Brand perception questions ask something satisfaction questions do not: not how the last interaction went, but what people think you are, who they think you are for, and how you compare with the alternatives they considered. Asked of your own customers, that is answerable in a few honest questions. Asked of a market you do not have access to, it becomes a research project, and this is a good moment to be clear about which one you are doing.
Ask your own customers, and know what that limits
You can survey the people who bought, and they will tell you what they thought before they did and what they think now. What you cannot learn from them is why the people who did not buy stayed away, which is the question a research panel exists to answer. Being explicit about that in the write-up stops the results being read as something broader than they are.
The questions that do the work
What would you say we do, in your own words. What made you choose us over the alternative you were considering. Who would you tell not to bother with us. The first tells you whether your positioning survives contact with a customer, the second gives you the comparison set you are actually in, and the third is the most useful and least asked of the three.
Positioning is a comparison, so ask for one
A brand does not sit alone in anybody's head; it sits next to two or three others. Asking what else they considered and what made the difference produces the map you were trying to draw, in language you can reuse. Asking people to rate abstract attributes on a scale produces averages that read well in a deck and change nothing.
Reading the answers without over-reading them
Twenty or thirty open answers from real customers will show a handful of repeated phrases, and those phrases are the finding. Resist turning them into percentages: the sample is your own customers, it is small, and a number implies a precision that is not there. Quote the phrases and say who said them.
Questions people ask about brand perception survey questions
How is this different from a satisfaction survey?
Satisfaction asks how it went. Perception asks what people think you are and who they think you are for, which changes far more slowly and is answered in words rather than scores.
How many people do we need to ask?
Fewer than for a score. Twenty or thirty considered answers from your own customers will show the repeated phrases; precision is not what this exercise produces.
Can we survey people who did not buy?
Rarely, without a panel, and that is the honest limit of doing this yourself. Say so when you present the results.
Should the answers be percentages?
No. Small samples of your own customers do not support percentages. Quote the language and say where it came from.