Spotify Discovery Mode - Is It Worth It? | Best Way To Do It
ฝัง
- เผยแพร่เมื่อ 15 พ.ย. 2023
- This video is a follow up to last week's video about Spotify's Discovery Mode.
I share the results of over 50 Discovery Mode campaigns and lots of tips on maximizing the program. For example:
How many songs to put in Discovery Mode?
Will putting more songs in get me more streams?
How to use the program to test songs in the algorithm.
What kind of results to expect and is it worth it?
Link to Spotify’s Discovery Mode page on Spotify For Artists: artists.spotify.com/en/discov...
Take action and join BandBuilderAcademy.com
For next level music marketing tips, hit that SUBSCRIBE button!
--------
Join Band Builder Academy today - We're educating musicians to win influencer support and get official playlists on Spotify and Apple. Let's clean up your music algorithm and apply the latest promotion tactics to your next song or album release!
About Todd McCarty
I'm the industry insider on your team. A former Senior VP of Sales at Sony Music (USA) and long time label manager of large indie label Fearless Records.
Get a quote for working 1-on-1 with me - found.ee/WorkWithUs
---------
Learn more about full academy access here: bandbuilderacademy.com
- Band Builder Academy offers the convenience of having direct access to Todd and monthly access to a complete library of courses covering streaming, promotion, influencers, and income.
- No-nonsense or hype with an easy to use guided roadmap. Not just theory. Detailed step-by-step methods.
- Active and supportive community (members only), one-on-one coaching support, and algorithm tool.
Don't wait, join today.
![USAB Showcase: Australia vs USA [OFFICIAL STREAM]](http://i.ytimg.com/vi/keiEr_MOaSQ/mqdefault.jpg)
Dude I wish I would’ve known this 9 months ago 😂😂 but hey, I have the results to easily know my top 4 songs based on data and there’s 5 hours left to update my campaign for this coming month… LETS MAKE THIS HAPPEN!! CAN ONLY MOVE FORWARD! Thank you!!!
No problem. And since the video was published 6 months ago, there is a change.. I think as long as your songs are getting more than a 65% "stream lift" you should put it into the program. So Yes, if it means putting all your songs in, do it. Unless there is a situation where you don't want a particular song in there for some reason. Use the program while it lasts. Also, I found out that Discovery Mode submissions of tracks goes by URI... so if you have multiple releases of a song (single, EP, Album) make sure to submit all URI's of a particular song. Todd
@@BandBuilderAcademy you are awesome!! Thank you for the advice !!
@@BandBuilderAcademy Can you elaborate more on the change, have you noticed a difference lately in the program?
@@PSPOD I think it didn't change, it's just that there is more inventory than I realized. There is more room for more of your songs. putting more songs in the program, doesn't guarantee you will get more streams, but you will have more chances to re-surface an older song. So it's worth trying deeper catalog, rather than just your top 5 performers. As long as the songs are getting more than a 65% lift, you might as well put it in. You will get more streams than before, and if you're over that 65% lift you will earn more money.
@@toddmccarty5355 Gotcha, I wonder how much catalog's past data affects the number of streams. I have a song on an editorial and another with a higher popularity score, but the program prioritizes an older song rather than the one's doing well organically. Have you seen any correlation between song age and stream lift?
Lately I’ve been noticing my results not being as grand. I think opting in less songs will help. I usually do everything that’s available, but next month I’ll see how just choosing my 5 best will do! Thank you so much for the information 🙏🏾
😅 There may have been a change of heart in the "best 5" versus "all"... The newest insight is that there is plenty of "autoplay" and "radio" inventory... so they can take more of your deep catalog. But the main thing is "stream lift". If you have higher than a 65% lift in streams, then it makes sense "royalty wise". See how it goes with your best 5, and maybe it moves the needle. If not, go back to how it was.
The reach has a lot to do with how many active listeners you have. So the limits of exposure are different for each artist.
@@BandBuilderAcademyThank you for the update! I really appreciate it 🙏🏾
@@ChiefTakinawa no problem
Good info! I Just got into the program after only having a consistent 1,200 monthly listeners, I am excited to see the types of results I will see. I also agree 2-5 songs is perfect to start with
thanks for letting us know. Please report back with your results :)
how did it go?
Tons of great info here....Thank you!
Credit where credit's due. It was YOU who inspired this video from your post in the Band Builder member community. 😎🙌
I really needed to see this a week ago...I have 15 songs in my first campaign that started TODAY (slaps palm across face) ah well, failure is the best teacher and I have never been disappointed with my results from actively pursuing failure. I'm sure some happy "coincidence" will yield useful information for next month! Thanks for the great video!
Thanks for checking out the video and the comment. No worries, maybe you'll uncover something unique in your campaign results and learn something new. These campaigns do seem to act differently from artist to artist and with different genres. So I'm learning new things about the results even since posting this video. cheers, Todd
Great tips, Thank you!
