Three Ways Publishers Go Programmatic — And Why the Choice Matters More Than the Tech
Most publishers don’t struggle with programmatic advertising because of bad technology.
They struggle because they never made a clear decision about what programmatic is for.
When teams talk about “going programmatic,” the conversation usually jumps straight to tactics: SSP selection, price floors, auctions, header bidding setups, or data integrations. But those are second-order decisions. The more fundamental question is this:
What role do you actually want programmatic to play in your business?
Programmatic isn’t a switch you flip. It’s an operating model. And if you don’t choose that model deliberately, you almost always end up with something accidental — a setup that creates internal tension, confusing price signals, and mixed revenue outcomes.
In practice, there are three distinct ways publishers run programmatic. Each can work. Each can fail. The problems begin when the model you’re running doesn’t match the model you think you’re running.
Model 1: Programmatic as Insurance
This is the model that dominated digital media for much of the period leading up to 2015, and it still exists today — particularly in premium, niche, or developing environments like digital out-of-home, audio, or podcasts.
The core belief behind this model is simple:
Inventory creates the most value when it’s sold through direct relationships.
In this world, the sales team is the business. They control access, protect pricing, and manage advertiser relationships. Programmatic exists primarily as a backstop — a way to monetize impressions that don’t get sold directly.
Programmatic here is not expected to set prices or discover value. Its pricing power is intentionally low. It sees limited inventory, limited data, and often limited transparency. Clearance typically follows a waterfall, with direct sales always taking priority.
This model is internally consistent and can work well in the right context. But it comes with tradeoffs:
- Limited exposure to broader market demand
- Weak pricing signals
- A risk of pricing discipline eroding before revenue does
If the market moves and demand shifts, this model often reacts slowly — because the feedback loop is constrained.
Model 2: Programmatic as the Business
At the other end of the spectrum is a model built on a very different belief:
Inventory creates the most value when it’s exposed to as much demand as possible and the market is allowed to set the price.
This is the model used by many of the largest digital platforms. Programmatic isn’t a backup plan here — it is the primary revenue engine.
Sales doesn’t disappear, but its role changes dramatically. Instead of clearing inventory deal by deal, sales focuses on:
- Managing the largest strategic relationships
- Shaping products
- Supporting the ecosystem around the auction
In this model:
- Programmatic competes directly with managed sales
- Pricing is discovered continuously in real time
- Market exposure is broad, often reaching tens or hundreds of thousands of advertisers
- The pricing signal is strong and immediate
When it works, this model is extremely powerful. But it also carries risk. Once everything is exposed to the market, differentiation becomes harder. If performance slips or demand slows, there are fewer levers to pull. The market knows the value of your inventory — and so do you.
Model 3: The Hybrid (Where Most Publishers Live)
If the first two models represent the extremes, most publishers operate somewhere in the middle.
The hybrid model exists for two reasons. Sometimes it’s intentional. More often, it’s the result of a series of reasonable decisions made over time.
A publisher might start with a strong direct sales model and gradually open up programmatic to capture incremental demand. Or they might aspire to a fully market-driven approach but lack the data, technology, or internal resources to run it end to end, relying instead on partners and partial solutions.
The hybrid model tries to balance:
- Relationship-led selling
- Market-based price discovery
It can work — and work well — but it’s the most complex to operate.
Pricing signals are often mixed. Market exposure varies. Incentives across sales, ad ops, product, and leadership are rarely perfectly aligned. Many decisions start with “it depends,” which isn’t a flaw — but it does require a much clearer story around yield management than most organizations have.
When this model breaks, it’s usually not because of technology. It’s because decision-making slows, internal friction increases, and different teams start optimizing for different versions of the truth.
The Real Risk: Misalignment, Not Mechanics
Every model described here can succeed. Every model can fail.
The real danger isn’t picking the “wrong” model. It’s running one model while believing you’re running another — or having sales, ad ops, and leadership each implicitly operating under different assumptions.
That’s usually when programmatic gets blamed.
But in most cases, the underlying issue is simpler: the operating model was never made explicit, and the hard conversations about tradeoffs were never actually had.
A Final Thought
Programmatic strategy isn’t about copying what the biggest platforms are doing, or about picking the “best” technology. It’s about being honest — with yourself and with your team — about what creates value for your business and for your customers.
If this framework helps you recognize the model you’re running today, or the source of friction you’ve been feeling, then it’s done its job.
And if you’re operating in that messy middle ground — trying to balance relationships with market signals — those are exactly the kinds of problems I enjoy working through.
James Deaker
The Yield Doctor