Beauty DOOH network economics at scale

A beauty DOOH network’s economics look very different at 10 screens, 100 screens and 1,000. The per-screen unit economics are only half the story; the other half is how fixed costs, demand and sales effort behave as you scale. This analysis traces the curve — what gets better with size, what doesn’t, and where the binding constraint actually is.

The per-screen base

Every network’s economics start from the single-screen model: gross ≈ foot traffic × fill rate × CPM × screens, and the operator keeps a share net of the ad-tech haircut and any venue revenue share. That model tells you what one screen earns. But a network isn’t one screen times N — the costs and the demand behave differently as N grows, and that’s where scale economics live.

What gets better with scale

Three things improve as a network grows, and they’re the reason scale matters:

Fixed costs amortise. The CMS, the platform, the ops team, the sales function — these are largely fixed. Spread over 10 screens they’re crushing; over 1,000 they’re a small per-screen cost. The fixed-cost leverage is the core scale economy.
One programmatic integration serves every screen. Wiring up to an anchor SSP is a fixed effort that makes all your inventory biddable — so the marginal screen reaches the same demand pool at near-zero added integration cost.
Track record compounds demand. Each completed campaign produces proof of play and a case study; a network with a history closes the next advertiser more easily than a network with none. Demand begets credibility begets demand — the flywheel that escapes cold-start.

What doesn’t scale linearly

But not everything gets cheaper with size — and ignoring this is how networks over-build:

Per-venue direct sales. Selling a specific advertiser a specific cluster is human, relationship-heavy work that scales roughly with effort, not with screen count. Programmatic helps, but premium direct deals don’t automate.
Venue acquisition. Signing salons one at a time — negotiating terms, installing, supporting — is linear, hands-on work. More screens means more venues means more of this.
Ops at the edge. Connectivity, uptime, hardware failures — these grow with the fleet. A >95% uptime requirement across 1,000 screens is materially harder than across 10.

So scale buys you fixed-cost leverage and demand compounding, but it doesn’t free you from the linear, human cost of acquiring venues and selling direct.

The binding constraint: demand, not screens

The single most important truth about network economics: the constraint is demand, not supply. Installing screens creates capacity; the lever that decides revenue is how many slots are live and sold, and early fill is low and lumpy. It’s tempting to scale screen count because installing is visible and controllable — but a network of 1,000 lightly-sold screens is worse than 200 well-sold ones. The discipline is to grow supply only as fast as you can grow sold demand to fill it, which is why the cold-start lesson — seed supply, then manufacture proof and demand — governs the scaling curve too.

The curve, stylised

Stage	Screens	Character
Test	~10	Prove the model; fixed costs dominate; direct/founding advertisers
Cluster	~100	Density in a market; programmatic wired up; fixed costs start to amortise
Network	~1,000	Fixed-cost leverage real; demand compounding; ops at the edge the new challenge

This isn’t hypothetical. The thesis is proven at scale: the live network this research draws on runs 1,000+ displays across 5 cities and 2 countries, serving ~22M+ impressions a month — a working demonstration that beauty venues monetise, advertisers buy, and the economics hold as the network grows. (The capital and operating breakdown is in How much does it cost to start a network?.)

The takeaway

A beauty DOOH network’s economics improve with scale — but only on the dimensions that actually scale (fixed-cost leverage, one programmatic integration, compounding demand), not on the human ones (per-venue sales, venue acquisition, edge ops). And the binding constraint throughout is demand, not screen count: grow supply to match sold demand, not ahead of it. The networks that work at 1,000 screens are the ones that earned the demand to fill them — not the ones that simply installed the most.