When Three Stakeholders Give Three Different Notes on the Same Visual, Who Is Actually Right?
Summary: Luka Fabry discusses the common challenge of managing differing feedback from founders, marketers, and designers on visual projects. The focus is on transforming subjective comments into testable decisions by creating a decision log that clarifies the purpose and constraints of the visuals. This approach encourages controlled comparisons to identify the best direction based on evidence rather than personal preference. The piece also emphasizes the importance of resolving foundational questions about audience and legal requirements before producing new visual alternatives, thus fostering more effective and accountable visual reviews.
Imagine a landing-page review where a founder asks for a more approachable visual, a marketer asks for a more premium one, and a designer says the composition is the real problem. None of those comments is necessarily wrong. They describe different concerns, but a team can easily treat them as competing answers to one question. The meeting then becomes a contest of confidence instead of a decision about what the audience needs.
The weak response is to keep revising the same image until nobody objects strongly. That often removes the clearest parts of the idea without resolving the original disagreement. A better response is to turn each comment into a testable decision. The team does not need to agree on adjectives first. It needs to define what evidence would make one direction more useful than another.
Before producing another visual, write a short decision log. It should be small enough to read during the next review:
Decision: What must this visual help the audience understand?
Constraint: What cannot change, such as the product context, brand color, or page layout?
Open question: Which part of the current direction is uncertain?
Comparison: What two or three alternatives would isolate that uncertainty?
Evidence: What observation would justify choosing one alternative?
Next action: Who records the choice and what happens after it?
This log changes the meaning of a request such as "make it more premium." The team can ask whether premium refers to lighting, typography, framing, subject matter, or the amount of information on screen. Each interpretation leads to a different visual change. Writing those interpretations down prevents a vague preference from becoming an endless revision request.
The next step is not to generate a large batch. Three controlled directions are usually enough for a useful discussion. One can change tone while keeping composition fixed. Another can change composition while keeping tone fixed. A third can preserve both and change only the visual hierarchy. The goal is not variety for its own sake. It is to isolate the decision so reviewers can explain why one direction serves the brief better.
Reviewers should then respond to the same questions. Does the visual communicate the intended category without extra explanation? Is the main action or object clear? Does the direction fit the audience named in the brief? What would be lost if this option were chosen? Shared questions make feedback comparable. They also make it easier to identify comments that are personal preferences rather than constraints tied to the product or audience.
The decision log should preserve rejected options as well as the chosen one. A single sentence is enough: "Option B was rejected because its tighter framing hid the interaction that the page needs to explain." That note keeps the next review from reopening the same debate and gives future contributors context that a folder of unnamed images cannot provide.
An image-generation tool can be a limited component in this process when a team needs a few controlled directions before committing design time. According to its current product page, Nano Banana 2 Lite accepts text and reference-image inputs. That makes it one possible way to externalize a disputed brief, but the tool does not decide which direction is correct. The decision still depends on the audience, the page goal, and human review. Current capabilities and usage terms should be checked on the product page at the time of use.
There are also cases where generating alternatives is the wrong move. If the team has not agreed on the audience or the claim the page is making, more images only create more surfaces for disagreement. If legal, accessibility, or brand requirements are unresolved, those constraints should be settled before visual exploration begins. The decision log is useful because it exposes those missing inputs early.
The practical question for founders is not how to eliminate subjective feedback. It is how to stop subjective feedback from hiding the actual decision. A small log, a controlled comparison, and a recorded reason for the choice can make the next visual review shorter and more accountable. How are other teams documenting visual decisions when founder, marketing, and design feedback point in different directions?
Nano Banana 2 Lite official website homepage showing the product interface and primary workflow
.png?_a=BAVMn6DY0){kind=small}