Test with a small batch, not a poll: a sale is the only answer you can trust. People answer polls for the person they would like to be, then buy what they always buy. So ask what they did last time, not what they would do, and let what sells decide.
Here is why a yes in a poll is not an order, how to ask better questions, and how to run a small batch that tells you the truth.
A yes in a poll is not an order
Asking feels like research. A candle label puts up a story poll, "Would you buy a fig candle?", and most of the answers say yes. Then the fig candle sits on the shelf while the vanilla sells out, as it always does.
Nobody lied. People say one thing and do another, because they answer for the person they would like to be: adventurous, generous, careful about the planet. At the till they buy what they always buy. That is what a habit is.
Large surveys ask the same kind of question. In PwC's Voice of the Consumer Survey 2024, people said they would pay about 9.7% more for sustainable products. That is a real answer about what people intend. Whether your customers pay more for your product, only your own sales can show.
A poll tells you who your customers would like to be. A small batch tells you who they are.
Ask what they did, not what they would do
The question you ask shapes the answer. "Would you try a fig candle?" invites a kind yes. "What was the last candle you bought, and what scent was it?" gets you a fact.
A skincare shop thinking about a night cream learns more from "What do you put on your skin before bed now?" than from "Would you buy our night cream?" The first answer shows the habit the new cream has to fit into. For more questions like these, and a way to sum up ten conversations with AI, see asking ten customers why they buy.
Run a small batch
Make a small number, sell it the way you sell everything else, and give it a fixed window. For the candle label, that is twenty fig candles, listed in the shop next to the rest of the range and announced in one post and one story. Three weeks, then count.
Sell it at the real price. A discount tests whether people like a bargain, not whether they want the product. And keep it where customers already shop, not on a special page, so it earns its place next to what they usually buy.
Read what the batch tells you
It sold out in the first week. Make more, and look at who bought. Regulars adding it to their usual order means it grows the range. New customers finding you through it means it brings people in. An hour with your recent orders shows which.
It sold slowly. Look at who bought it and whether anyone came back for a second. A slow start that brings repeat orders earns a second batch.
It barely sold. You learned that for the price of twenty candles, not a full production run. That is a good result, and the next idea is ready for its own batch.
A bakery reads a new loaf the same way. Two Saturdays on the counter, and the tray at closing time is the answer.
What CP3® does here
Market Trends shows what your audience is searching for, scored for fit with your brand and refreshed every month. A search is something people do, not something they promise, so it is a better place to find your next product idea than a poll. Making the batch, pricing it and counting what sells stay with you.