audiokite_1266x704

What Artists Should Know About AudioKite

Has this ever happened to you? You think you’ve written your best song yet, but an offhand remark from a friend plunges you into self-doubt. Wouldn’t it help to have feedback from music fans of your genre who have no incentive to sugar-coat their opinions?

Sure, you say! I’ll just use SoundOut, or ReverbNation Crowd Review (also powered by SoundOut). Unfortunately, my experience with SoundOut, and those of most of the commenters, left a lot to be desired. I’ve also received a mostly useless – but free – focus group from Music Xray, and even repurposed Jango aka Radio Airplay to create my own focus group.

AudioKite has built a better mousetrap. Here’s why:

Amazon Mechanical Turk. Listeners are enlisted from Amazon Mechanical Turk rather than SliceThePie, the listener end of SoundOut. I don’t know the nuts and bolts of the operation, but the end result is that listeners comment coherently and seem to actually listen to the whole song. There’s no sign that anyone is trying to game the system by listening to only the first minute and copy/pasting generic reviews.

I asked Alex of AudioKite to explain how they ensure listeners are actually listening. Here’s what he had to say:

Due to the fact that a fair wage for mTurk workers is based on the amount of time they take to complete a task, and the fact that songs have varying lengths, listen time is a data point that changes from song to song. But despite our minimum listen time requirement of 40 seconds, nearly all respondents listen more than that – exactly 1 minute 44 seconds on average as of this writing. The average listening time for your song was 2 minutes 58 seconds, which is pretty high. We track how long each respondent listens and actually use that data point in your “commercial viability” percentage rating. The thought process here is the better your song is, the more time people will spend listening to it. We also go back and give bonuses to mTurk workers who write long reviews or listen longer.

Genre-targeted reviews. With SoundOut, I felt like a healthy percentage of the listeners simply don’t appreciate my style of music. With AudioKite, you can select from seven genre categories to have your music reviewed by fans of your genre. Currently, the genre categories are a bit broad. Electronic is grouped with Avante-Garde, whatever that means, so I chose Pop/Dance instead. Hopefully that will change as the operation expands.

Pricing. SoundOut starts at $45 for 125 reviews. ReverbNation Crowd Review costs anywhere from $9.95 to $99.95 for 20-200 reviews at various levels of reporting. AudioKite starts at $29.99 for 100 reviews, which is the sweet spot in my opinion. You can get that down to $19.99 with a $9.99 Pro subscription, which includes some extras:

Free response question. The default question listeners are asked to respond to is, “What did you like or dislike about this song?” With a Pro account you can change that to whatever you’d like. The pricing is low enough that you might consider getting several reports over the course of writing and recording a song, adjusting as you go based on listener feedback. In that case, you might ask something like “Is the chorus too long?” or “How would you improve the mix?”

Promotional link. With a Pro account, you can include a link your web site or a destination of your choice, so listeners can hear more if they so desire.

Another big difference between AudioKite and SoundOut is that the reviewers are all located in the US. I asked Alex to clarify:

It was a choice, for a couple of reasons. The first is that in our research before we set this up, musicians were mostly concerned with the opinions of people in the United States when it came to music insight. However, we’ve been getting people asking about Canada and the United Kingdom. As we grow and our understanding of the market develops, expanding internationally is inevitable – it’s definitely something we’re looking at.

The second is that by opening it up to all countries, we would unfortunately and inevitably receive a large number of low-quality responses from click-farm-esque operations in India, China, etc. who try to game mTurk. You mentioned SoundOut before, which uses SliceThePie for their reviews – according to Alexa.com, around 20% of SliceThePie’s traffic comes from Kazakhstan. I don’t know about you, but my band’s marketing efforts aren’t really based around selling to people in Kazakhstan, so why would I want 20% of my responses from there? To ensure high quality responses, we’re keeping responders US-only for now.

One thing remains the same across both services: the reviewers can be brutal!

