Mind the gap

Voice-first platforms have reached an inflection point.

In digital media, there is often a gap between what is popular and what is profitable. Because tech platforms can subsidize usage, either through venture capital or profits from a “cash cow” business unit, mass adoption often comes before meaningful monetization. The best time to invest often lies on the way to mass adoption, but before meaningful monetization. We saw this play out in search, social networks, mobile, and streaming video. Now, we’re going to see it again in two new fields: voice and eSports. I’ll talk about voice this week, and maybe get to eSports next week.

44 million Americans used a smart speaker in 2017, and eMarketer projects penetration will reach 77 million by 2020. Nearly 2/3rds of these devices will be enabled by Amazon Alexa. According to Comscore, over 20% of U.S. households now have a smart speaker, with the fastest growth occuring in households with very young, or very old users. This latter statistic speaks to the new market opportunity for voice as a platform: to enable those who don’t currently read or have their own computing device (the young), or those who aren’t yet fluent with conventional technology (the old). BMW, Ford, Toyota, and VW are beginning to bring Alexa natively to new lines of cars in 2019. The touchpoints for voice as a platform have broken through a clear barrier of adoption and are on their way to pervasiveness.

Monetization for voice platforms, however, lags adoption. The Information reported that only 2% of Alexa users have made a purchase through the platform. Of those users, only 10% have made a second purchase. Other studies report billions of sales from voice commerce, but engagement statistics like these make me doubtful. Voice purchasing is likely less than a billion in sales today and would need to grow engagement significantly to make a dent in Amazon’s top line.

Advertising could also drive voice platforms to new levels of monetization, but I remain skeptical of the efficacy of ads in the format Alexa presents today. A user asking Alexa a question is looking for a specific item or answer. The faster the result is returned, the better. Pre-, mid-, and post-roll ad formats don’t work with voice. An ad product like Google SEM doesn’t work in voice because no screen real estate exists to redirect the user’s attention to an ad unit. Same for most app install ad formats, like you’d find on Facebook or in the app stores. The best opportunity for ads in Alexa may be upselling subscriptions within a free service you already use, or perhaps promoting other skills. In any case, none of these formats make money today, nor will they until voice apps can sustain engagement comparable to popular mobile apps.

Podcasts and voice media are at a similar stage in their evolution. A longitudinal study by Edison Research in 2017 found over 60% awareness of the term “podcasting” (up from 46% in 2012). 40% of those surveyed had ever listened to a podcast, 24% were monthly podcast listeners (up from 14% in 2012), and 15% were weekly podcast listeners (up from <7% in 2012). That latter number implies 42 million weekly podcast listeners domestically, or only 19% of the 229 million Americans who listen to AM/FM radio each week.

Podcast listeners have some favorable demographics compared to the overall population. They skew younger (51% aged <34), more wealthy (45% over $75K in HHI), more college educated (57% college or higher), and more employed (63% full-time). They also tend to use social media at a far higher rate than the general population and are twice as likely as the general population to own an Amazon Alexa device. While Americans as a whole spend only 2% of their audio time on podcasts, podcast listeners spend 30% of their audio time in this new medium. The compensating reduction in “share of ear” for these podcast listeners comes out of AM/FM radio, which is cut in half to accommodate podcast listening and a bit more streaming audio, which comes in second place to podcasts at 19% “share of ear.”

Like voice platforms, monetization is also early in voice media. IAB/PWC expected ad revenue in podcasts to be $314 million in 2017 and projected that number to double by 2020. 2/3rds of ads were host-read, demonstrating the importance of native formats in this medium. In comparison, classic radio ad revenue is flat year over year. 2017 saw $17 billion of ad dollars spent in that format, with only 8% digital ads. If we extrapolated the above statistics, we’d likely see an accelerating decline in AM/FM radio minutes and ad revenue, since podcasts are a meaningful substitute by younger, wealthier, social media savvy listeners. While still a sub billion dollar market today, podcast and voice media ads could overtake older formats in the coming years.

Voice is on the cusp of something big, even though monetization is admittedly early. Both voice platforms and media have reached 20% user penetration domestically. Both are sub-billion dollar addressable markets today, but are growing quickly. It feels like a good time to invest heavily in what could become the next multibillion dollar consumer media market.

I originally published this post in my weekly-ish newsletter “Drinking from the Firehose.” Sign up below to receive columns like this (and more) in your inbox: www.tinyletter.com/ataussig