Google’s New Music Crawler – How to Game the Algorhythm
Following feedback from the Google Play service in the US, insiders indicate that Google is set to roll out its own music ranking service later this month. Internally referred to as GTunes, SEO engineers are falling over themselves (and their spellcheckers) to christen it “The Google Algorhythm.” There’s a good chance it will revolutionise music as we know it.
Google’s low key start with its Google Music offering – recently redubbed “Google Play” as part of its app store renovation – has been informed by past mistakes in its quest for total informational domination. The aggressive push for content under the Google Books and Google Scholar programs attracted some high-profile criticism from libraries and copyright owners. The lessons learned have clearly informed the “softly, softly” approach being taken with artists and music collections in the new scheme, listening to their ever important concerns about piracy and royalties.
Meanwhile, consumer habits have shown that simply having the right songs at the right price doesn’t guarantee sales: being able to find what you want, and what you didn’t even know you wanted is critical. With its unparalleled insight into a world of data, Google is better placed than anyone else to serve up a catalogue with that kind of insight.
Slaves to the Algorhythm
To this end, Google is tweaking a subset of the existing search algorithm – already being dubbed “the Algorhythm” in certain quarters – in order to index and rank the music tracks listed on the service.
Most interestingly of all, Google is setting out to “discover” and rank music online by actively crawling the web and downloading tracks and MP3s it finds on user websites. Initially this will just be those who are logged in to a Google profile, but pending a successful trial it will be extended to all web users.
The basics for preparing for this important change are becoming widely known. Although it is expected that a user-submitted metadata file will be analysed at the point of the track submission (much like the video sitemap file currently supported by Google), it is also relatively easy for the Algorhythm to parse the audio amplitude and plot a waveform pattern; similar to what is seen on bass and treble volume/level indicators.
From this, Google is able to draw comparisons between the type of music specified in the data file, and the tempo, volume and regularity of the waveform extrapolation. Social approval signals and the amount of track listened to (before being exited or skipped) is also expected to further inform the actual ranking position.
A New Age for SEO?
As SEOs, the possibilities that these ranking priorities create are exciting, and the issues they present are numerous. Will certain genres and styles have an unfair advantage when being recommended to the music-listening community at large? Will we see the creation of a new generation of SEO-savvy artists, lacing their lyrics with strategically placed keywords? Only time will tell.
There’s already evidence that the music industry has caught wind of these changes. Justin Bieber’s “Baby” may have been dismissed as the death-rattle of human civilisation upon its release, but watch it in light of the upcoming changes and it seems like a masterwork of Machiavellian keyword-stuffing.
I guarantee that he’ll be number one for “Baby,” “Woah,” “No,” and “Oh” the second the update goes live.
Some key points to consider:
- It’s important to remember that many Google philosophies are likely to be expressed in how the new music search engine ranks music;
- Drum-machines, music sequencers and similar devices are expected to fall afoul of Google’s stance on auto-generated content;
- Status Quo is reportedly researching at least two extra chords after it was suggested that a large percentage of its 60 chart hits could get flagged as duplicate content;
- The potential for breaks and samples to be viewed as scraped content may achieve a change that decades of copyright lawsuits have failed to affect. If the bling-laden fingers of the Hip-Hop fraternity aren’t shaking on the triggers of their MAC-10s, they ought to be;
- There’s also the issue of how the new measures will interpret music from fringe genres and experimental artists.
One SEO-savvy friend working for a major spoken word label told me, “I’ve got beat poets phoning me up, asking how it’s going to affect their royalties. The optimistic ones need to be talked down – yes, people are finally going to value your original, environmentally-conscious lyrics, but droning monologues are no longer going to cut it. It’s an Algorhythm change, not an Al Gore-rithm change.”
Preparing for the Algorhythm
If we’ve impressed upon you the potential chaos of this latest update, you’re probably asking: how can I safeguard my site and get ahead in the Google optimised audio search? We’ve already had some thoughts, and observed some phenomena which may pre-empt upcoming changes:
- An early conclusion from our limited testing shows that the GTunes algorhythm takes a large cue from the metadata associated with the file submission, and in particular, microformat data available when the audio file is crawled and scraped;
- This may be a pilot project in the much-hyped “semantic web” development very much on Google’s agenda, so it pays to ensure that any track information is tagged so as to be content-rich;
- Google appears to be utilising its own in-house schema as opposed to adopting any of the proposed schemes currently on the web (frustrating, but hardly out of character), and it’s expected that this will be able to hook in to the social signals given from the G+ platform and tracks added to user circles;
- The weighting applied by the algorhythm is harder to quantify, but according to Google’s maxim of “relevance” it would be suggested that any tunes categorised as “rock” for example, would be looking at a 4/4 time signature and an average bpm (beats per minute) of at least 160. A more sedate category, such as a waltz, would be looking for a 6/8 time signature and a far lower bpm.
Some users may not wish Google to visit their music pages and take their music content for GTunes. This feature is of most interest to those wishing to control copyrighted material, but it may also appeal to those who do not want the internet at large to discover their online directory of embarrassing music.
<meta name=”gtunebot” content=”noindex”>