Aaron Master was with SoundHound, Inc from 2006 until early 2013, serving as Director of User Experience, iOS Product Manager and Research Engineer. The views expressed in this article are his own and do not necessarily represent the views of his current or former employer. Disclosure: Dr. Master holds investments in SoundHound, Amazon and Apple.
In December 2017, Apple announced it was acquiring Shazam, a music identification service, for $400 million. Verto Analytics had the opportunity to speak with digital music industry veteran, Aaron Master, to discuss the potential implications of Apple’s latest acquisition, and what it means for the highly competitive streaming music market.
Q: Were you surprised that Apple acquired Shazam, and why do you think they did so?
A: No – in fact, I’m surprised it took this long. Apple has had a long and fruitful partnership with Shazam: for example, in the early days of the App Store, Apple used Shazam to sell iPhones through its “there’s an app for that” ad campaign during the 2009 Super Bowl. The Shazam app has regularly been given banner status on the App Store since then, to make sure iPhone customers could use a brand they’d recognize to identify a song. Shazam has also been integrated into Siri since 2014, providing answers if you ask “what’s this song?” (And it’s been integrated into iMessage since 2017, for some reason.) These partnerships provided increasing brand awareness for Shazam, while providing Apple with an opportunity to sell identified songs on iTunes or as part of an Apple Music subscription.
However, Apple might want more control over the terms of the relationship. A few theories have surfaced about what Apple wanted from Shazam that it couldn’t already get through its existing partnership: (1) technology; (2) data including music recommendations; or (3) referrals from the Shazam brand. Right off the bat, I thought “it’s about the referrals,” which are the links in an app that send traffic to a partner (e.g. Apple Music or Spotify). Acting on its own, Shazam could send that traffic wherever it pleased. Apple, of course, wants that traffic for Apple Music; Apple is trying to compete in services but still lags Spotify by a factor of two in paid subscribers. Shazam is one of the biggest sources of referrals for newly discovered music, including on platforms like Android that don’t offer Siri.
For context, Spotify captured a big lead in the U.S. market after it launched in 2008 and now has 60 million paid subscribers globally. Other rivals include Amazon, which currently has 90 million Amazon prime subscribers and 180 million users in the U.S. alone, some of whom contribute to its 16 million music subscribers that an industry source reported this past summer. Apple has employed a number of tactics in its attempts to catch up, including traditional marketing and some controversial changes to the iTunes (now Apple Music) user experience.
Q: If the acquisition is about referrals, how will this manifest in terms of the user experience? How will Shazam be integrated in Apple products, from an app and hardware standpoint? Will Apple shut down the Shazam app?
A: Let’s start with the basics. Apple wants to use Shazam as another pathway for listeners to join Apple Music, which means they want to create a bigger presence for Shazam as well as a bigger presence for Apple Music in Shazam. On a practical level, we could see new iPhones ship with Shazam preloaded onto the devices, Shazam auto-included with an iOS update, or further promotion (or even automatic activation) of Shazam inside Siri. The Shazam app itself might also see a redesign that reshapes it into one big marketing funnel for Apple Music, even on non-iOS platforms (Apple Music has been available on Android since 2015).
Where things get interesting is how Apple Music will choose to handle competition, both for Shazam and Apple Music. Judging by the online reviews (and because I cannot provide confidential information), lots of customers appear to use Shazam or SoundHound to auto-generate Spotify playlists. If Apple commits to a blanket “kill the non-Apple Music links” strategy, it could alienate existing Shazam users and drive them to SoundHound or another competitor. There is also a chance Apple could alienate Shazam’s (now Apple’s) content partners, whose cooperation ensures that their database is current, accurate, and able to compete with SoundHound. If Apple makes it inconvenient to use competing services like SoundHound, there could be a backlash against the iOS platform, similar to what happened with the Apple Maps fiasco.
