AlexaDev Tuesday: Which Alexa Skill Should You Create?

Publication Date: 2/27/17 – Alexa skills can be written in any programming language that can parse, and output to, JSON. So choosing your coding language is easy. Just go with the one you’re most comfortable and experienced with. The more difficult part is deciding: which Alexa skill should you create?


God Kills A Kitten


Here are some questions that should help you zero in on skill concepts that are most likely to be successful, and more importantly, help you avoid going down a development path you’ll later regret.


1. Are there many other skills like it already available?
Do some keyword searches in the Alexa app or on Amazon in the Alexa Skills Store (UK readers click here) to see if you can find anything similar to what you have in mind. Finding one or more hits doesn’t necessarily mean you should give up, but if you find a half dozen or more that are a close match you should probably move on to your next idea. It will be difficult for your skill to stand out from the crowd.

It doesn’t really matter how those other skills are doing with reviews, either. If most are being positively reviewed, what reason will the user have to choose yours instead? Alternatively, if most are being panned, your skill will be judged negatively by association.


2. Is it a good fit for the platform?
Alexa and other voice-activated digital assistants have been around long enough now that the novelty of literally talking to a software program has worn off. Now, users want and expect voice applications to either provide a type of experience they can’t get elsewhere, or perform a useful function that truly makes their lives easier—and can’t already be performed equally well by some other device they already own.

There’s nothing Alexa can do that some other device and/or app can’t do. However, there are plenty of things Alexa does better, or more conveniently. For example, controlling smart lights. Sure, they can be controlled with a mobile app on a phone or tablet, but the convenience of voice control with no need to find and turn on a device is something the app can’t match.

So getting back to your skill idea…ask yourself: if the thing your skill will do can already be done by a different device, mobile app or computer program, will moving that function to a voice-controlled platform substantially improve the user experience?

Use existing, native Alexa functions as touchpoints for your thought process. Think about how Alexa has improved the music playback, alarm clock, timer, dictionary, and handheld calculator experience for the user, for example. Will your skill offer a similar improvement? If there are areas of native functionality where you feel Alexa currently falls short, will your skill help to bridge that gap?


3. Does it solve a problem for the user, or improve the user’s life in some way?
This sounds like a difficult bar to clear, but don’t overthink it. Some of the most successful and best-reviewed skills are the simplest.

For example, my most popular skill is Focus Word. In all honesty I thought of it as a throwaway at the time I wrote it, something simple to get me familiarized with the Alexa Skills Kit and learn the broader AVS system architecture. But Amazon’s engineers immediately predicted it would be a hit, and a lot of users have made it part of their daily routine.

It’s basically a daily affirmation engine. Lots of people want to make daily affirmations a part of their lives, but having to open an app or look something up in a book to get the day’s affirmation introduces minor obstacles and delays, making it less likely the person will follow through. When all you have to do is ask Alexa to read the affirmation, it becomes much easier to form this habit.

Anything that saves the user time or steps can make for a worthwhile skill. Even if the functionality you want to deliver is already available elsewhere, ask yourself if delivering it via Alexa will make it faster, easier, or more convenient for the user.


4. If it’s a just-for-fun type of skill, is it truly novel or surprising, or does it have the kind of wow factor that will make users want to talk about it and show it off?
Say what you will about the infamous fart skill, it was hugely popular right out of the gate because it had wow factor. With it, people were able to make a digital voice assistant break wind, and they thought that was hilarious.

If you want to write a trivia or factoid skill, it needs to be far and away more interesting, useful or funny than all the skills of those types already out there to get any traction—at least, to its target audience. With those types of skills it’s smarter to shoot for a niche audience than trying to please all of the people: choose a lesser-known subject that a smaller group of users is very passionate about.

For example, even though Star Trek is a hugely popular franchise, as of this writing there’s only one Klingon skill out there right now, Klingon Dictionary, and it’s rated 5/5 stars.


Hopefully, these questions have got you thinking about the types of skills that are most likely to be successful. You don’t have to reinvent the wheel, or realistically simulate Jarvis. Just remember to think in terms of the user experience, and you’ll be on the right track.