What I thought was the “hit single” from my upcoming EP scored a mere 5.4 out of 10, so I tried again with another track, which managed a 6.3. At that point, I was officially addicted, so I tried a third song, which scored a 6.0.

Here’s most of the highest scoring track and how it fared (I’ll embed the whole thing once the EP comes out):

AudioKite elements

AudioKite Purchase

AudioKite Average

AudioKite Word Cloud

Some sample responses at both ends of the cruelty spectrum:

“I immediately thought of Napoleon Dynamite when I heard the beginning of the song. After it progressed, I couldn’t really take it seriously. I felt like it would be on an SNL skit.”

“Really don’t like the fluctuations in the singer’s voice. I don’t know if that was auto-tuning, but it just makes it sound whiny at those parts. Lyrics were on the poor side of mediocre, beat was generic synth-pop. Overall quite uninspired.”

“I like electronic music, and this song has a nice upbeat feeling to it! I also like the lyrics and the vocalist’s delivery; his voice works very well with the notes in the instrumental, and it all blends together nicely.”

“Felt familiar enough to what’s on the radio now, but also had enough distinction and character to break it out of the run-of-the-mill radio mode. It’s hard to imagine getting tired of hearing this on the radio, because it both calms and makes happy at the same time.”

Check out the entire report here.

Bottom line, I’m impressed. And hooked.

This is the best market research for musicians I’ve seen yet. I like the service so much that I asked for an affiliate link, which Alex was gracious enough to provide. You can support this site and get a $10 discount on all reports except the cheapest one (100 listener, non-targeted – $5 off) when you use AudioKite coupon code AK-PASSIVEPROMOTION.

Discount Code AK-PASSIVEPROMOTION

I’ll close with my favorite reviewer comment:

“I want to say this is the third Color Theory song I’ve heard thanks to AudioKite? I really loved the lyrics, they painted such a great picture in my mind. At this rate I might just end up a Color Theory fan despite my initial misgivings.”

Have you tried AudioKite? What did you think? Any other options out there beyond SoundOut? Let’s talk about it in the comments!

24 thoughts on “What Artists Should Know About AudioKite”

  1. So… why let offhand remarks from a friend send you plunging into self-doubt when you can pay for deliberate self-doubt-provoking remarks?

    1. The comments I picked were definitely at the extreme ends of the spectrum. If you review the report (link above), you’ll see that most fall somewhere in the middle. I scored worst on my vocals (um, ouch), and highest on sound quality (one would hope, as I’m a mastering engineer by day).

  2. Hi – as mentioned above, here is my report. I should not have entered a dance remix as many of the listeners did not understand that the tune is designed for the dance floor, not for kickback listening. The comments at the bottom do not match the overall score of 68%, and thats good news, because most are positive.
    http://www.audiokite.com/report/?report-id=OTI= via

    If you want to listen to the tune I submitted, here is the link:
    https://soundcloud.com/self-same/fragments-featuring-shaz

    thanks again for enlightening me to this service!!

    1. Okay, wow. This is a fantastic track!

      That’s the best score I’ve seen so far, and I’m sure Alex from AudioKite would agree it’s a high one. While I think you might have scored even higher with a radio edit, it seems most of the listeners had the patience to hold out for the meat of the track.

      Interesting that you scored particularly high on the east coast!

      Quite a contrast with my reports, in many areas. This is good stuff!

      Thanks for sharing!

      1. thank you – yes, Alex wrote me and stated my 7.2 is very high in comparison to other tracks. He also mentioned that my song element scores were also unusually high. This makes me feel great because I have invested a lot of money & resources on this EP, hiring the right producer was key. I am finally recording an album of music that I want to hear – (no more smooth jazz!) – so this has been very encouraging, to say the least!
        I added another track to the service, a chill ballad – I’d be curious how it works. Thanks so much for the discount code, and your blog. I’ve learned a lot from this website!

        if I may, here is the link to the ballad. (the EP will be released in July):
        https://soundcloud.com/self-same/beauty-working-title

        1. You may! ;) Beautiful track!

          Smart of you to hire a producer. Too many people, myself included, insist on doing everything themselves. I could’ve saved a lot of time by learning from a pro from the start!

