early adopter of voice technology.
"I saw Alexa and the Amazon Echo devices taking off and was immediately excited to build experiences that would help shape the future of ambient computing," says Moore.
In April 2018, Moore built the skill that can play continuous color noises such as white and brown noise, nature sounds, and other ambient sounds that people can use to help them focus, relax, and sleep. With the Alexa White Noise skill, users can play up to ten hours of uninterrupted audio.
"You'll never hear an audio gap during playback, which can wake you up if you're asleep," says Moore. "That's what makes this skill stand out."
As the user base for White Noise grew, TMSOFT's hosting and streaming costs for the skill's audio skyrocketed. Moore realized that developing in the cloud required a different approach from building smartphone apps.
"Alexa is a cloud-based service, and the White Noise skill was our first venture into having a fully cloud-based audio streaming solution," says Moore.
In March 2019, Moore learned about ISP and decided to monetize the White Noise skill to cover costs and facilitate ongoing development.
Adopting ISP to support growth and development
TMSOFT enabled ISP a little over a year after launching the skill. The skill is still free to install and streams up to 50 hours of audio. After 50 hours, the skill prompts customers to sign up for a subscription to continue using the sounds and unlock premium audio quality.
The move to a subscription-based model has proven successful for TMSOFT. According to TMSOFT, the company's conversion rate from free to paid users on Alexa is about 22 percent. And the skill has at least doubled its revenue each year since 2019. Furthermore, the Alexa skill represents 18 percent of all revenue for TMSOFT, and that percentage continues to increase.
"I'm pleasantly surprised at the conversion rates and the number of users that are willing to upgrade to a premium subscription," says Moore. "It's easy to continuously support, upgrade, and add new features and sounds when you have a customer base that's willing to throw you a couple of bucks a month."
Tips to grow revenue with In-skill purchasing
Below, Moore provides four tips to developers looking to grow their revenue with ISP.
Tip #1: Provide users with the full experience first – free of cost
Moore recommends that developers should let users fully experience their skill before upselling premium content.
"Give away a lot for free and get people hooked," says Moore. "You want customers to fall in love with your skill initially, so you don't want to limit it too much."
When users appreciate the core functionality of your skill, you can then upsell them with enhanced features. Because they are bought into the experience, they are more likely to convert to the paid feature.
White Noise offers its full catalog of 20 sounds for free for 50 hours. That way, people can see it's worth the price.
Tip #2: Be transparent about the reason for your ISP upsell
Moore recommends that developers be transparent about what they are charging their users for and why.
"Explain to your user why you're charging them money," says Moore.
His original upsell message in his skill was – "Streaming uninterrupted audio from our servers for up to 10 hours at a time is costing us thousands of dollars. We had to make a tough decision to either call it quits or ask for help in funding this service. We are not quitters. We are betting that you are willing to help. After 50 hours of free streaming, you'll be asked to purchase a monthly premium subscription. The pricing for the monthly subscription is inexpensive compared to other streaming services. Premium subscribers also receive higher quality audio. Thanks so much for your support! To get pricing, say, 'Purchase Subscription'. If you would like additional details, say 'Subscription Help'."
In addition to being clear about what his customers will get with a premium subscription, Moore explained how additional income from this subscription enables him to offer customers new features and great service.
Moore also picked the lowest possible price he could charge. "We have to charge something. We picked the lowest price we could charge, too, and I think that helped."
Tip #3: Improve skill discoverability
Only if a skill has traffic can it be monetized, so developers need to focus on solving skill discoverability first. Moore suggests that developers leverage Alexa features such as Name-free Interactions (NFI) and CanFulfillIntentRequest to help customers find their skill organically. When a customer makes a request without mentioning the skill name, Alexa looks for skills that might fulfill that request and determines the best choice among eligible skills. To enable customers to launch a skill without knowing or having to remember its name or invocation, developers should consider adding the Name-Free Interactions (NFI) container to their skill and provide a signal for Alexa to consider their skill when routing name-free requests.
“When we added multiple launch phrases and the CanFulfillIntentRequest intent, revenue for White Noise really started growing,” says Moore. “It greatly helps with discoverability when you can do name-free interactions.”
Tip #4: Use analytics to continuously improve your skill
Analytics plays an essential role in understanding the user base, including which phrases developers should support and what users expect from their skill. Moore recommends using analytics on the Alexa developer console to discover points of friction in the skill and smooth them out.
“There are a lot of phrases people might use to try to play sounds, and we’ve discovered those and added them to the user experience along the way,” says Moore.
Apart from analytics around utterances, Moore suggests reviewing ISP conversion metrics, such as the offer conversion rate, to periodically refine the upsell message. He also recommends keeping an eye out for customer reviews and ratings on the Amazon website to get a pulse on customer feedback.
The White Noise Alexa skill continues to grow for TMSOFT, and the company plans to release a new version that will support multimodal experiences in the near future.
“Monetization works on Alexa,” says Moore. “ISP has made it possible to continuously upgrade White Noise with new features and sounds, and I’ve enjoyed working with this new business model.”
Get started with ISP by reviewing this technical document and design an optimal ISP experience by reviewing the Alexa Design Guide.