A while back I was scoping a display rollout for a distributed consumer brand - screens going into dozens of locations. The default framing was simple: pick the hardware, install it, move on. A line item. A cost.
But there was a second way to look at exactly the same project. Those screens would sit in front of an affluent, urban, wellness-motivated audience, several times a week, for ten to fifteen minutes at a time, off their phones. That's not a cost center. That's media inventory, and media inventory is something advertisers pay real money for.
The hardware was going in either way. The only question was whether we'd design it to earn or design it to sit there. That reframe - cost center vs. revenue asset - is the whole game, and it applies to far more than screens.
The default mistake: infrastructure as a cost center
When you treat a deployment as a cost, every decision optimizes for spending less: cheapest hardware that works, simplest software that displays. Nothing wrong with that - except it quietly forecloses the upside. The screens that only ever show your own content can't be retrofitted into a revenue source without ripping out the platform underneath them.
The decisions that unlock monetization aren't expensive. They're just decisions you have to make before you buy, not after. Miss the window and you're looking at custom middleware and months of rework to capture revenue that should have been free.
The uncomfortable truth: someone is already selling your audience
Here's what made the case undeniable. The largest digital out-of-home ad exchanges already sold lookalike audiences for this kind of customer segment - advertisers were paying to reach the same people on other media, using adjacent signals as the targeting hook.
Read that again. The audience value already existed. Brands were already spending against it. We were generating that value and capturing exactly none of it. A media network simply flips the dynamic: instead of your audience being sold elsewhere, you sell access to it, on your own surfaces, on your own terms.
If brands are already paying to reach your audience somewhere else, you're leaving money on a table you built.
The market is already there
None of this is speculative. The demand infrastructure is mature and growing:
$4.7B US DOOH market (2024) · 24% → 65% programmatic share by 2029 · Top 3 wellness as ad category
Programmatic digital out-of-home (pDOOH) is being plugged into the same demand-side platforms that buy connected TV and digital display. Wellness, lifestyle, and retail environments are already major advertiser categories in the channel. And the model is validated - distributed venue networks already run in-location media businesses generating real, recurring revenue. The playbook exists; most operators just haven't run it.
The decision that changes the economics: the CMS is the monetization decision
The single highest-leverage choice in the whole project turned out to be the content-management platform. Choose a CMS with a native integration to a supply-side ad platform (SSP), and monetization becomes a configuration step - set your category blocks, your floor prices, your content-to-ad ratio, and inventory flows to the ad exchanges automatically. No new hardware. No custom engineering.
Choose a CMS without that integration and you've signed up for custom middleware and months of delay before a single dollar comes in. Same screens, same audience - radically different economics, decided entirely by a platform choice made early. That's the kind of leverage you only get if you're thinking about revenue at spec time.
Protect the experience, or you destroy the asset
This is the part most "retail media" pitches get wrong, and it's the part that matters most. The audience is only valuable because the experience is good. The second you degrade the experience to squeeze out ad impressions, you erode the very thing the advertisers are paying for.
So the design has to be disciplined about where monetization is allowed:
- Entry / pre- and post-experience: a controlled ad load (think 30–40% paid, the rest your own content), category exclusions, premium floors.
- The core experience itself: nothing. No ad units where people came to be present. The room is the product; you don't interrupt it.
- Experiential sponsorship: the rare exception - a brand-funded moment that genuinely enhances the experience, packaged as a creative partnership, never as an ad slot.
Constraint here isn't a limitation - it's what keeps the inventory premium. A network that protects its experience commands higher rates than one that floods every surface.
The math, conservatively
Built this way, the numbers are not small. A 3–5 location pilot can produce first revenue inside 90 days, validating the model before any full-scale commitment. At full network scale, a conservative base case - modest fill rates, mid-market CPMs, a mix of programmatic and direct-sold inventory - can land in the seven figures of net annual revenue. The optimistic case, with premium private marketplace deals and category-exclusive sponsors, runs several times that.
And critically: most of that is margin. The hardware was already a sunk cost of the rollout. The incremental cost to monetize is a small operational overhead. The revenue largely offsets the infrastructure that produced it - the deployment pays for itself.
The principle generalizes
Screens are just the clearest example. The underlying principle is the one I bring to every system I build, physical or digital:
Build infrastructure, not features. Every system should produce something other systems - or other budgets - can consume.
A sound-monitoring platform that also feeds coaching dashboards and compliance reports is worth many times a standalone meter. A sensing layer that drives both lighting and analytics earns its keep twice. A display network that shows your content and sells inventory funds its own existence. The most valuable infrastructure is the kind that was designed, from the first spec, to give more back than it costs.
That's the lens. When you're about to treat a deployment as a cost, stop and ask the other question: what would it take to make this thing earn? More often than you'd expect, the answer is a decision, not a budget.
Building something that should earn its keep?
I design AV, sensing, and experiential systems that produce value, not just output. Let's talk about yours.