A likely scenario is that Apple applies the link takeover selectively, only funneling new customers to Apple Music while letting existing customers (or those who dig into settings) link to Spotify, Pandora, or whatever service they prefer. It depends on whether Apple thinks there is more growth to be had from new streaming subscribers or existing ones. If the latter, they will also want to do a better job helping users move Spotify or other playlists to Apple Music, which currently requires third party apps. Even though Apple is known for shutting down and integrating companies it acquires, it is very unlikely that Apple will shut down Shazam. I mentioned referrals above, and those referrals only exist if Shazam exists. This is especially relevant on the huge Android platform, where Shazam may be Apple’s strongest presence and Siri does not exist. Apple is buying Shazam’s brand, mindshare, and customers (Shazam even has its own TV show). I expect Apple will leverage all these things to its own advantage.
Q: So it’s not about the core music recognition technology?
A: The short answer is “not exactly.” Shazam’s core recognition technology has really only been able to compete on speed, accuracy, or database size with audio tech leader SoundHound since about 2014. If this acquisition were about Apple wanting the best technology and being willing to pay hundreds of millions for it, they would have acquired SoundHound years ago (before it was worth a reported $830 million), or licensed SoundHound technology for Siri.
However, I don’t think this acquisition, or the earlier integration of Shazam into Siri, would have happened if Shazam had failed to improve its technology. Referrals and preference data only exist if the technology works. While there are many companies that provide automatic content recognition (ACR) for the “low noise” case, the “environmental ACR” problem that SoundHound and Shazam solve is a harder one. Facebook launched a version of ACR that didn’t work as well and seems to have been removed from the app. Google also has made a version called “Now Playing” for the Pixel 2 that doesn’t require an internet connection, but it has only a tiny fraction of the database size of SoundHound or Shazam.
Q: Shazam’s other technology and initiatives include their AR experience, Shazam for TV, Shazam for TV ads, Shazam for print ads, auto-Shazam, and Shazam music recommendations. What about those?
A: With one significant exception, I don’t believe the acquisition is about these other initiatives either. First, there isn’t evidence that these technologies have succeeded. If these technologies were expected to increase the value of Shazam, they would have been expected to do so over the past two years, as they are mostly recent initiatives. Yet during this time, Shazam went from a paper value of $1 billion to an acquisition value of $400 million. Shazam, which has historically been quick to share usage numbers, has not posted encouraging data about use of these technologies. Personally, I have not seen a TV ad (nor a print ad) with a Shazam logo* in years.
Second, as a UX designer, I don’t see synergy between most of Shazam’s newer technology initiatives and Apple’s iOS. In fact, the availability of Siri and other assistants should be viewed as competition likely to stifle Shazam usage. Why would a user scan an image of Pepsi or tag a Pepsi ad in Shazam when he or she could just ask, “Siri, tell me more about Pepsi?” It’s worth mentioning that some experiences, like SoundHound’s Hound assistant, do benefit from OS synergy, meaning that the user experience is markedly improved if the app is accessible from the lock screen rather than being three clicks deep into a mobile interface.
Which brings me to the exception: a combination of Shazam technologies that I do think has OS synergy for Apple, even if not for consumers. Shazam has been able to detect TV shows since 2012, including live and recorded TV. It’s been able to continuously auto-Shazam content since 2013. It doesn’t take a great leap of imagination to consider that (1) the TV feature could be expanded to cover every TV show and movie in iTunes, as well as every item from Amazon, Netflix and others and (2) that the auto-Shazam feature could be built in to Siri, which is already listening for users to say “hey Siri.” I’m not sure if consumers want a list of every content item they experience on a given day, but I’m nearly certain that Apple does. Constant individual monitoring is the holy grail of media measurement and facilitates ever better ad targeting. Notably, Apple would need to convince its customers that this can be done in a way that respects their privacy.
