You know you need to run a survey. You want to get the feedback your organization needs to improve your services. And you know one of the best ways to get feedback and improve your organization is to run a survey. Surveys let you know what your strengths are, and most importantly what your weaknesses are, and where you can improve.

But even if you know all that, it doesn't matter if your survey doesn't return accurate results you can actually use. There's a big difference between running a survey, and running a survey that gets you the information you need.

There are 4 simple principles to keep in mind that will help you create a survey that delivers solid, actionable data.

1. Define Your Goals

This is the absolute first thing you need to do when deciding to launch a survey. You have to know what your goal is - what the burning question you need answered is - before you go any farther.

Gather your team and sit down to decide what your goal for the survey is.

This  goal has to be attainable. Narrow your goal down to something you can realistically know that will improve your business. Instead of "how can we serve our clients better," you would be better served with a specific goal like "what services do our clients find the most room for improvement for?"

Choose that one goal to determine what questions you need to ask in your survey.

2. Craft Your Questions

It's not just what you ask; it's how you ask it.

Your survey respondents are doing you a tremendous favor by taking your survey. So you don't want to waste their valuable time. Keep your surveys as short as possible and don't waste their time. When you send the survey out, be sure and let people know the survey should only take a few minutes or less to answer.

If your survey has to go long, you need to keep in mind that the accuracy of responses may trail off after the first few questions. So you'll want to get your most important questions right at the front.

If you have the option, randomize the order of questions so that you can make sure you're getting an accurate representation of responses.

Make most of your questions closed-ended. These are simple "yes" or "no" questions, or multiple choice, ranked, or checkbox questions. These take less time to answer and leave less room for interpretation.

Free response questions (where the respondent writes out their thoughts in a comment box) are fine, but keep them to a minimum and at the end of your survey.

Most importantly, make sure you don't ask leading questions that might influence how a respondent answers.

3. Determine Your Audience

Your audience matters in more ways than you might think.

For starters, you need to consider your sample size. Generally speaking, the more responses, you have to a survey, the more accurate your data is going to be.

For an example, if you asked a question of two people, and one answered favorably and one negatively, you'd have a 50% favorable to 50% unfavorable rate. But if you sent that same question to 10,000 people, you might find that 8,000 people answered favorably and 2,000 unfavorably. That would be an 80% favorable to 20% unfavorable rate. Now you have a much more accurate picture.

You also need to target your survey to your customer or client base. If you only serve people who are experiencing homelessness in a west coast city, you don't want to send your survey to high-income people on the east coast. You'll be better served by selecting existing or past customers, or people in that same demographic.

If your budget allows it, you might wish to offer an incentive to survey respondents to increase chances of participation and completion.

4. Get Ready for Launch

Now it's time to send out your survey right? Not so fast. Take some time to make sure it's the way you want it. Send your survey to coworkers or family members who weren't involved in creating it and have them go through it like a respondent would. This user testing will let you know if there are any questions that don't make sense or are confusing.

It's also worth having your questions and responses copy-edited by someone. This is one last chance to catch any typos, grammatical errors, or misspellings that spellcheck might have missed.

Finally, send the survey to yourself. If you're sending it as an email, spend time crafting a compelling subject line that will entice people to open it. Mailchimp has a handy guide for subject line best practices. Once you've written some subject lines, check out their scores on Send Check It's subject line tester.

All set? Now it's time to send your subject line and wait for the results to come in.

‍