Alex White co-founded Next Big Sound in 2008, while in his last semester at Northwestern University. The analytics service measures daily music consumption and purchase decisions around the globe. White and his co-founders have been featured in Billboard (10 best music companies), The New York Times, CNN, Forbes (30 under 30), Techcrunch, BusinessWeek (25 under 25) and more. Next Big Sound works with thousands of customers, from individual artists to major labels, and licenses two charts to Billboard magazine. In 2012 the start-up announced a $6.5 million dollar series A financing from IA Ventures, Foundry Group, SofttechVC and other notable angel investors.
By Jerry Dawson
You started Next Big Sound in your last semester at Northwestern University. What did you see in the music industry and/or the tech world that inspired you to create Next Big Sound?
My background is in the music industry, so I came at it from a music approach, not a tech startup approach. First, at my internship at Universal I saw the weekly Soundscan reports that the industry executives were using to figure out whom to sign, what the consumers were purchasing, and how the marketing was working. At Northwestern, I was booking concerts and wanted to book acts within our budget that would sell out, and I asked real industry agents to refer me to the sources that they were using. They referred me to Poll Star and Celebrity Access, but those didn’t feel like they were capturing the way that my peers were interacting with music.
Those were two formative experiences that showed me that the data that industry professionals were relying on wasn’t coinciding with the way that consumers were increasingly interacting with music. There was a discrepancy between how people discover new acts vs. how the industry was measuring that. That’s the lane that we carved with public social data and now moving towards streaming, downloads, and that sort of thing.
In addition to writing for Kickshuffle, I also teach high school students. Data is huge in education today, just as it seems to be in the corporate world, the sports world, and now the music industry. Why are data analytics viewed as so valuable?
People have always wanted more data to make better decisions, except the sources and technology were prohibitively expensive or impossible to track. People use lots of proxis. For example, National Purchases Diaries used to have housewives track the purchases they made, and they’d see, “People who bought bananas also bought this kind of lipstick,” and it was based on phone calls, surveys, and focus groups because that was what was available, that was the only means for measurement. What has happened over the last thirty years is one by one the price reduction and the data availability has coincided with different industries being transformed by data. The first real areas with a low cost and high reward for tackling data and acting upon it was finance, so you had hedge funds and traders using all these statistical techniques because their storage costs were super high but the amount of money that could be made was so great.
Now you see sports using the data, and instead of daily box scores being compiled and rolled up, now everyone knows every ball and strike as it comes across the plate. Music was the first entertainment category to move almost entirely online, have all these data points trackable, and that’s where we started with MySpace in 2009. You can’t make high frequency stock trades using artists’ numbers, but you can make smarter decisions on who to sign. There just had to be that point where the lines crossed where the decision you make and the financial returns of that decision is in line with the data and the price of that data.
What are some of the misconceptions people may have about using data to make predictions, draw conclusions, etc?
I think thats the been the story of Next Big Sound– we mostly educate people. Here’s what the numbers mean, here’s what you can do with these numbers. There’s a lot of skepticism from entrenched industry players that we’ve been able to overcome. From the beginning we’ve been interested in metrics that are associated with dollar signs, so that really gets the attention of the executives who’ve always kind of viewed this as more of a nuisance than something to pay attention to. People now realize that new media aren’t going away, and they are part of the marketing campaigns.
Part of the marketing campaign, or increasingly, a majority of the campaign…
Yeah, exactly. These numbers are hard to ignore. When you see Justin Timberlake has 89 million Vimeo views, that’s an attention grabbing number, especially when you see that alongside nearly a million in first weeks album sales. People are getting used to seeing spikes in social media in tandem with sales spikes.
Your intro video says, “Next Big Sound: Knowledge to Navigate the Music Industry.” Data and knowledge are different, though. Be as broad or as specific as you’d like: how do you turn raw data into knowledge?
That’s a brilliant question, and very prescient. We have a data hierarchy pyramid posted in our office: at the bottom level is data, the raw numbers from all of the major sources. We turn that data into information, which is the second layer of our pyramid. Those are time series graphs, color-coded metrics, simplified versions of seeing the numbers move, and it all allows for easy reporting. That kind of defined, clean, intuitive interface is what we’ve spent years working on and perfecting. That’s really how we got our foothold in the industry: making the data useful and friendly, not requiring a PhD to master. The reason we raised our series A last year was to invest in the data science team to take the data to knowledge to intelligence. The knowledge is a step beyond information.
An example would be when Justin Timberlake has a TV appearance, we want to say here’s how your numbers moved compared to similar artists who’ve done the same thing. Basically we’re saying things like after you were on SNL, you saw an average increase in 60% of page views on Wikipedia, and other similar artists have had a spike of 20%. We’re adding more knowledge about where artists stand in relation to others. The first layer is the data: you had 500 thousand new page likes last week, you still don’t know if that’s a lot or a little, you don’t know where that puts you, until we tell you that makes you the fifteenth most-liked artist on Facebook last week, then the higher level of intelligence is when we say, other artists have seen even greater increase in page likes when they’ve appeared on these shows, or done these types of things. We have data scientists and PhDs working on this data to present it in a digestible way.