No problem, and thank you for being part of the inspiration for this video and sharing your results in the Band Builder member community. 🙌
This is the exact video I’ve been needing for months. I’ve been on discovery mode for most of this year. I have been adding every song possible every month. It makes so much sense that I’m not seeing the increase I feel like I should be seeing if it’s spreading these streaks across the board on songs that aren’t as popular, thus hurting their chances in the algorithm
Bingo. Glad the video helped you. 🙌
@@BandBuilderAcademy submitted your Google form to get in contact as well
Very informative video, thank you 🙏
You're welcome Alex. Thanks for checking out the channel. Todd
Thank you
No problem 😎
Thank you for posting this video and sharing your insight. So when determining which songs to put forward into Discovery mode are the only metrics that really matter: Saves, Playlist Adds, and Intent Rate? We had a song with ridiculous stream and listener lift percentages but no saves, playlist adds, and 0 intent. Sounds like that would be one to remove the following month?
Hey @midnightsyndicate Thanks for the comment. I checked out your music (really good) and I think your results might be unique due to the types of listeners (from instrumental, movie soundtrack, neoclassical, goth, etc.) These types of genres perform a little different and some are more of a passive (ambient) listening experience, so might explain the lower engagement. Hard to tell for sure without a closer look.
I am currently running a 2-song campaign, which is going well, and was considering throwing the full catalogue (34 total songs) in for next month to see how they perform against each other before cherry picking. I take it you'd recommend keeping the stronger performer of these two and putting in 1-4 others alongside it rather than dumping in the full catalogue? I feel either way would give an accurate read on which songs raise their hands, just maybe not at as high of a return rate in the initial month.
There is something I found in subsequent research and hearing from artists with larger catalogs 30-100 songs or more. If you have a bigger catalog Discovery Mode can handle more volume. So try the strategy you mentioned above.
Another consideration I’m learning about (and sharing) is the Stream Lift Rate. If that ever falls below 65%, it’s time to remove that song from the program ( because it risks not being profitable any more)
Todd
@@toddmccarty5355 perfect. super helpful. Also just curious since we are in our first month... what should I expect the metrics to look like on a track that carries over into its second month? I would expect it would not see the amount of lift it saw from the 28 days of NOT being in the program to its initial month of being in the program, so how do you evaluate that? Or should we expect continual growth?
So, a strategy I think I'm seeing in your spreadsheets is testing a group of 3-5 songs for about 3-4 months and possibly replacing one song in that group each month and then evaluating the intent rate to decide which songs to keep. However, i've noticed that the intent rate varies from month to month on songs. I've had a month where every one of my songs managed to get a +2% intent rate but others where that same song would get 1% or below. Would comparing the intent rate from a song month to month be worthwhile, it seems inconsistent to me.
Hey @PSPOD yes, I think that's a fair summary of the strategy. To answer you last question about comparing the intent rate from a song month to month be worthwhile... I would put more focus on the algorithmic play, and song popularity that you can estimate from your Spotify For Artists analytics. There was intent rate consistency looking at over 50 campaigns. Looking back at the data I do see some cases of fluctuation, but the sample size of streams was far lower than the top performing tracks. So when in doubt trust what you see on Spotify Analytics over the longest time period you can find, and also cross reference those on Apple and TH-cam.
Another few questions, When you find those 4-5 songs that do well in discovery mode do you just keep them on or how long until you need to test another song?
Also, have you noticed discovery mode impact popularity score? I'm not sure if it provides enough streams to say a sub 50k monthly listener account to have a meaningful effect on a song's popularity score, but if we narrow the songs that do work using your recommendations have you seen the score go up? I've noticed older songs with lower popularity scores seem to get more streams than a somewhat newer song with a high popularity score.
Seems to me that a possible strategy would be to discover which one of your songs work well in DM, have that songs in for a few months until their popularity score increases enough to hit Discover Weekly and Radio/Autoplay on its own but are the provided number of streams through DM enough to do so?
I don't have specific analysis on popularity score and haven't tested in that way. If you do learn more after several months of testing, please report back with your findings. Everybody would appreciate that insight.
Running 5 songs in Discovery Mode every month, I would continually keep your best 2 or 3 songs in (as long as they continue to perform). Then use the other 2 or 3 slots as a way to test other songs. Experimentation can help uncover things. I would stick with quality songs though. If your goal is simply to elevate a song's popularity score to get it more algorithm play, that only works if people react to the song "way above average". The algorithm is looking for "amazing" songs. Spotify have plenty of "good" songs. The goal of their algorithm is to surface the ones that perform way above average.
@@BandBuilderAcademyUpdate on this, so I have noticed an 1-3 point increase on popularity score on a song that most like, but wasn't getting the most algorithmic love. This is for a 12-15k monthly listener account. From what I've seen on a 25k monthly listener bad, their monthly listeners went up to about 100k and they opted in every song. The songs before discovery mode weren't getting much algorithmic play at all (12-20k streams after being out for 3 years) but after discovery mode, popularity scores increased from 17s to 30s. It seems like Discovery mode can have a meaningful impact on scores if your listener count is 20k before opting in and letting DM take over.
