Spotify Marquee and Discovery Mode: ROI Calculation for Independent Artists

What Marquee and Discovery Mode actually do

Marquee is a paid recommendation unit that appears in the Spotify Home feed as a full-screen takeover to listeners likely to engage with your track, while Discovery Mode is a discounted-revenue promotion that boosts your track in algorithmic surfaces in exchange for a 30% royalty haircut on streams generated during the campaign window.

Marquee launched in 2019 and is now the dominant paid-acquisition tool for independent labels and self-releasing artists on Spotify. You set a budget (minimum $100, practical minimum $250 for any meaningful test), pick a target market, and Spotify serves a full-screen recommendation card to listeners whose taste profile suggests they would save or follow the track. You pay per impression, not per click or stream.

Discovery Mode is structurally different. You opt a track in, and Spotify boosts it in Radio, Autoplay, and the listener's daily algorithmic feed — but the boost only applies to listeners who would not have organically encountered the track. In exchange, you accept a 30% reduction in the per-stream royalty on streams generated during the boost window. The catch is that you only know if a stream was "boosted" by looking at the post-campaign report from your distributor.

Both tools are designed for the same goal: generating bridge listeners who will save and continue streaming after the campaign ends. The economic difference is that Marquee costs cash up front, while Discovery Mode costs a share of future royalties.

Marquee ROI: the actual mathematics

Marquee ROI in 2026 depends on three variables you can control (budget, audience size, target market) and two you cannot (organic save rate and post-campaign stream behavior), and the break-even CPM for an independent artist at 100k-500k monthly listeners is roughly $0.35-0.55 per stream delivered.

The standard 2026 CPM (cost per thousand impressions) for Marquee is $5-9 in the US/UK/EU, $3-6 in LATAM, and $2-4 in Southeast Asia and India. Click-through rate on a Marquee impression averages 4-7% — meaning 40-70 clicks per 1000 impressions. Of those clicks, 60-80% convert to a 30-second stream. So 1000 impressions in the US market typically yield 30-55 paid streams at a cost of $5-9.

The cost per paid stream is therefore $0.10-0.30. To break even, the average value of those streams (royalties + future streaming) must exceed that cost. At a $0.003-0.005 per-stream rate (the 2026 average for independent artists), 1000 paid streams recover $3-5. That is far below the $5-9 cost. So Marquee alone does not pay back in royalties.

The case for Marquee is the listener-to-fan conversion that happens after the impression. If 5-8% of paid-stream listeners save the track and 1-2% follow the artist, a $500 Marquee campaign delivering 2500 paid streams typically generates 125-200 saves and 25-50 follows. The lifetime value of those fans (continued streaming, catalog streams, future-release streams) at $0.005/stream over 24 months is roughly $15-40 per fan, meaning a $500 campaign can produce $375-2000 in lifetime value if the math holds.

Discovery Mode ROI: a different calculation

Discovery Mode's ROI is calculated against the 30% royalty reduction during the campaign window, not against cash spend, so the break-even point is a function of incremental streams delivered versus the per-stream royalty given up.

When you opt a track into Discovery Mode, Spotify's algorithm will boost it in algorithmic surfaces, but the boost only applies to streams from listeners who would not have organically encountered the track. Every such stream pays out 30% less than the normal per-stream rate. If the average per-stream rate is $0.004, a Discovery Mode stream pays $0.0028, meaning you forfeit $0.0012 per incremental stream.

The break-even point is reached when the post-campaign stream behavior justifies the royalty sacrifice. The 2024-2025 data from distributors like DistroKid and TuneCore shows that tracks with strong post-campaign retention (saves, follows, continued algorithmic placement) typically see Discovery Mode deliver 15-40% more streams in the 28-day window, with about 60-70% of those being "incremental" (new listeners) and 30-40% being "cannibalized" (listeners who would have streamed anyway).

The net economics: if Discovery Mode delivers 100,000 streams in 28 days and 30,000 are incremental, you forfeit $36 (30k x $0.0012). If those 30,000 incremental streams translate to 1500 new saves and 300 follows that drive another 200,000 streams over the next 12 months ($800 in royalties at $0.004/stream), the campaign is a 22x return on the royalty sacrifice. The risk is when the post-campaign retention is weak — then the 30% haircut is pure loss.

When to use each tool in 2026

Marquee and Discovery Mode are not substitutes — Marquee drives top-of-funnel awareness from cold listeners, while Discovery Mode is a mid-funnel conversion tool for tracks that already have a stable audience but need algorithmic amplification.

Use Marquee when you are launching a new release to a cold audience and have cash budget for testing. The minimum viable test is $250-500 in a single market. Below 50k monthly listeners, Marquee is rarely cost-effective because the algorithm does not have enough taste data to target accurately. The sweet spot is 100k-1M monthly listeners with at least one prior track that performed well.

