Decision density in media buying: why the number of signals isn’t the point

Decision density is the number of live signals a system actually weighs together before it makes a single media buying decision. It’s different from the amount of data the system has access to. A higher number doesn’t prove a better decision on its own: what matters is whether those signals shape the same choice at the same moment, and whether the system acts on the combination.

What decision density means

Decision density measures how many signals inform one decision at the moment it’s made, not how many signals exist somewhere in a dashboard.

A bidding rule that pauses a campaign when CPA crosses a threshold is making a decision. A system that weighs 200+ variables before touching that same budget is also making a decision. Both are real. The difference isn’t how much data sits behind each one, it’s how many of those signals feed the same choice at the same time.

Rule-based decisions vs multi-signal decisions

A rule checks one signal against a threshold, while a multi-signal decision weighs several possible causes before acting. That changes what the decision can explain.

Rule-based decisionMulti-signal decision
Question it asksHas this number crossed this line?Why is this number moving, and is it worth acting on now?
What it can doPause, raise or lowerDiagnose, then reallocate or wait
What it can’t seeAny cause outside the one signalSignals it isn’t connected to

Take a rising CPA. A rule sees only the rise. A multi-signal decision can weigh whether it’s seasonality, a competitor’s bidding, creative fatigue or a shift in audience quality, and which of those deserves a response right now.

Why more variables don’t automatically mean better decisions

Counting variables proves nothing by itself. A system logging 200 signals that sit unused isn’t denser than one using 20 that actually shape the outcome.

Decision density becomes real when three things are true: every signal genuinely feeds the same decision, each one is weighed against the others rather than checked in isolation, and the system acts on the combination instead of just having access to it. The number describes an architecture. It doesn’t guarantee a better call.

How to evaluate decision density in any media buying system

Any provider that quotes a number of variables should be able to answer three questions without revealing proprietary logic.

Which signals feed a single decision, as opposed to which signals are collected. Whether those signals are weighed together or checked one after another. And whether you can see, for a specific decision, which signals moved it. If the answers stay at the level of “we use a lot of data,” the number is marketing, not density.

Decision density at Mainkore

Mainkore evaluates 200+ variables per decision, in around 20ms, weighed against each other rather than checked one at a time. Our position is that the useful test isn’t how many variables a system uses, it’s whether it can tell you why a number moved.

That’s the practical difference between a system that pauses a campaign and one that diagnoses it. Decision density belongs to the same family as how often a system actually optimizes and whether it decides across channels: all three describe what a system can really process, not just how fast it moves.

FAQ

What is decision density in media buying?

The number of live signals a system weighs together before making a single decision, not the number of signals available in its dashboard.

Does a higher number of variables mean a better decision?

Not by itself. What matters is whether those signals genuinely inform the same choice, how they’re weighed against each other, and whether the system acts on the combination.

How many variables does Mainkore use per decision?

Mainkore evaluates 200+ variables per decision, in around 20ms, weighed together rather than one by one.

Decision density is one of the eight criteria in our media buying models guide. Want to see what goes into one decision on your campaigns? See how Mainkore makes a decision.

Mainkore. Signals weighed together, not counted separately.