Just as the “next big sound” is a very volatile concept today, the “next big social media platform” is equally in flux–What does NBS do to keep up with the various platforms and their relative popularity?
Good question. We could spend all day integrating new sources, because of our client base, which includes all of the majors and thousands of individual artists and managers. Every new tech start up wants us to integrate their data. We make determinations based on a few factors: we want to hear the request to integrate that source from multiple clients over time. For example, big and small artists all wanted us to add Instagram. Also, there are sources requested where we don’t have easy access to the data. Shazam and Soundhound keep their data private and internal. Lastly, we look at if the numbers are available, how they line up to other data sources we track. If you are a great source with great data but not many artists, then it doesn’t make sense to integrate it, with the hundreds of thousands of artists we track. Every new source we add has to go through fifteen daily, rigorous tests, and adds a huge amount of overhead, so if the data isn’t universally applicable, we try not to add it.
What are your thoughts on crowdfunding music projects? Does Next Big Sound look at these numbers?
I love the movement and success of Pledgemusic, and Indiegogo and the more mainstream Kickstarter. There’s a lot of interesting stuff we can do…we’ve been asked to track that data. Those are quick campaigns, and it’s unclear how we can systemically track that. In terms of Next Big Sound, I’m interested in using our data to predict whether the artists will be able to hit their funding goals before they start campaigns. We would partner with one of them, get a list of successful and unsuccessful acts, and we would run regression models on our side to see how early on we could predict the success of those campaigns. That study hasn’t been run, but that’s the kind of interplay I’m interested in.
Is the Next Big Sound able to differentiate between artists who surge in popularity because of, let’s say a gimmick type approach–think OK Go’s music videos, or something like “Harlem Shake”– and an artist that is truly developing a grassroots following?
We very much want to help our clients differentiate between the two. Because this all ties back to the music business, the key second word being business. If it’s a one hit wonder but is structured like a multi-album contract, then it’s a mismatch for everybody, and what was successful and popular could be a money-losing endeavor. We can see size and dramatic growth in “Harlem Shake,” and it looks different than The Lumineers and their rise. We can see across YouTube, iTunes and Facebook and we can start to discern when an artist is going to continue to run away and climb the charts as opposed to when the artist has hit a cultural zeitgeist on the head but is not going to maintain that popularity.
There’s a parallel to venture capital where, lets say it’s a gaming company, and the investors have to know, is this a lucky break and the company has one game, or does the developer have a long term plan, know how to sell ads, know how to routinely and systematically develop these games. Then the venture capitalists invest appropriately.
You participated on some panels at SXSW, both with great names– “Big Data: The New Oil or the New Snake Oil” and “Man vs. Machine: Data Science and the Future of A&R.” What were your takeaways from those discussions?
I liked the varied perspectives. For the “Big Data” panel, I heard David Lowry, a professor at UGA and a friend of the company, and longtime user of Next Big Sound. It was good to hear how he as an artist uses the data in his workflow. He’s a savvy artist who’s checking in frequently.
For the “Man vs. Machine Panel,” it was great to hear Jacob Fain, who does A&R at Sony/ATV talk about his processes for researching hit records. We have a new product rolling out that’s geared towards research guys like Jake. It was nice to have an open dialogue about that, and Billboard did a nice write up and coverage of that. But really it was about how far does a machine go and where is a human ear required to make the final decision. Next Big Sound is a tool, but it isn’t going to write your song, or record it, or mix and master it, or go on tour or do any of these things. The whole goal is to provide data to inform any kind of decision in the music industry. That’s where we require a deep understanding of the biggest questions our clients face, and how we can use the data to help them make those decisions. That was the higher level point of both of those discussions, along with the point that we are twenty-two folks here in New York City working day-in day-out, who’ve come a very long way, and for whom the next year or two is even more exciting.
The idea I find especially interesting is your emphasis on the fact that Next Big Sound can’t make the decision for managers, labels, etc, but you can provide all of that data available possible to inform those decisions.
We can’t negotiate a deal, or explain why an artist should sign with one publisher vs. another, but we want to automate what we can because the cost for tracking all the data has plummeted, the amount of information available is greater than ever, and computers can do a lot of stuff now that was previously very manual or completely impossible to track.
What’s up ahead for Next Big Sound?
We’ve spent a lot of time developing the data and information layer, and now its about moving up the data hierarchy pyramid. The products we’ve built help support heads of global marketing at global, major labels, but there’s a lot of room and focus now on making those tools and that information available to every artist on the planet. I also think that the research A&R product we have in the hands of early users and customers is going to completely change the way that artists are discovered and talent is identified, and that’s the focus for the company this year.
You were recently named on Forbes “30 Under 30” list for music. From that list or elsewhere, who else should we be looking out for in the music tech world?
We did a mixer last week with our new friends from the company Songza, who are doing some great stuff. I love their product. There’s a new company called BioBeats, who’s doing some great bioinformatics stuff as it relates to music, and other than that all of our data sources–Spotify, Soundcloud, Rdio, are doing great things as well.