What's a reliable survey?

Subject: What's a reliable survey? Author: mom_pop

I've heard that a survey isn't valid unless its done on a random population of 1,000 or more people.

I'd never have the time to read the completed surveys of 1,000 people.

Am I just out of luck or is there another way?

Subject: Re: What's a reliable survey? Author: Dennis S. Vogel

Hi:

This IS a confusing subject. My views on it may not be the most popular, but I'm sure they're realistic.

If you let people pick up a survey on a "Please Take One" basis, some people will say that the results aren't valid because respondents SELF-SELECTED."

Realistically, all respondents self-select because (in a free country), anybody can refuse to answer any/all of the questions. If you approached everybody at random so that everybody had an equal chance of being picked to participate then 1) It would make the results statistically valid; 2) You can safely bet that some won't respond. That wouldn't invalidate the research, otherwise there'd be very few valid research projects.

My point is, that for a small business, if you went through the trouble to make your survey statistically valid, the results would be invalid. Why? Because very few, if any, businesses are going to appeal to everybody in its trade zone unless the products/services are absolute necessities and there are no competitors. So, trying to survey a group that's representative of a whole population is a waste of time, energy and materials.

If you're just starting your business, you should have a good idea before you start who will want and be able to afford what you plan to offer. You should survey THEM ONLY.

If the survey isn't vital to people, not everybody will respond anyway. In many cases, those who don't respond aren't true prospects anyway. They don't respond because they're not interested in the products, services or subject matter the survey is based on. So, it's not worth their time to participate.

If you know that your business will appeal only to blue-collar people, don't survey executives, you ought to know what their answers will generally be anyway.

I doubt that any survey will be totally accurate to the point of being completely definitive. The results should be INDICATIVE.In other words, the results should indicate what a business should do and to whom it should make its offers.

If you don't know for sure if some people will want and be able to pay for what you offer, then be sure that you survey enough of them. If you need a customer-base of 200, and there are 2,000 prospects, don't just randomly survey 20 of them.

NOTE - Representative means the selected people are much like everybody in the group as a whole.

A randomly selected, representative group is chosen because surveying the whole group of thousands or more people isn't affordable. If your target market is small, do your best to reach all or enough of them to find out if your business will be profitable.

For example: If I survey 100 out of 300 small businesses to find out who may be true prospects for my service, I wouldn't care if 80 respond negatively or not at all, if I get 20 clients. 20 clients would be all I could handle at a time.

Realistically, I'd need to reach out to a few more prospects than I need just in case I lose a few clients along the way.

My conclusion is - The research is successful if we get the results we need, even if they're not the results we wanted.

Thank for the question.

Thank you for using my blog. Please let me know if I should clarify anything.

Copyright 2016 Dennis S. Vogel All rights reserved.
When you compete against big businesses with big budgets you need powerful marketing strategies & tactics. You'll find them here-
https://thriving-small-businesses.blogspot.com/
http://www.voy.com/31049/