When validating a new idea, it's best practice to use a variety of research methods: customer interviews, surveys, and experiments.
Research methods that gather opinions like questionnaires and customer interviews are great ways to discover unmet needs and find potential problems with your idea. But they are terrible at predicting how your idea will perform when it's launched.
The say-do problem
Surveys and interviews are useful for developing product and service ideas but are poor at predicting how good those ideas are because of the say-do problem - psychological studies consistently show what people say and what they actually end up doing are two very different things.
Take shopping carts, for example, when polled about their shopping habits 75% of people said shopping with a cart didn't increase the amount they would spend in store.
Yet when an in-store experiment was conducted researchers found that using a large shopping cart increased people's spending by 40%.
Measure behavior, don't collect opinions
When trying to find out how your idea will perform when it's launched the best approach is to measure potential customers actual behavior in a realistic setting.
This is why Seedling collects behavioral data, do people click the call to action button, how long do they interact with a page, do they request more information... from customers interacting with your idea, opinions can't be trusted when making investment decisions.