Is Spotify's Discovery Mode Worth It?

Since the launch of Discovery Mode in November 2020, I've not paid any attention to this opt-in service that Spotify offers to artists. Recently however, I decided to experiment, in order to see if there are any positives to be found.

What is Discovery Mode?

In a nutshell, Discovery Mode is Spotify's own take on pay-for-plays. They claim that Discovery Mode will highlight/boost your presence in algorithmically generated playlists, earning more streams and artist-profile growth in return for a 30% share of that track's streaming revenue.

The somewhat obvious catch however is that Discovery Mode can only be worthwhile if the opted-in track's revenue exceeds its previous earnings by more than 30%. For example, if you only achieve a 20% increase in streams, your earnings will be down by 10% (compared the previous month), due to Spotify taking 30% of Discovery Mode's revenues. In other words, it's a bit of "risk/reward" a gamble.

More nuanced however are other factors such as whether an increase in streams (of the Discovery Mode-enabled track) will attract more listeners to stream your other music (which is not Discovery Mode-enabled). If that were to happen, whilst the Discovery Mode-enabled track might not achieve a notable difference in earnings (whilst in Discovery Mode), the net benefits to your account as a whole (follower-count and overall stream-count) may result in more favourable earnings. Therefore, whilst you might take a 30% hit on that track's earning, your account as a whole may see a positive uptick in streams, followers, likes, etc, that more than makes up for the 30% drop in revenue from a single track.

The Plan

I intend to take a cautious approach to this risk/reward gamble of Discovery Mode with one of my tracks, in order to learn more about the potential gains that Discovery Mode claims to offer.

I have a fairly large catalogue of music on Spotify and a few of these tracks hit respectable numbers. For this experiments I'm going to leave those tracks alone, as I don't want to risk losing out on any revenues with those tracks.

The rest of my catalogue however, quietly ticks over without generating much in the way of streaming revenue. Therefore, if I pick one, lesser-streamed track to test out Discovery Mode with, it will not matter whether or not it generates a few more, or a few less pennies of revenue. The stats however will hopefully be insightful, regardless of any money earned or lost.

I released the album Sunrise Express on January 24th of this year. In its first month, its title track garnished about 2,500 streams. The bulk of these came in two waves; the weekend it was released, and a secondary (smaller) wave following a mail-out campaign to my BandCamp mailing list (approximately 4000 inboxes), to encourage Spotify streaming.

The above diagram shows that after an initial spike in interest (followed by a secondary spike that coincided with my mailing-list email), there is a steady decline in this track's streaming through February.

If Discovery Mode does what it claims, then this track should theoretically see an upturn in its streaming stats through March. I will also be able to ascertain what, if any impact Discover Mode has on my wider profile, followers and subsequent streams of other tracks in my catalogue. I will check my data at the end of March, endeavour to unpick any noteworthy findings, and report back...

The stats are collated, so let's dig into them for more insight...

Total Streams (organic & Discovery Mode)

The diagram below shows that my track Sunrise Express earned 548 streams in March. This is approximately $1.65. 30% of this track's earnings (approximately $0.50) go directly to Spotify, as payment for being in Discovery Mode.

Discovery Mode Only Streams

Of the 548 streams mentioned above, only 92 of those were as a result of Discovery Mode. Let's break this down...

• 548 streams is approximately $1.65 in streaming revenue

• Discovery Mode cost 30% of the streaming revenues from Sunrise Express

• 30% of 548 is 164 streams

• 164 streams would earn approximately $0.50 - This is what Discovery Mode cost me

• My Discovery Mode earned 92 streams

• 92 streams earns approximately $0.28 - This is what Discovery Mode earned for me

Therefore, my Discovery Mode achieved a supposed 17% increase in streams, as a cost of 30% of its revenues. I.e. it cost me $0.50 to generate $0.28, meaning I lost $0.22 using Discovery Mode.

More concerning however is that the streams for this track were down significantly on the previous month. In February (which is three days shorter than March) Sunrise Express achieved exactly 1400 streams; almost three times what it achieved in March (when it was in Discovery Mode).

Compare February alone (above) with February and March combined (below) and it's clear to see that March's performance was notably down on the previous month. Considering Sunrise Express was in Discovery Mode for March, I'm simply at a loss as to how or why this may have happened.

However, whilst I'm out of pocket at face value, perhaps being in Discovery Mode yielded other positives and benefits in other areas of my profile, such as likes, playlist adds, followers, etc.

Likes, Playlist Adds, and Followers

My Discovery Mode achieved 66 unique listeners, 25 of which were people who don't typically listen to my music. According to Spotify's own stats, it clearly states that this is a drop in performance by 41% on the previous (organic) month. Another confusing result, considering Discovery Mode is intended to improve and enhance performance over and above a track's organic growth.

Sunrise Express earned one playlist add and no saves whatsoever as a results of Discovery Mode. For context, prior to March, Sunrise Express was in 27 playlists. It's now in 28 playlists. It currently has 25 saves, none of which were achieved by Discovery Mode.

The Discovery Mode stats page doesn't state if the track resulted in any new followers directly. The diagram below shows my total follower count for February and March, and it shows that March's follower count continues a fairly uniform trajectory from February.

In February I gained 48 new followers. In March I gained 49 new follows. Factoring in that March is three days longer than February, that's actually a downturn of -1.5 follower growth. I.e. my Discovery Mode follower-growth performance was worse than my previous (non-Discovery Mode) month.

Conclusion

Below are all of my Spotify streams for February and March. There's no obvious difference between the two months.

Spotify tells me that my overall performance in March was down 3% compared to February. This sort of variance is insignificant and fairly common month-by-month across a year. As mentioned above, Sunrise Express saw a drop in streams of approximately 61% in March (compared to February) and new listener exposure was down 47% (compared to February).

When we consider that this track was supposed to receive preferential treatment in algorithmically generated plays, it's strange and confusing to see its performance continue to decline, during a month of streams that were only 3% down on the previous month.

To be honest, I'm a little disappointed. I went into this (somewhat cynically) expecting to be underwhelmed, but I didn't expect to experience a sense of shock at just how poorly it had performed. It didn't even get close to matching the previous month's "organic" growth 🤷🏻‍♂️.

Spotify (sort of) acknowledged that this track didn't perform well. At the bottom of my Discover Mode stats page it says:

"Unless a song is doing well through other metrics, consider not including it in your next campaign. You can also see if performance changes over the coming months."

Does this mean we should only put music into Discovery Mode that is already performing well? Moreover, is a track that (organically) earned over 2700 plays in its first month not performing well?

On the other hand, I would have thought that if music is performing well, it is in less need of being "discovered" compared to music that isn't performing as well. Maybe I'm missing something here.

Anyway, I think these stats speak for themselves, and that just about wraps up this little experiment. Nothing ventured, nothing gained, I suppose. At least now I know that Spotify's Discovery Mode isn't quite what its marketing would like us to believe, and it is in no way a golden ticket to improved streaming revenue and/or audience growth. Needless to say, I won't be using it again any time soon, and I'll leave you with a screenshot of the overall underwhelming statistical summary.

Side note: If you're reading this and you've personally had positive results with Discover Mode, I would interested to hear about it. Either drop me a message, or hit me up on socials.