The brief starts with a hypothesis
A request such as “make a beautiful post” provides no useful criteria for creation or review. A useful brief states the hypothesis to explore: which problem to name, who the message is for, and what shift in understanding is expected. The asset may be organic or paid, but its job must be clear before anyone chooses the color, image, duration, or aspect ratio. For each topic, record the evidence and assign an owner before turning an observation into a decision; this reduces assumptions in the next review.
Frame the hypothesis as a testable relationship without promising an outcome. For example, explaining a process may address a recurring question, while a demonstration may help someone assess fit. Then state what will be observed and what the evidence cannot establish. This gives the team direction without treating the publication as an experiment with a predetermined result. When a topic changes, compare the recorded source and conditions with the observed outcome so the update does not rely on memory alone.
- Name the problem, audience, and viewing context.
- Describe the expected shift in understanding.
- Define the signals of learning and the limits of any conclusion.
- Separate organic content from media goal.
Provide enough source material for the message
The creative team should not have to invent evidence, timelines, or service features. Provide the offer, scope, image sources, authorized examples, legal restrictions, and questions the audience commonly asks. If something has not been confirmed, mark it as pending. This transparency prevents design from solving an information gap by creating an attractive but false phrase. For the theme “Provide enough source material for the message”, register the evidence and the responsible before turning reading into decision; this reduces assumptions in the next review.
The brief should also state what the asset must not do. It may be inappropriate to promise outcomes, depict a setting that does not exist, use an unauthorized testimonial, or label every interaction an opportunity. Clear limits do not diminish creativity; they help the team find expressive metaphors and compositions without overstating the available evidence. When a topic changes, compare the recorded source and conditions with the observed outcome so the update does not rely on memory alone.
- Attach offer, source and authorization of assets.
- Resolve or flag outstanding facts before final copy.
- List promises, images and forbidden terms.
- Explain what the asset must not imply.
Specify the experience, not just the file
A testable asset must account for where it will appear, how long it will be viewed, whether sound is available, its dimensions, and the action users can take. Define the opening message, visual sequence, supporting copy, caption, and destination. The final file is only one part of the experience, which also includes loading, accessibility, ad context, and continuity on the landing page or in the conversation. For the theme “specify the experience, not only the file”, register the evidence and the responsible before turning reading into decision; this reduces assumptions in the next review.
Do not expect a single creative asset to explain the entire business. Choose one main idea, support it with proportionate evidence, and use the destination to provide more depth. When a format imposes character or aspect-ratio limits, create focused alternatives instead of squeezing a paragraph into an unreadable font. Well-defined constraints usually produce clearer assets that are easier to compare. When a topic changes, compare the recorded source and conditions with the observed outcome so the update does not rely on memory alone.
- Describe the channel, placement, aspect ratio, and sound conditions.
- Define main message, support and destination.
- Include caption, alternative text and necessary variations.
- Prioritize readability over information density.
Review the outcome without conflating signals
After publishing, compare what was observed with the original hypothesis. Views, clicks, responses, and conversations are distinct signals; none proves that the asset drove a sale on its own. Record the period, audience, media spend, and any simultaneous changes. If the comparison is not controlled, disclose the limitation instead of choosing the most favorable narrative. For the theme “review result without confusing signals”, register the evidence and the responsible before turning reading into decision; this reduces assumptions in the next review.
Creative review also looks at quality. Do comments reveal misunderstandings? Does the destination confirm the message? Did the customer-service team receive useful questions, or did it have to correct expectations? Use these answers to adjust briefing, offer and script. A test does not have to declare a winner; it may reveal that the hypothesis was too broad or that the operation could not support the next step. When a topic changes, compare the recorded source and conditions with the observed outcome so the update does not rely on memory alone.
- Compare the piece to the hypothesis recorded.
- Separate exposure, interaction, inquiry, and qualified opportunity.
- Include qualitative feedback from destination and service.
- Document limits before deciding the next variation.
Reference sources
- Meta — Ads Guide — accessed on 6 September 2026.
- Meta Business Help Center — accessed on 6 September 2026.