          1. Thanks for the kind words!
            Yes – it was a good move, but an incredibly expensive decision. That said: one good sync license could pay off the entire budget!
            I just got word today that Beatport approved the track as well as Pandora. SiriusXM will be the most difficult. Extremely excited about Beatport as it’s curated, thus, lots of good company surrounding the music.
            My advice to anyone wanting to make a pro album – hire a legit team: producer, PR, and consultants. It’s worth it. I’m fortunate to have a distribution deal with The Orchard, which has helped open many doors – but, my knuckles are bloody from knocking on so many of them!!

        2. Hey Tony, congrats on Beatport & Pandora, that’s awesome! Again, really nice job on your track.

          Knocking on doors is for sure the right way to go about it – this article wouldn’t have happened if I hadn’t knocked on Brian’s! It sounds like you have the right formula for success, and we all look forward to watching your career flourish.

        3. Out of curiosity how long did it take Pandora to approve your tracks? I submitted my music somewhat recently and I was wondering about the timeline. Thanks!

          1. They’ve opened up the requirements quite a bit since I last submitted, so I’d imagine they have quite a backlog. I honestly can’t remember how long it took last time. Weeks?

          2. Hi Bob!
            I didn’t know they reduced the requirements – I’m surprised they did, as curated experiences seem to be the next wave. (Unless they plan to have dedicated independent artist channels?)
            IMHO, One of the issues with all of the platforms is that more bad music is pushing down good music, thus requiring promotional commerce to dig out a hole to be heard, just like terrestrial radio. (Another topic, right!!) I digress…
            That said, my older contempo-jazz stuff was added years ago and it only took about a month from soup to nuts. Pandora approved my song ‘Fragments’ almost 3-months ago, with great fanfare from their ‘congrats’ email. It has yet to go live…

          3. My “opened up the requirements” comment was in regard to them no longer requiring a physical CD on sale at Amazon. Digital-only submissions are now allowed.

          4. ah – that’s right. one needed a physical CD prior. Its good that the requirement still consists of music that distributed through an e-tailer. I’d be curious to know how they choose the music, however. I know a few people who have made excellent music, but cannot get approved on Pandora. Oddly, I’ve never Googled that…

  3. I really think more artists should use AudioKite as I am gaining some valuable feedback. One needs to remember to not allow the crowd to dictate your art, especially if you are only querying only 100 folks. I plan to use the large pool on my next test. Here is the link for the results of my chill ballad ‘Beauty.’ – scored a little lower than ‘Fragments’, but I really got a kick out of the comments. The comments, IMHO, are the most valuable aspect for gleaning hints for flushing out the kinks…
    http://www.audiokite.com/report/?report-id=OTU=

    the tunes are all on Soundcloud – search self.same

    thanks again for the awesome post!

    1. Good stuff Tony!

      Still a good score. And I agree, the comments are key. Some can be disregarded, but some are very insightful, with suggestions that are relatively easy to implement.

      I’ll be curious to hear how a larger sample size changes the experience for you. I’ve only tried 100 listeners so far.

  4. Hey Brian! Just started following you recently! Thanks so much for your reviews you posted on SoundOut and AudioKite. I use TuneCore and they provide a service by SoundOut so having read your blog, I decided to try AudioKite instead!

    Such an awesome service! And yes as you said, some of the comments can be a bit brutal! But hey that’s what we sign up for!

    I was pretty happy with my score, considering it was a christian acoustic style song and it’s not exactly professionally mastered or anything! I used your promo code too to wipe off some $$ thanks!

    If I’m allowed, here’s my report: http://www.audiokite.com/report/?report-id=MjE3

    And my song: https://soundcloud.com/cruz-folmli/secret-place

Share Your Thoughts