What a paying subscriber actually costs you
Every channel on one definition. Signups are free, so the paying number is the only one that counts.
Paid spend, 30 days
£13.8k
Meta and X combined
New subscribers
826
From 11,710 signups
Blended cost per sub
£16.71
Lifetime value is about £20 monthly, £59.99 annual
Best paid channel
Meta
£17.27 per subscriber
What I'd do this week
Ranked by impact. We go through these together every week.
Move £2,000 a month of X budget onto the face creatives on Meta
Meta returns a subscriber at £17.27 against £25.77 on X, and lifetime value is about £20. X is losing money on every one it brings in.
+40 to +75
extra paying subscribers a month
Put real effort into Discord as an acquisition channel, not just a community
It converts 14.0% against X's 6.0%, sends 44% annual buyers, and costs nothing. Your best channel, untreated as one.
3x
conversion rate versus X, at zero media cost
Cut the approval-map static and replace it with two more gameplay clips
£33 a subscriber against a £20 lifetime value, and fatiguing at 3.4x frequency.
-£700
monthly spend released for testing
Trigger a second scenario at the end of a first career
38% of churn lands on players who finished a term with nothing queued next.
30% to 24%
monthly churn, if it lands
Where that gets you in 90 days
Same spend, different allocation.
Current trajectory
826
subscribers / mo
£16.71
per subscriber
30%
monthly churn
If we act on the four above
1,010
subscribers / mo
£13.70
per subscriber
26%
monthly churn
Assumes the budget shift holds its efficiency at scale, which it usually will not hold perfectly, so the real number sits between the two columns. Churn improvement assumes the end-of-career trigger ships. Recalculated every week against what actually happened rather than left as a forecast.
Channel performance
Sorted by cost per paying subscriber. Community channels carry no spend.
| Channel | Spend | Signups | Subscribers | Signup to sub | Cost per subscriber | On annual |
|---|---|---|---|---|---|---|
M Meta Paid | £3.8k | 3,180 | 220 | 6.9% | £17.27 | 22% |
X X Paid | £10k | 6,420 | 388 | 6.0% | £25.77 | 14% |
R Community | None | 1,240 | 96 | 7.7% | No spend | 31% |
D Discord Community | None | 870 | 122 | 14.0% | No spend | 44% |
What this view is for
- Annual buyers are worth about three times a monthly one. Discord sends the highest share of them and costs nothing. X sends the lowest share and costs the most. That gap is invisible without this table.
- Signup volume is a vanity number here. X sends twice the signups Meta does and converts a smaller share of them into subscriptions.
- The decision this drives is where the next thousand dollars goes, reviewed every week on what the table says rather than on which channel feels like it is working.
The hidden cost
£1.6k
Inference spent on the 10,884 signups that never subscribed
Signup is free and every turn fires dozens of AI calls, so a badly targeted click costs twice. Once in ad spend, once in tokens. Cutting the sources that produce free-only players is the fastest saving available.
Concept view on sample data sized against your real inputs: £5.99 monthly or £59.99 annual, roughly 30 percent monthly churn, and the current X budget. Live numbers replace these once the pixel and events are wired.