Use Discovery Mode when you have a track that is already getting some organic algorithmic placement but is plateauing, or when you want to push catalog tracks that have not broken through. Discovery Mode works best for tracks with 10-50k existing monthly streams — enough to give the model training data, not so much that incremental lift is negligible. Avoid using Discovery Mode on a track with fewer than 5k monthly streams; the model does not have enough signal.

Stacking the two tools: a $300 Marquee campaign in week 1-2 of a release, followed by Discovery Mode in week 3-6, is a common pattern. Marquee generates the initial bridge listeners; Discovery Mode amplifies the algorithmic pickup those listeners trigger. The order matters — running Discovery Mode first often cannibalizes organic streams without delivering the cold-audience boost Marquee provides.

Benchmarks and 2026 data

Industry data from distributors and from Spotify's own 2024-2025 earnings disclosures suggests a 2026 benchmark of $0.10-0.30 per paid stream via Marquee and a 15-40% incremental lift via Discovery Mode for tracks with stable baselines.

Marquee benchmarks in 2026 vary significantly by genre and market. Hip-hop and pop in the US/UK deliver 25-55 paid streams per 1000 impressions at $5-9 CPM. Indie folk and ambient in the US deliver 15-30 paid streams at the same CPM. Electronic and dance music in Germany, Netherlands, and Scandinavia deliver 30-60 paid streams at $4-7 CPM. The genre-market combination is the primary determinant of cost per stream.

Discovery Mode benchmarks are harder to source because Spotify does not publish per-track data. Distributor reports from 2024-2025 show median incremental lift of 22% in the 28-day campaign window, with a 70th percentile around 40%. Tracks that underperform tend to be those with mismatched metadata (the algorithm cannot find the right listeners) or those in over-saturated genres where the boost is diluted by competing recommendations.

Follower conversion from Marquee campaigns runs at 1-2% of paid streams for most independent artists, with a 2026 industry median of 1.4%. Save rates run at 5-8% of paid streams. These are the metrics to track in Spotify for Artists post-campaign — raw stream counts are misleading because they do not capture the long-tail listener value.

Common mistakes and when not to spend

The most common Marquee mistake is running a campaign before the artist has enough audience signal for the targeting to work, and the most common Discovery Mode mistake is opting in catalog tracks with weak baselines, which leads to royalty losses without meaningful lift.

Do not run Marquee on your first-ever release. The targeting model needs at least 3-5 prior releases with stable monthly listener data to identify your listener cluster. A first-release Marquee campaign typically delivers 50-70% of expected streams because the algorithm is targeting broadly rather than precisely.

Do not run Discovery Mode on tracks with fewer than 5k monthly streams. The model's confidence interval on incremental lift is too wide, and you risk giving up 30% of royalties on streams that would have happened organically. The same is true for tracks that are already spiking — the lift you get is on top of an already-elevated baseline, so the incremental percentage is small relative to the royalty haircut.

Do not stack both tools on the same track in the same window. The signals cannibalize each other, and you end up paying twice for the same incremental streams. The standard pattern is Marquee in week 1-2, then Discovery Mode in week 3-6 after the initial push has stabilized.

Finally, do not evaluate ROI on 28-day stream counts alone. The whole economic case for paid promotion on Spotify is the lifetime value of the listeners you acquire. A Marquee campaign that breaks even in month 1 but delivers 2000 catalog streams over 12 months is a 4x return. The mistake is treating Spotify paid tools like a performance ad platform with short attribution windows.

Marquee vs Discovery Mode vs organic push

Marquee + Discovery Mode rollout sequence

1. Confirm baseline metrics: Pull your Spotify for Artists data: monthly listeners, save rate, and current algorithmic placement. Both tools underperform without stable baseline signal.
2. Choose the lead track: Pick a track with at least 10k existing monthly streams and a save rate above 3%. Tracks that already have some algorithmic placement benefit most from both tools.
3. Set Marquee budget and market: Allocate $250-500 for the first Marquee test. Target one to two markets where your listener concentration is already high to maximize CPM efficiency.
4. Run Marquee in week 1-2 post-release: Launch the Marquee campaign to coincide with editorial playlist consideration. The combined signal improves the model's confidence in your track.
5. Wait 7 days, then opt into Discovery Mode: After Marquee delivers the initial wave, opt the same track into Discovery Mode for 28 days. The two-week gap prevents signal cannibalization.
6. Track save rate and follower conversion: Monitor Spotify for Artists daily. A 5%+ save rate on paid streams and 1-2% follower conversion are healthy signals that the campaign is working.
7. Evaluate lifetime value at day 90: Do not judge ROI on 28-day streams. Calculate lifetime value by looking at catalog streams from acquired listeners over 90 days, then compare against the campaign cost.