Finally, there has been speculation that the acquisition is all about the music discovery data meaning that Apple wants to predict hits or know what songs to recommend to their subscribers. While there is indeed a lot of interesting data here from 100 million users a month doing occasional Shazaming, I don’t see it as worth much money to a company that already has 30 million streaming music subscribers. If anything, Apple may want to keep this data out of the hands of an upstart competitor. Of course, if Apple implements the always-on Shazam feature I just mentioned, the data becomes much richer and takes on more value.
*These partnerships took the form of TV ads for non-Shazam products like Pepsi (most famously in the 2012 Super Bowl), but devoted significant time and space to mentioning Shazam. These generated millions of dollars in marketing exposure for Shazam, but did it have a net benefit for partners like Pepsi? There were indications in 2015 that the answer was “no” and that still appears to be the case today.
Q: Given Shazam’s implied $1B valuation in 2015, isn’t a $400 million acquisition in 2017 a disappointment for employee shareholders and investors? If Shazam’s usage and relevant technology haven’t changed much since 2014, why didn’t the acquisition happen back then?
A: It’s a mixed bag. Who wins and loses (or “wins less”) in this deal depends on things like liquidation preferences of investors and exercise prices of employee options. Even though Shazam lost 60% of its paper value from two years ago, the $400 million sale price is much greater than the $125 million in funding it received.
As I mentioned elsewhere, the lower price likely indicates a difference in optimism – at both Apple and Shazam – about Shazam’s initiatives in 2015 versus now. That includes Shazam for TV, ads and images, as well as their referrals for song downloads. As widely reported, the download segment has been getting hit hard lately and there’s no reason to expect that referrals to downloads have been exempted from the decline or that streaming referrals have made up the slack. Consumers are also listening to more streaming radio than ever before, which reduces the need for music recognition.
This also helps address the question of why Apple didn’t buy Shazam back in 2014. At the time, Shazam had taken on investment valuing the company at $400 million, and was gearing up for an IPO at a valuation of “$1 billion or more,” a price Apple may have thought too dear.
Q: What risks does Apple face with this acquisition?
A: The risks here are relatively small. If Apple does something wrong, they can always revert to their existing practices and simply be out a manageable chunk of their cash hoard. The biggest risk is that they would botch some of the details in their referral strategy as I mentioned earlier.
Q: How will the HomePod play into Apple’s strategy? When it comes to smart speakers, Amazon’s Echo line seems to be angling for e-commerce dominance, do you think Apple is trying to make a similar play for music and/or streaming content overall? Is music going to become Apple’s market differentiator?
A: It depends on whether Apple thinks Apple Music will sell HomePods or vice versa. If the former, Apple is facing major headwinds: while they’re one of the leading music streaming services in the U.S., they don’t have a large subscriber base compared to Amazon or Spotify, meaning the HomePod wouldn’t succeed until Apple could get more subscribers. Shazam could be a big part of changing that, as I mentioned before. On the other hand, if Apple thinks the HomePod could sell streaming subscriptions, which I see as more likely, they could throw in a short-term Apple Music subscription with the HomePod (or discount an unlimited subscription as Amazon does for Echo owners) or just hope that the HomePod is so brilliant it will tempt people to subscribe to Apple Music. A less likely option is that Apple could make the HomePod version of Siri actually work with Spotify or another service. It sure doesn’t work now; just try telling your iPhone “Siri, play songs by Bruno Mars in Spotify” and watch her brusquely reject your request.
But I don’t think Apple is thinking this small with the HomePod. I expect the HomePod to serve as the Apple-Siri entry in the competition to own the interface to the internet of things, much as the Echo products are for Amazon-Alexa. I could also see the HomePod serving as a next generation Apple TV: its specs suggest it will outperform a lot of TVs’ built-in speakers, and HomePods should be able to serve as streaming hubs for video content. E-commerce is another beast, though, and Apple would have to make major moves to catch up with Google or Amazon in that arena.
