When you hand out free accounts to a digital service -- especially in the age of Facebook Connect -- you're likely to find most people just aren't that into you. That's what Spotify and Deezer have found, notes music industry analyst Mark Mulligan. Whether or not it's a problem -- and how to fix the problem -- is open for discussion.
Inactive accounts represented 70% of Spotify's registered accounts at the end of 2011 and 73% of Deezer's registered accounts this year. [Update: Deezer CEO Axel Dauchez tweeted on March 29 that Deezer’s inactive users account for 62% of registered users.] The important point here, writes Mulligan, is that "streaming services as a whole have a problem with churn." The term churn means loss of customers. Churn is expensive to companies because it is always more expensive to gain a new customer than to keep an existing customer.
But worrying about an unconverted, inactive group of registered users breaks with the logic of the Internet. Inactive users are just a part of doing business online. Getting people to register is just the first step. Not everyone will become a frequent visitor or paying customer. As Mulligan later noted in an update to the blog post -- after an exchange on Twitter with Spotify CEO Daniel Ek -- "this is a problem that affects all businesses that have a free tier that requires registration."
Inactive accounts are a part of doing business for many online services. For instance, a 2012 study found 70% of Facebook pages are inactive. Last year Semiocast found that 73% of Twitter accounts are inactive.
Retention is difficult elsewhere. A recent study by Flurry shows a third of apps were able to retain 37% or more users over a subsequent 30-day period. Email lists have a similar problem with inactivity. Various estimates put the average email list's inactive addresses at 65% to 70% of the entire email list in a given year.
All this inactivity by a large number of users brings to mind the 80/20 rule that says 20% of customers account for 80% of a company's sales. (It is said to hold true in other areas. For example, 20% of computer bugs cause 80% of crashes.) The other 80% of customers spend less and buy less often than the super-active 20%. Some are first-time customers who will never return. Others could be infrequent customers who buy once every year or two.
The 80/20 rule is also known as the Pareto principle. People in the music business may recognize the name "Pareto" from the book "The Long Tail," whose familiar graph is that of a Pareto distribution. We see something resembling the Pareto principle in Internet radio. Pandora says only 4% of its users are likely to be affected by the mobile listening caps it recently installed. That is to say, just 4 out of 100 active users listen to Pandora more than 40 hours per month. The Pareto principle also applies in the amount of time people will put into creating, editing and viewing Internet content. It's called the 1% rule or the 90-9-1 rule. This rule says that only 1% of people on the Internet will create content, 9% will edit content and the other 90% will view the content without contributing. All over the Internet, relatively small groups of people are accounting for large groups of activity.
Mulligan believes subscription services can reduce their level of inactivity and manage churn by making music more social. The solution -- it would be an market-wide solution, so don't hold your breath -- would be to make the playlists portable and universal so they can be played on any service. The idea is that the sharing of playlists, and thus participation in subscription services, will increase just as use of mobile phones adoption increased once customers were able to call and text people on other networks. Universal playlists may help adoption and reduce inactive accounts on the margins, but only to the extent people discover music through friends. The fact that services are now putting so much of their effort into curation and editorial suggests they discovery may be less social than once thought.
I think there's an important lesson in these inactive accounts that should not be overlooked: most people say they love music but won't invest much time in it. Nielsen's recent study "The Buyer and the Beats," unveiled at SXSW, indicates this fact very clearly. The survey found that most people don't enjoy creating playlists with specific themes or sending playlists to friends. Most of them don't read music blogs. Even the groups of consumers that account for over half of all spending on music don't really care to receive emails from favorite musicians.
Yes, it is unfortunate that music services have so many inactive accounts. But since this is also an issue faced by other Internet services that offer free registrations, where does one draw the line between an acceptable and inacceptable amount of inactivity? Where does it cease being a typical problem and start being a problem that needs a solution?
I think a better way of looking at this issue is to follow the most important metric: subscribers. The paying customer is a far stronger driver of a music service's revenue than advertisements that target the free customer. To get subscribers a service has to register new users, and it’s going to register more users if registration is free. Some users will become paying subscribers, others will remain active users and others yet will become inactive users. Along the way the service can take steps to prod inactive users to rejoin the flock and add to subscription growth.
A freemium model like Spotify and Deezer will certainly register many ambivalent music consumers who are better served by Internet radio. These services just aren't for everyone. Some inactive users may return once functionality improves or new features are added. A key lesson of the 80/20 rule is that companies should focus on the 20% of customers from which they derive nearly all their revenue. In the case of Spotify and Deezer, that means catering to the relatively small number of consumers who will pay for the service.