The numbers
US$2.25M priced seed.
We plan our cash on the Benchmark case, which uses rates near published industry medians. On Benchmark the raise lasts to month 26. We raise the Series A on six gates read by month 18.
Revenue and cash
| 36 months from January 2027 | Benchmark | Plan |
|---|---|---|
| Revenue 2027 | US$234k | US$571k |
| Revenue 2028 | US$273k | US$1.32M |
| Revenue 2029 | US$284k | US$2.14M |
| Monthly players, month 18 | 38.7k | 114k |
| Recurring run-rate, month 36 | US$258k | US$2.39M |
| Cash runs out, with no Series A | Month 27 | Never |
Benchmark sets every driver near its published median, and we plan our cash on it. It doesn’t reach Series A numbers on its own. Plan is our top-decile target, not a forecast. It is near breakeven in 2029 on the seed alone. Tax credits and a Series A are left out of every case. The floor is Election alone, with nothing else working: margin stays negative and cash lasts to month 32, which is our signal to cut free play or stop.
Most 2029 revenue comes from the tiers
- US$1.87M Platform tiers: Basic, Pro and credits, with Max from March 2029 (Plan case)
- US$226k Election on Steam
- US$43k Mob Machine contributions
Every paid tier earns well above its cost to serve
| Per player, per month | Price | AI compute |
|---|---|---|
| Free | US$0 | About US$0.15 |
| Basic | US$7.99 | About US$0.80 |
| Pro | US$19.99 plus credits | About US$8.30 |
| Max | US$49.99 | About US$22 |
Benchmark case. We sell on the web and keep 95% of each payment, not 70%. Platform gross margin is 66% at month 36 on Benchmark and 68% on Plan, after creator earnings.
Use of the raise
- US$1.58M Team, salaried and contract: three growing to nine
- US$222k Launches, community and paid acquisition
- US$224k Contingency, 9%
- US$166k Overheads: tools, legal, travel, German subsidiary
- US$144k Freelance art, audio and testing
- US$64k AI and development compute, after credits
Benchmark case, months 1 to 24: US$2.44M of spending, partly covered by about US$0.5M of revenue.
What we prove, and when
- Cost. Election at 5 cents or less a player-hour by month 6. Time Bandits at US$3.44 or less an era by June 2027, or it stops.
- Retention and sharing. 15%+ of Election players still playing at day 7, and 8%+ at day 30. 8% of runs shared, and 30% of new players arriving by a shared link.
- Trending. 15%+ of play-hours on worlds another player generated or remixed, by month 15, with plays per shared world reported monthly.
- Scale, by month 18. 100k monthly players on one title or US$0.6M+ recurring run-rate, with 3% of players paying.
Where it leads
Series A trigger, month 18
≤US$3.44measured cost per Time Bandits era
100kmonthly players on one title
3%of players paying
The lab path
Each game that passes its gates joins the same tiers. Players remix on Max. The Plan case reaches about US$62M of revenue in FY2032. At 5 to 8x that revenue, a sale is worth about US$0.31 to 0.50B (estimate).
Franchise path
A breakout becomes a series, with a partner or as an IP sale.
Category exits, for scale, not comparables: Tencent and partners paid US$8.6B for 84% of Supercell in 2016 (CNBC), and Sega paid €706M for Rovio in 2023 (VGC).
Questions investors ask
How is the round structured?
A US$2.25M priced seed, flexible to US$2.5M, at a proposed US$9M pre-money, targeted to close in March 2027. SAFEs taken before the close convert into it at the round price.
Why plan on benchmark rates?
So the money lasts even if Election only performs like an average AI app. At those rates the raise lasts to month 26, and the Series A starts at months 15 to 18.
What are the biggest risks?
Cost per era is modelled, not yet measured. Free plays are the largest cost, so they are capped. People make the story and art direction, and we say where AI is used. No real faces or voices in Election. Publishing and cash-out are for verified adults.
How does Election handle misinformation?
Candidates and parties are fictional. It never makes a real person’s face or voice, it labels what the AI writes, and it follows each state’s AI election rules. Real polls set the scenario; players argue inside the game, not on the record.
What happens at the Series A?
We raise it on six gates read by month 18: cost, sharing, trending, retention, a second game in Time Bandits, and scale.
Keep going