The consumption of streaming music is increasing every nano-second. Usage is up 76% year-over-year, and just recently Goldman Sachs predicted that the music streaming industry would reach $ 28 billion annually by 2030. But that is. doesn’t mean that there aren’t a lot of grunts from certain camps within the music industry. Not on the appropriate payout percentages (this is a whole other topic of controversy) but on the technical verification of the number of flows themselves. Some even have made the claim that there is no valid system in place that actually verifies the counted streams for top musical rankings. This is especially intriguing given the huge rate at which Millennials consume music and more on digital platforms. But how valid are such worries and how can it all play into what can sometimes breathe life into or spell the end of songs and albums: the almighty record position?
First, let’s take a look at the technological process that creates the streaming counts and determine if it’s as easy to play with the system as some might think. Antonis Karalis, Product Architect and Co-Founder of HPCmusic, explains: “This is how it works. Streaming providers use data mining to manage streaming as well as to count manipulation incidents. Data mining and regression allow data scientists to track, store, and visualize all the streams on the music catalog. “He continues,” Then they use statistical processes to estimate relationships between different variables such as the user behavior, social trends regarding the artists, the sound characteristics of the file and several other macro and micro trends. ”Karalis says that while these tech experts work, their absolute goal is also to create and refine a model that will allow them to those who try to manipulate the system.
“These models are really accurate,” he says. “Even without AI (artificial intelligence), the level of confidence that the industry operates to identify manipulation is over 90%, which is more than acceptable. At the cutting edge of technology, when AI enters the market mix, the confidence interval (CI) goes through the roof. “
To that end, a spokesperson for Billboard and Nielsen Music (which provides information on sales, streaming and distribution of Billboard charts) confirms that the company “works closely with each data provider to ensure the both the accuracy and legitimacy of the reported streaming volumes. . A system of safeguards exists to identify and rule out any irregular and excessive streaming pattern. “
Yet these assurances do not seem to satisfy the artists, managers and other disgruntled people who work in the arena of the music industry. So, is there a way around such high tech to simulate human behavior the same way a bot is used in other arenas or are these people just falling victim to sour grapes and egos? ? “A company like Spotify can track the behavior of users and user agents to assess whether a given user is a real person or a bot,” explains Jordan Mendler President and CTO of The Veloz Group. “They can use data analysis to track listening behaviors of humans, for example, to see whether or not the user is streaming music to a Bluetooth device or if they are doing things like pausing songs. , go forward, adjust the streaming volume or start songs from the beginning or if they stream from devices like iPhones and web browsers or only stream music on one type of device. “
However, unlike Karalis, Mendler says dishonest behavior might go unnoticed. “It would be pretty easy to manipulate the system just by mimicking the behavior of a real user,” says Mendler. “You could study the behavior of a real person playing music and develop software to mimic that behavior. For example, you could emulate an Android device using an automation platform to search for music, play it. , rent it, and simply move it. to the playlist. “
If so, are such tactics used? If so, how often could they be used and could this impact the music charts? “For the most part, we just trust all of these facts and figures released by the streaming services,” said Paul Resnikoff, editor of Digital Music News. So when a company like TIDAL, for example, says it has 8 million subscribers (or whatever it claims), we really have to take that as the truth. Same with any other service. “
“But I think the biggest problem with the Billboard charts – which artists like Meek Mill and Tyler the Creator etc. point out – is that all of these fancy algorithms for figuring out what the ‘biggest album’ is are. really arbitrary, ”Resnikoff adds. “The measure is tricky, and besides, the album means less and less to the music industry every day, even though it was once the gold standard and the measure of success. really a stretch, and leads to all kinds of distortions. “
A lot of noise for nothing ?
Eric Holt, assistant professor of Music Business at Belmont University, recently named one of the top 10 music schools in the country, agrees to some extent: “Billboard has been a measurement tool for the industry since then. long time. I think the Billboard is relevant. , but here’s the news: maybe not for long. “Holt adds,” Aside from the possible flow manipulation, when you watch artists like Chance the Rapper get the best success in the business bypassing all the sheets. routes, and innovations like Jay Z mixing partnership pre-purchases, it makes sense to me that to stay Billboard needs to create better measurement tools to fit current and future realities in the music industry rather than running to follow them after the fact. “
In the meantime, what is on the horizon, technically, to create more assurance of validity for the charts that are valued today? “Companies like Spotify use data mining techniques to monitor the behavior of their users. They measure things like engagement and listening patterns to determine if a given user is a bot or a real human, ”says Karalis. “They have a team that enforces user quality by identifying auditors that are machines and either blocking them or asking them to verify that they are human by completing things like Captchas.”
One thing is for sure, questions and analysis will always remain when one item is ranked above or below another. Technology could complicate or simplify such situations. While music has never been so popular, other disruptions and controversies will surely continue to follow your favorite tracks.