Atualizado: 6 de out. de 2020
Different from Apps or websites that can be easily searched in the web using google, voice apps are tied to the inherent premise of ambient computing. If the premise is that you do not need to rely on one hundred apps to each one perform one small delimited task, you need to allow for the agent to find on behalf of the user what is the best app to solve his problem.
This is not only important for the user, to have his request fulfilled but important for companies that want to stay relevant in the voice space.
Amazon's approach to this can be divided into 2 major initiatives, launched in May 2018 and June 2020. Both of them relates to the same topic, Name Free interactions. On the first release, an interface called CanFulfilIntentRequest was released to skill developers so they could indicate in their interaction model which one of their intents could be fulfilled by user's requests. This works by guessing how a user would utter Alexa to solve a specific problem that could be tackled by your skill. After giving the appropriate information, your skill would now be eligible to be a candidate every time Alexa would need to query their database of (100.000) skills to fulfil a user's need.
The second and most recent effort from the Alexa team is called Name Free Interactions Toolkit (NFI) and was released earlier this year in June. It relies on the user's ability to create five suggested launch phrases and let Alexa derive others through machine learning. The eligibility formula is not shared with the users but it can rely on aspects other than the utterance, like the time the user stays in your skill after reaching it through a name free interaction. This second revised approach to discoverability is something to be used alongside CanFullfilIntentRequests.
You cannot pay for any of these spots which, at least for now, makes the market a bit more meritocratic at this point
Unfortunately, both of those initiatives are only available outside Anglophonic markets. If you develop in any other language you are stuck with:
A 50 character-long skill title;
160 character of short description;
1600 characters for the long description;
This is what Alexa takes in consideration to discover and recommend your skill so I advise you to do your homework and try to be the more descriptive possible during your submission process.
Alexa may help you increase your discoverability if you do a pretty good job of driving customer activation and retention through what they call "Editorials" - shortcuts to activate the skills other than their invocation names, "Spotlights" which are the newsletter sent to all Alexa users weekly, "Trending Topics" the space in the Echo Show Screen and a "Feature Skill" banner on the Skill Store. You cannot pay for any of these spots which, at least for now, makes the market a bit more meritocratic at this point.
All other traditional digital and non-digital marketing channels are valuable resources to explore when aiming to drive traffic to your skill so be sure to resort to that if you want to increase discoverability for now.