Key Takeaways
- A global publisher found there is no single right length for branded content. It depends on the funnel stage.
- A creator agency found that audience size alone barely predicted how much a creator moved people.
- Both answers came from the same approach: start with the decision, list the tested work, add your own data, and go one layer at a time.
- The findings are examples. The approach is the part you can reuse.
Case 1: a global publisher asks, "How long should our branded content be?"
What Ellie knows. The funnel results for each creative, its length, and how viewers reacted second by second.
The approach. The team asked Ellie to list their tested creatives. Then they compared results by length, one funnel stage at a time. Finally, they looked inside the creatives at the moments where viewers drifted and where surprise landed.
What the client added. Their own objective for each piece.
The finding. There is no single right length.
- Short creatives built awareness just as well as long ones.
- Longer creatives earned their time lower in the funnel.
- Creatives where surprise rose near the end got about twice the lift.
So the question changed from "how long?" to "what is this piece for, and does the ending earn it?"
Want the second-by-second view yourself? Read about Ellie's respondent-level biometrics skill.
Case 2: a creator agency asks, "Which creators should we back next?"
What Ellie knows. How far each tested post moved people through the funnel, and its Stopping Power™ and Staying Power™.
The approach. The team pulled the results for every post they had tested, then compared the tested creators side by side.
What the client added. Views and fees for each post.
The finding. Audience size alone barely predicted impact.
- Some creators reached a lot of people but moved few of them.
- Others moved people well but needed more reach behind them.
That gives the agency two different conversations: where to put more reach behind a creator, and where a big following isn't paying its way.
A note on scope: this works across the creators you have tested with us. Ellie compares the work she has measured; she doesn't rank creators she hasn't seen.
What's the approach behind both?
- Start with the decision. Not "what does the data say?" but "what do we have to decide in the next few weeks?"
- Get the full list of tested work first. Ask Ellie what you've tested before asking what it means.
- Add your own data. Budgets, fees, views, objectives. Ellie knows how people responded. You know what it cost and what it was for.
- Go one layer at a time. One funnel stage, one comparison, then the next.
Your answer will be your own
Neither finding is a rule for every brand. A different publisher could land on a different length, and a different roster could tell a different story. The approach carries over, so the answer you get is yours.
Get onboarded
Ellie is your decision partner for marketing. She works inside Claude, ChatGPT, Gemini Enterprise or any other MCP-enabled AI tool, on top of the creatives you've already tested with us. Bring one decision you have to make in the next two weeks, and we'll work through it with you.
Get onboarded