In customer validation and development how effective are surveys?

They can be effective--you just have to ask the right questions in the right format. We've done this successfully for several clients looking to find a product/market fit. This is the format we follow:

Introduction
Very succinctly and clearly explain your product concept before asking any questions. What it does, who it's for, how it works. If you've created user stories these are helpful in writing this. Make it no more than 2-3 sentences.

Behavior
Ask question about their current behavior as it relates to your product. For example, how do they current address the problem your product will solve? If it's a consumer app, ask about their habits.

Product Reactions
Talk about the features of your product and ask questions directed around its use. For example, explain a feature then ask the respondent if they would be interested in that particular feature. Position the features as if you're solving a pain point for the user. Example: "If the product had the ability to synch all your leads to every device you own, saving you the process of manual entry..."

Likelihood of Use
This is where you want to be very direct--ask them if they will use the product based on how you presented it. Ask them questions about whether they would be more or less likely to use the product based on how it works. Never ask if they will buy it or how much they will pay--people buy based on emotion--you won't get accurate data. You simply want to know if they like the concept and if they would use the product if it was available to them.

Keep your survey short. There's no rule of thumb on questions but try to keep them to yes/no, multiple choice so they can be answered quickly and make sure respondents can finish the survey in 5 minutes or less. Always give them the ability to provide free-form feedback however if they decide they want to do that. Make it optional. I hope this helps.

Weaknesses of surveys are pretty well know, despite them being a pretty aolid tool to gather information. However, there are a number of biases and just plain issues that you may only control for if you happen to use surveys (non-exhaustive): 1. Confirmation bias - there is a tendency among survey makers to prove themselves or a particular reality as true. Consider going 3rd party or fresh eyes to evaluate your questiin set 2. Self selection - Most ppl who take yiur survey will already be a fan or angle the survey to see their wishes come true. Randomize particilation where you can and normalize results accordongly. 3. Returning surveys - this is the biggest pain; most aurveys arent returned and incentives to participate often distort results. Select multiple channels for aurvey distribution. In my experience, surveys are exactly that, a means to poll for whats out there, where certain issues "spike", and the correlations between those areas. In short they tell you whats worth your attention. Surveys should not be the only method you use to determine product market fit, they just happen to be one of the least costly.i If you are assessing product/market fit, I assume you are working toward an MVP - a test that proves or may validate a belief you have about your product or service. Surveys here are useful in deciding which MVPs to create and allow you to identify externlities that will become success criteria once your MVP launches. Hopefully some of this is helpful.

I suggest you conduct face to face interviews/needs analysis to really understand the details first then use the results to inform the survey questions and the weight you assign to each. you really need to clearly understand who your ideal prospect is, their pains and passions.  i think you'd be more successful first having real conversations with real people then use surveys to scale and refine your  understanding of customer needs.