How to Score Campaign Ideas So the Best One Actually Wins

Your team keeps choosing the familiar idea. Here's a four-criterion system that changes that.

There's a dynamic I've watched play out in marketing meetings for thirty years. The team brainstorms. Three ideas go on the whiteboard. Everyone looks at them. The room slowly gravitates toward the one that feels most comfortable — usually the one that looks most like something you've already done, because that's the one nobody can argue against.

The best idea in the room doesn't always win the room. The safest idea does.

I'm not saying your team is wrong to be cautious. Campaign resources are real, timelines are real, and a bad bet costs more than a skipped opportunity. But "safest" and "best" are not the same thing, and the way most teams evaluate ideas — vibes, precedent, and whoever speaks most confidently — is not actually measuring what you care about.

Here's a system that does.

The problem with gut selection

Before I describe the system, it's worth naming what gut selection actually measures, because it's not nothing. Experienced marketers have genuine pattern recognition. When something feels wrong, it's often worth pausing on that feeling.

The problem is what gut selection doesn't measure well: effort. In my experience, teams systematically underestimate how hard an idea is to execute, especially when they're excited about it. The high-production video concept that everyone loves at the whiteboard stage looks very different at two weeks before deadline with half the budget already spent.

The other thing gut selection doesn't measure well is differentiation. Familiar ideas feel safe partly because they feel proven — but what made something work last year may not be available to you this year, especially if competitors have since moved into that territory.

Four criteria that actually matter

The scoring system I use with AI-generated campaign concepts has four criteria. Three you add together; one you subtract. That last part is the part most people push back on, so I'll explain it.

Impact measures potential business return. How much could this concept actually move the number you care about — leads generated, revenue influenced, retention improved? A concept with high impact addresses a significant problem your buyer has and positions your offer as the meaningful solution. Score 1 to 5.

Differentiation measures how unique the concept is relative to what your competitors are currently doing. This is where the competitor filter comes in: before you score, you check each concept against what's already running in your category. A concept that occupies genuine whitespace — territory competitors aren't claiming — scores a five. A concept that looks like your main competitor's current campaign scores a one. Score 1 to 5.

Channel Fit measures how well the concept suits the distribution channels you actually have and can execute on. A brilliant interactive quiz concept scores low on channel fit if your primary channel is a weekly email newsletter. A high-production video concept scores low if you don't have video production capacity or distribution. This criterion is a reality check, not a creativity check. Score 1 to 5.

Production Effort is scored inverted — 1 means easy, 5 means hard — and subtracted from the total. This is the criterion people argue with, so here's the logic.

A concept with a combined score of 13 across the first three criteria but a production effort of 5 nets 8. A concept with a combined score of 11 but a production effort of 1 nets 10. The second concept ships. The first one might not — and a campaign that doesn't ship has a real-world value of zero, regardless of how exciting it looked in the meeting.

Subtracting effort is a forcing function that makes the cost of a complex idea explicit instead of abstract. It doesn't eliminate ambitious campaigns; it ensures you're choosing them with full awareness of what they cost.

Using AI to run the scoring

Here's where AI becomes genuinely useful rather than just fast. After a brainstorm session — which I now run with AI generating thirty or more concepts so there's actual range to choose from — I ask the AI to score the filtered set against all four criteria at once.

The prompt is specific: score each concept on Impact, Differentiation, Channel Fit, and Production Effort (where 1 is easiest and 5 is hardest). Calculate a total as the sum minus Effort. Sort the results from highest to lowest. And — this is the part that matters most — include a one-sentence rationale for the effort score.

The rationale for effort is important because effort is the most subjective criterion. What's easy for one team is hard for another. When the AI explains why it scored something a 4 on effort, you can correct it based on your actual resources, your agency relationships, your team's skills. The table surfaces the trade-offs; you adjust based on the reality the AI doesn't know.

What to do with the score

The score is structured input, not a verdict. I want to be direct about this, because there's a version of this system that becomes a way to abdicate judgment to a number rather than to improve judgment.

A concept that scores second on the table but fits a strategic window you know is opening — a competitor's stumble, a seasonal moment, a product launch with runway — may be the right call. The scoring system cannot know that. You can.

What the system eliminates is the meeting where the loudest person wins. It gives everyone the same framework for evaluating trade-offs, it makes the cost of effort explicit before you've committed resources, and it creates a basis for disagreement that goes beyond "I have a good feeling about this one."

That's the goal: not to replace judgment but to give it better material to work with.