Category: Blog
-
Programmatic advertising transparency: what an advertiser should actually be able to audit
Programmatic advertising transparency usually means one thing: knowing where the money goes. How much reaches the publisher, how much stays with intermediaries, which fees apply along the supply chain. That question matters, and the industry has made measurable progress on it. But it is only half of transparency. Cost transparency tells you where your budget went. It doesn’t tell you what you got for it, how that result was measured, or who answers for it if it falls short. Full transparency starts before the campaign runs: an agreed objective, a committed KPI, a clear way to measure it and explicit…
-
What not to build, and how we test what we did
Most of what I hear about AI in advertising falls into two buckets. A smarter bidder inside one channel. Or an agent with the keys to your ad accounts, trying things to see what sticks. When we started building Mainkore, the first decision was what not to build. The obvious move is to put an LLM on top of ad accounts. We didn’t. LLMs are great with language. Media buying is a long chain of decisions under a budget, with feedback that comes late and is measured in results. Knowing the next likely word won’t tell you what to pay…
-
How to compare media buying platforms without comparing the wrong layer
Before comparing media buying platforms, check that they work at the same layer. DSPs and ad platforms are execution layers: they buy inventory inside the environments they’re connected to. An autonomous decision layer works above them and decides across all of them. Putting both in the same feature matrix mixes two different jobs, and leads to a shortlist that answers the wrong question. Execution layer vs decision layer in media buying Media buying technology works at two layers: the execution layer, which buys inside a platform, and the decision layer, which decides how budget and bids should be distributed across…
-
Running several media buying models at once: who owns each decision
Many brands don’t use a single media buying model: they run an in-house team, an agency and programmatic or autonomous execution at the same time. That isn’t a transition problem to be solved by migrating from one model to the next. The real risk is that nobody has defined which layer owns which decision, so rules contradict each other and some decisions get made twice while others aren’t made at all. Why media buying models coexist instead of replacing each other Media buying models coexist because each one holds something the others don’t, so replacing one with another usually means…
-
Does autonomous AI replace your media agency?
No. Autonomous AI takes over the execution decisions that were always mechanical, even when a person made them: bids, budget shifts between channels, pausing a creative that stopped working. The agency keeps the work that needs context a system is never given: strategy, brand judgment, media negotiation, and setting the boundaries the system operates within. What changes is where agency time goes, and how its value should be measured. What autonomous AI takes over from the agency Autonomous AI takes over the layer of execution decisions that follow rules and signals, and that were limited by review cycles rather than…
-
Performance guarantee vs performance target in media buying: who carries the risk
A performance target is a result the provider projects and will try to reach. A performance guarantee is a result written into the contract, with a defined consequence if it isn’t met. The difference comes down to who carries the risk when the number is missed: with a target, the advertiser absorbs the gap, and with a guarantee, the provider does. The difference between a performance target and a performance guarantee A performance target is a best-effort projection, while a performance guarantee is a contractual commitment with a remedy attached. The two are easy to confuse because they’re often described…
-
How to choose a media buying model based on how your organization decides
The right media buying model is the one that matches three things about your organization: how many buying decisions it generates, how fast it can approve a change, and how much risk it accepts in a decision nobody reviewed first. Features come after that. A model that decides faster than your organization can absorb its decisions doesn’t remove the bottleneck, it moves it to the approval chain. Start with your organization, not with the models Choosing a media buying model is first a question about how your organization makes decisions, and only then a question about technology. Two companies with…
-
Walled gardens in advertising: why owning the inventory isn’t the same as deciding well
Walled gardens are advertising ecosystems, such as Google, Meta and Amazon, that control three things at once: the inventory, the data about the people who see it, and the measurement of their own results. Their advantage is real. Logged-in user data and inventory like YouTube or Prime Video can’t be bought anywhere else. Their limit is structural. Each garden optimizes toward what happens inside its own walls, and none of them is built to tell you how much of your budget should go to it rather than to the garden next door. That decision sits one layer above the gardens,…
-
How often does your media buying actually optimize?
Media buying optimizes as often as it’s actually re-evaluated, which is often the human review cycle, weekly or monthly, rather than the real-time capability of the tools involved. Optimization frequency is how often a system really re-evaluates and changes what it’s doing. It’s a different claim from what the system is capable of, and the gap between the two is where performance quietly leaks. Optimization capability vs optimization frequency Optimization capability is what a system can technically do. Optimization frequency is what it actually does, on an ordinary day, without anyone prompting it. “Continuous optimization” is a common phrase in…
-
Cross-channel media buying: where single-channel optimization stops
Cross-channel media buying means deciding budget, bids and audiences across every channel as one connected system, instead of optimizing each channel on its own. Single-channel optimization can work very well inside its environment. The loss happens at the seam between channels, in the decisions no single-channel system can see. What single-channel optimization can and can’t see A DSP optimizes inside the auctions it runs on, and smart bidding optimizes inside its own ecosystem. Both can do that job well. What neither can see is what happens outside that environment. The budget sitting in another channel, the audience already reached through…