@@PSPOD thank you for that follow up. Let's enjoy Discovery Mode while it lasts. I don't believe it will be around forever. Also, most of the artists I've analyzed have between 3 and 15 songs eligible. Choosing the best 5 makes sense. But a few artists have 40, 70, 100 tracks eligible. For those artists with larger catalogs, it's fine to put 20 or more songs in Discovery Mode. The amount of tracks you enter should be relative to how many total songs you have.
Thanks for this info! Question the listeners and streams metrics:
-Are these radio/autoplay listeners / streams from the promoted songs in discovery mode ONLY OR are the overall song listeners / streams?
-As such is the lift calculated as radio/autoplay lift OR overall song lift?
All the numbers shown on the Campaign Results report on S4A (and my spreadsheet) are streams coming from Discovery Mode ONLY.
I do see that since publishing my video Spotify added "My Daily Mix" into the equation. Not just "Radio" and "Autoplay". So that's cool.
Also... I pulled this text from their website:
"Discovery Mode will be active in other Spotify Mixes - including Artist, Genre, Decade, and Mood Mixes - in the coming months.
You'll only be charged a commission on streams of your selected songs in Discovery Mode contexts during the campaign period.
All other streams of the same songs in other areas of the platform are commission-free."
support.spotify.com/us/artists/article/discovery-mode-contexts/
Todd
@BandBuilderAcademy thanks for your answers! Other question: have these artists NOT seen less discover weekly plays since relative to when they started using discovery mode?
Because if you gain X streams in discovery mode but lose that same X amount of streams in discover weekly then the net end result is negative because of the 30% commission on discovery mode streams...
@@NewMarshallPlan I have not noticed any decrease in Discovery Mode. I never thought to look at that. I see no reason why Discovery Mode would have any negative affect on Discover Weekly. Discover Weekly is it's own detached algorithm from the others. It has a very specific purpose. If a particular song is performing well in Discovery Mode, it would only have positive effects on other algorithmic functions on Spotify. They're just looking for songs that perform exceedingly well. And if they do, the algorithm pushes it. It's not like they would say, "we're giving them more money from Discovery Mode so we're going to take away their streams on Discover Weekly".
@@BandBuilderAcademy thanks for your answer! I am asking because I remember seeing a video (but forget which) where someone actually mentioned this happening...
@@BandBuilderAcademy I found the video mentioning this effect, it's from Tom DuPree III. Curious what you think about it: th-cam.com/video/AcfnKb9emSM/w-d-xo.html
I wonder if including tons of lesser performing songs (everything and the kitchen sink approach) could also have a negative impact by increasing an artists overall skip rate. What do you think?
Hey @MsDrwalker Good point. Songs are always being tested, so if you know certain songs haven't performed well in the best of circumstances, they won't perform well in Discovery Mode. An exception I can think of is if an artist had a great song a while back, but it ended up on some bad playlist (one of these bot farm type lists) it most likely killed the song in the algorithm. So after some time has gone by, this might be a fresh way to test that song in ideal organic circumstances.
I have many songs, but it won't work with any of my new releases, only ever have 3-4 songs eligible, which sucks because only 1 of them is indicative of the sound I am now pushing and the album it's part of. Any tips on how to get your really good new songs eligible? :p
You'll need to promote your newer songs to get them to a bigger level in the algorithm. I recommend you join Band Builder and you can learn all the ways to get your new tracks going in the algorithm.
I was just wondering about the costs, will I get a bill in year that I cant pay?
No you wont get a bill, and it doesn't work like that. It's not a hard cost that you pay. Spotify deduct it from royalties that the pay to your distributor. And in every case you will earn more money than you did - if you were not "in the program". The volume of your streams will increase each month you are in the program, but you make slightly less per/stream. Watch the video again or visit the Spotify website in the video description for full details.
@@BandBuilderAcademy wow man thanks you really clarified on the issue for me, now I get how the tool works and won’t be worried to use it in the future anymore 🙏🏾♥️
@@JASSEY.06 glad to help
Spotify is now dumb stupid they wannt to take the royalties off of songs that are under 1k streams
And the discovery mode is only for people that have allready made it.
Bro with 25k monthly listeners i would be rich.
Nicht From spotify but the sytstem i would use
I give this a thumbs up, because musicians gotta keep speaking up about this and discuss their ideas about a more fair royalty system. Just to make 0.5% more money, Spotify sold out a huge segment of new & young musicians. It's short sighted, and surely there are other ways to make money than taking it from new artists. Wall Street greed is the driving force, but Spotify leadership is where the blame goes.
@@BandBuilderAcademy right now i got so many songs at 800 and 900 streams that is adding up i can buy a few pizzas from them having that removed is big.
Thats just to make it harder for peolle to blow up and its sad.
There is enough space to have everyone famous 😅
Well with 25k listeners you make around 100-130€. I wouldnt call that being rich to be honest. But yes, dicovery mode doesnt help small artists, with that i agree. But its also kinda your work to grow with your music. Music promotion is as important as the quality of music. I always spend every penny i had into promotion, tried pretty much everything and now after some years, i got into the discovery mode. even without 25k listeners! So a lot of work and dedication always pays off in the long run! So keep it up mate!
@@exenyenot from spotify ive meant with my system 😅