Sachin Rekhi

Despite the over 200 music startups that launched in 2008, I am disappointed with the lack of startup innovation in the space. While many of them definitely nailed building something people want, most failed to make something people will buy. When developing a startup, figuring out a viable business model is as important as producing a compelling product. And unfortunately there was little in the way of innovation in business models in the music space.

This is an even more pronounced issue in music, where content licenses are owned by an oligopoly of four highly litigious record labels. The labels have been eager to shut down or sue a variety of music startups, including Muxtape, Songbeat, Seeqpod, Project Playlist, and others. Yet what the labels are struggling most with is recouping lost revenue from the 45% drop in CD sales since their peak in 2000. What they are most desperate for is not a large settlement, but a new sustainable revenue stream. What they need are startup partners seeking to help them generate new commercial value with their content and through relationships between artists and fans.

There are a few notable exceptions who did focus on business model innovation. LaLa, for one, introduced a new a la carte on-demand streaming pricing model, charging users 10 cents per track for unlimited on-demand streaming. Sure one could argue that there are already too many free alternatives for on-demand streaming and thus there is a significant uphill battle for LaLa to gain traction. But LaLa is focusing on trying to build a compelling enough user experience with a clean playback experience that would be worth paying for. And at the same time, it's still to be seen whether the existing fully ad-supported music streaming services are going to be profitable businesses. To be sure, LaLa is far from a success, but at least it went out on a limb and tried something new.

However, the most interesting potential bright spot in online music monetization is monetizing time spent listening through online audio ads. While the online audio ad market is still nascent, the offline radio advertising market is still a $21B industry. The time is now for these dollars to move online in a meaningful away.

There have been various attempts at developing the online audio advertising space to date. Unfortunately early attempts were limited in success. Some have tried to bring the same offline radio ad spots from broadcast radio to online music streaming services. These are your typical 30-60 second national brand ads, like Geico car insurance ads. The attractiveness of this format is that it is much easier to convince ad agencies to take their existing audio creative and do a test buy online. For broadcast agencies, it is exciting to finally have access to actual impression data. To them it is a novel concept to actually know exactly how many people viewed their ad spots, since in the offline world much of the metrics are based on sampling. Yet this traditional radio ad format is completely inappropriate to the web world, where oftentimes users choose to move to online music services to get away from these long obnoxious ads.

What is needed in the online music space is an entirely new audio ad format that is appropriate for the medium. Just as how the online video space has developed its own unique format - the video ticker overlay ad unit - a similar new ad format needs to be developed to adjust to user behavior and expectations regarding online music.

And that's exactly what we are finally starting to see happen. imeem, for one, is serving a unique 5-8 second audio ad spot at a low frequency with a companion medium rectangle (300x250) banner that takes over the player during ad playback. Pandora just introduced 15 second audio ad spots with an initial frequency of one spot every 2 hours. TrueAnthem has developed an 8 second audio spot with a companion banner played at a frequency of once per session. Jingle Networks, which provides a free 411 voice service, has been serving 10-12 second audio spots from both national advertisers and geo-targeted local advertisers. They have had no trouble filling their inventory and have even launched a complete voice ad network to allow the insertion of in-call audio ads for other services.

There are several existing online audio ad networks already out there, including Ronning Lipset Radio and TargetSpot, which recently merged to create the largest online audio network. Unfortunately a lot of the units they carry tend to be the longer more traditional 30 second ads. This month Google announced that they would be shutting down their broadcast radio advertising business that Google has pursued since its acquisition of dMarc Broadcasting in 2006. While many picked up on that news, I think the most interesting piece is that Google also said that they are going to now actively pursue online streaming audio ads. I'm excited to see Google enter this space and I hope they bring with them advertisers willing to try shorter and more innovative formats, even if it means Google or someone else taking on the challenge of producing this new custom audio creative. This is exactly what VideoEgg had to do in the early days to get advertisers to embrace the video ticker overlay.

Early data suggests that in-stream audio ads have higher CTRs and greater brand recall than traditional online banner ads and can be minimally detrimental to site usage when done tastefully with shorter audio spots at low frequencies. Instead of seeing yet another slick AJAX interface for creating and sharing music playlists, we need more entrepreneurs to optimize the audio ad format to figure out the right set of parameters to make this work. Whoever does that may be able save the music industry and line their pockets all at once.