There is another way that the Shazam acquisition could be relevant to the HomePod: much as Siri with auto-Shazam could monitor all content around an iPhone user, a HomePod with auto-Shazam could monitor all content in a room – or that the HomePod itself outputs. This is thorny legal territory as Vizio recently found out when it had to pay hefty fines to the FTC for failing to get “affirmative consent” to monitor content shown on its users’ TVs. But if done with a user’s permission, this would give Apple still more data about consumer habits – even when users consume content through a non-Apple app.
One last quirky Shazam-HomePod possibility deserves mention. To the technical-minded, who understand that automatic content recognition is only “really needed” for unknown items, there is little reason that a user would “use Shazam” on a HomePod, because the HomePod, like all streaming devices, generally “knows” what it is playing – be it from Pandora, Apple Music or anything else. But in my view, that isn’t necessarily how consumers think of Shazam. Consumers don’t think Shazam means “use algorithms to search a database of songs for this unknown item,” they think it means “make an electronic note that I like this song and I’m not sure I know its name.” It’s possible that the HomePod would allow users to say “Shazam this song” meaning add it to their Shazam list, even if the HomePod already “knew” the title and artist of the song.
Q: From the Beats Electronics (and Beats Music) acquisition to Shazam, will we continue to see a consolidation of competitors in the music streaming market? How does SoundCloud stand to compete with a new CEO attempting to revamp the company?
A: I’m tempted to say there is consolidation coming, but the industry constantly surprises me. People have both emotional and practical attachments to their chosen methods of music consumption. On the other hand, we’ve seen services shut down amid financial difficulties even with dedicated subscribers hanging on.
Q: We’ve been looking at the acquisition from the perspective of what Apple wants. Is it possible Apple is playing defense here, by preventing Shazam from being acquired by another company?
A: Absolutely. Just as Apple hopes Shazam will help it succeed in music streaming, data collection and with the HomePod, it also may fear that Shazam could help competitors. The first to come to mind is Google, which offers the limited music recognition product mentioned before, as well as the Google Home smart speaker, and unlimited music through both Google Music and the ubiquitous YouTube. A combined Google Music/YouTube Red service is reported to be launching this March; one wonders if early news of such a launch encouraged Apple to act now to blunt the impact. The second obvious competitor is Amazon as mentioned before. Streaming leader Spotify has the most to lose, but I’m not sure Apple would need to play keepaway there; recent financials suggest that a $400 million acquisition might have put Spotify in too vulnerable a cash position.
Q: What does Shazam’s acquisition mean for SoundHound? Are the companies competitors?
A: For several years now, based on public statements and job openings, the two companies have been charting very different courses. Shazam has doubled down on the media ID strategy while SoundHound has done a deep dive into the speech and AI space and launched Siri-killer Hound and the Houndify platform, enabling the first conversational AI. I think it would be impossible for the companies to pivot into one another’s strategies.
Nonetheless, it’s true that SoundHound and Shazam have been competing in music recognition for a while. Based on app downloads, SoundHound’s music apps are a major driver of traffic and mindshare for SoundHound as a company. Given this, the biggest impact on SoundHound will come down to Apple’s strategy for directing customers to Apple Music as discussed above. If they make the Shazam UX miserable for Spotify or Pandora users, for example, expect defections to SoundHound. I can expect that Apple will not want to encourage usage of SoundHound assuming they can’t guarantee the same referrals to Apple Music they can get through Shazam, which is a substantial risk for SoundHound on iOS.
In the bigger picture, though, the acquisition is good for the industry including SoundHound: it shows that there is substantial value in the combination of good technology and an engaged consumer. At the same time, it also suggests that SoundHound made the right decision several years ago by diversifying its technical efforts to include speech and AI. While music recognition quality may be approaching its theoretic limits (more unique songs are identified more quickly by both SoundHound and Shazam), there is a lot more room to grow and value to be created in speech and AI. With its heavy tech investment and consumer focus, SoundHound has positioned itself well for the years to come.