In our third post, we looked at some creative executions and attempts at getting the word out. In this post, we sum up some of the problems and opportunities with ASK.
Development Process Pain Points
I mentioned in our second post that troubleshooting a skill was time consuming and tough. Here are some reasons why:
Local ASK development environments
I didn’t have a local environment, so I would make blind changes, upload, and test. Not super efficient. Turns out, some people have done work on local development environments. Good to know for next time.
Pushing to Staging
You have to manually upload a zip to AWS Lambda after each change. I’d love a simple command line interface similar to ‘git push heroku master’ for AWS Lambda functions. Maybe there is one and I didn’t see it.
Debugging Tools on Staging and the Device
The interface in the developer portal and the Echo for staging testing really just tells you if there’s an error. There is very little in regards to diagnostic. This makes a local environment even more important.
Assign Staging Lambdas
One issue arose after the skill was published. The only place I had been uploading the code was now live and public facing. To test any future changes, we’ll have to setup a separate staging skill. Another reason why a CLI for zipping and uploading to Lambdas would be helpful.
Platform Discoverability Issues
Discoverability for standalone skills is tough. Amazon has over 15,000 skills in their store. The most successful skill types out there are flash briefings from existing news organizations, or existing services or products providing another customer interface — Spotify, Dominos, Uber, etc.
One root of this problem is that Amazon doesn’t appear to be trying yet. That, or they’re so buried in the volume of skills being submitted that they’re struggling to keep up with things more serious app developers would like to see implemented, for example, monetization options.
- Their advertisements are extremely powerful for showing off how simple the device is to use, and have played a huge role in the adoption of a brand new technology. BUT. They don’t show off that you can add skills, or that there is a skill store.
- Within the skill store there does not appear to be a big push in categorizing, curating, or creating staff/editor favorites.
- Apple’s app store policy includes language that allows them to reject applications due to unoriginality or poor design. Amazon’s policy does not (both share some similar common sense ethical and safety baselines).
- The skill store has no priority placement in the Alexa companion mobile app which is where you initially set up the device.
- Within the Alexa companion mobile app, there is more emphasis on built-in features like messaging across devices.
The people in the driver’s seat for encouraging certain user behaviors to the masses is Amazon’s advertising teams, followed by their skill store team.
Another problem is that there aren’t any built-in features in the hardware that encourage communication among users. Discoverability is easier when there is word of mouth behavior to piggy back off of, such as texting & screenshots on phones.
- There’s potential in the devices new messaging and voice features, “Alexa, add Sarah to the Domino’s order and tell her when the pizza is en route.”
- There’s potential in the UI around the mobile app, the Echo Show, and visual cards that skills generate, but this also feels like a step backwards in regards to interface friction.
- There’s potential within skills themselves to prompt the user to recommend it to a friend, similar to mobile app ranking prompts.
- Another interesting area of potential, when you ask Alexa something she does not know how to handle with built-in features, she will prompt you with a skill that may be useful for it. This could pair well with aggressive advertising of non-built in skills to help encourage users to naturally explore phrases with less ‘fear’ of a negative response.
- Even then, all these would just be guesses until there are more robust ways to measure the effectiveness of them. In addition to no skill store landing page metrics, it doesn’t look like there’s a way to get any analytics around card interactions or if a user responded to a prompt.
This all will just be a matter of time. Even if Amazon doesn’t solve them, every one of these issues has some workaround waiting to be figured out.
Wrapping Up
Parting ASK Platform Thoughts
The platform owner sets the tone for new devices adoption.
Apple ran this to perfection for touch screens and mobile devices. Now Amazon is leading voice. Amazon is off to a good start on the basics. Soon, they’ll likely benefit from showing off the skill store and highlighting developers.
The other thing the platform owner decides is how much they want to share with developers and how much they want to keep for themselves.
There isn’t enough built-in functionality in ASK yet to really facilitate word of mouth growth, or take payments if we wanted to enable paid acquisition.
Maybe Amazon is holding onto some features too tightly, maybe they just haven’t gotten there yet.
Parting Background Baseball Thoughts
Data is the big restriction we have right now (ok…time and money…but we’re happy to open source) in regards to the feature set. It would be incredible if more of these broadcasts were opened up by the people who hold the rights to them.
Until then, we’re limited to the data from what’s hosted on the public domain archive. With more, we could build in the ability to request a game by decade, announcer, player name, or only games where your team wins.
One other thought, had our data honed in on one particular era, say the 60s and 70s, we could have taken the branding in an entirely different area. Just think no-nos on LSD.
Given that the skill itself is about mood setting and the creative sets the tone for how to use it, this doesn’t seem like a small thing. Especially if the popularity of flash briefings and podcasts continue to grow and the AI assistant storefronts and devices become primary entry points for podcasting and news brands.
Background Baseball is an Amazon Echo skill that allows you to stream classic baseball games from the 1930s-70s.
If you have an Echo, you can enable it in Amazon’s App Store.
If you like it, do us a solid and do one or more of the following:
1. Leave a review on Amazon
2. Give us an upvote on Product Hunt
3. Share it with friends
We aren’t making money off this, but having people using it and enjoying it make us feel good. Thanks for the help spreading the word.
