Things We Are Learning About Smart Speakers

Technological change in the commercial radio business has always seemed slow and a bit begrudging. Whether it be streaming, HD, smartphones or podcasting, the business has not experienced the ferocious erosion of other media which has forced rapid change and innovation. Think print. So, it is remarkable to look at radio’s prompt interest in smart speakers and not be somewhat amazed at how fast the business is embracing these devices.

Certainly, a large part of the interest is getting radios back into homes; a key area that has eroded significantly over the years. And it certainly makes sense for radio to be on smart speaker devices, since music, and audio in general, top the list of uses. From that standpoint, it doesn’t feel like a big digital reach to get a station onto smart speakers. In fact, TuneIn and iHeart have done it for most radio stations by default.

However, based on what we are seeing, just "checking the box" by getting your station onto smart speakers is not likely to move the needle.

Back in April, my company Amplifi Media, joined forces with Jacobs Media/Jacapps to create Sonic AI, a business that develops “skills” for smart speakers. We work with radio stations and podcasters to create great user experiences on these devices, and with some 20,000 “skills” already in existence for Amazon’s Alexa, fighting for attention and repeat use on smart speakers is quickly becoming more critical.

Here are some observations we have gleaned from our own work:

Having the right invocation name is key – There is still great confusion among potential users in accessing stations by name, especially if the name is common such as Mix, Star, or Lite. We have seen a lot of stations make mistakes by developing non-intuitive invocations. One thing we know, is that searing language and ideas into listeners minds takes clarity and time. An invocation that seems logical and simple to recall tops the list.

Although smart speakers are hot, almost 90% of people don’t have one yet – Education, teaching and familiarization are important. We council stations on exactly how to talk about smart speakers on the air and present on mobile and web. People are becoming more comfortable with voice command devices (Siri, Cortana, Google Home etc), but there is still a learning curve.

The Amazon “dial” is more competitive than the local radio dial – We have conversations with stations assuming they will see an immediate “pop” with listening once their station's skill is up and running, but with 100,000 radio stations, 400,000 podcasts and 5 large streaming services, many stations have been disappointed that their station streams aren’t getting lots of traffic. Hey, remember, radio stations haven’t exactly taken off in the streaming world - compared to "pure-plays" - and a lot of that has to do with generic content, bloated commercial loads and plenty of "pure-play" competition. Some radio stations streams still suffer from technical issues and Smokey The Bear PSA's.

Thinking beyond the stream – We are content guys. Since we pooled our resources, our common thinking has been informed by what we have learned with our own digital initiatives – Jacobs/Jacapps with deep learnings about what content works on mobile devices and Amplifi on what “prints” in podcasting. In each case, it is significantly different from over-the-air AM/FM.

Our contention from the start has been that a station’s stream may not be the big win on these devices

**Entercom KISW, Seattle morning show BJ & Migs skill is part of the overall "Open 99.9 KISW" skill**

Our contention from the start has been that a station’s stream may not be the big win on these devices, so we have focused our resources on creating and developing engaging listener experiences that are easy for listeners to follow and highlight top featured segments of station audio. We use custom audio from talent to host “skills” and enable client stations to feature top benchmark content, podcasts and original “Alexa-only” moments.

Skill design can be brutal – Data we are seeing shows that only 3% of the 20,000 skills developed for all brands are used more than once. Sure there is the novelty of playing around with these new devices, but who really needs to go to “Taco Facts, or "Nicolas Cage Trivia" more than once. By the way – CNET has a fun list of "30 Alexa skills nobody needs."

We have seen the “Mcdonalds-ization” of "skills" where you can buy one for $59 dollars. If only everything were that simple.

Get more quarter hour credit – Picking the right content can result in PPM credit. It is “early innings” and we can’t say yet stations are blowing off the doors with increased quarter-hours, but we are seeing benchmark features out-pull full morning show podcasts and the same appears to hold true with Alexa skills. People seem to gravitate to bite-size content such as “The Bizzare Files” from the skill we developed for Beasley's WMMR, Philadelphia. It's the top feature of the "Preston and Steve" morning show.

Our learnings are rapid.

We have more we are eager to share. One thing we know for sure is that simply being “on” a smart speaker isn’t going to be enough for most radio stations and repurposing the stream merely places a station in a more crowded sea of choices than the local radio dial.

More about Sonic Ai here