How to review and test a customer survey - Survey Monkey - Tools and Technology

BottomUp - Skills for Innovators

008 - Preview & Review - Survey Monkey - Tools and Technology

[00:00:00] Mike: [00:00:08] Hello and welcome to the BottomUp podcast. I'm your cohost Mike Parsons, and as always, I am joined by the Survey Monkey himself. Mr. Chad Owen.

[00:00:18] Chad: [00:00:18] Well, here we are in the fourth episode in the Survey Monkey series. We've given you an intro to why quantitative research is important. Talked about, what do you need to do before you even dive into Survey Monkey to prepare and create a research plan, talked a little bit about designing [00:00:34] surveys in the last episode, and here we are with the, my hand is hovering over the button.

[00:00:39] I'm about to launch it to get a thousand responses and oh wait, there's a missing question or a huge typo, or we forgot to upload the image.

[00:00:48] Mike: [00:00:48] Yeah. Hold the press. Just hang on a second. Because before, which will be in our next show, we talk about how to get people into your survey. A tried and tested warrior's advice from Chad and myself to all of our listeners. [00:01:08] Please preview and check your survey before you send it out.

[00:01:12] Chad: [00:01:12] And then check it again and then have someone else check it.

[00:01:16] Mike: [00:01:16] All right, so here's the thing, why this is a bit different to maybe publishing a website. Let's say you've got a typo on your website; you can update it and then you know from now on it's all fixed. The problem when it comes to quant research and doing a survey in Survey Monkey is you've often invited people to fill it out, and once you start changing the questions midstream through recruitment of people to [00:01:42] respond to your survey.

[00:01:43] You are basically invalidating the data because not everybody saw the same set of questions. So, you can't draw a really broad analysis of the data because you essentially have to split all the data up. Doubling down on this, Chad, the big booboo is when you've paid to recruit people in and then you realize you didn't preview and check your survey, and so it's not only cost you time,

[00:02:06] it's costs you some dollars.

[00:02:08] Chad: [00:02:08] Yeah. And if you're sending it to more than a hundred people and you're changing things, or you're getting back data, that could be a several hundred-dollar mistake. [00:02:16] The numbers only just go up if you're sending it out to even more people.

[00:02:19] Mike: [00:02:19] And more specialized people as well. You could easily find yourself spending a couple of thousand on recruitment, so you want to make sure it's right first time.

[00:02:28] Chad: [00:02:28] Yeah, and Survey Monkey actually gives you some really great tools to preview your surveys on not only just desktop computers but mobile and tablet. Mike, I think we probably get over half responses coming from mobile devices. So, there's a great way for you to check that out and preview it straight from within Survey Monkey.

[00:02:45] Mike: [00:02:45] Yeah, because the catch is you're often building this on a computer, on a laptop, but it's [00:02:50] actually often responded to via mobile as much the same in terms of, you know, newsletters and so forth that it's a majority mobile these days. I think the greatest thing to do, simple thing, take the survey yourself.

[00:03:04] And actually you should be in surveymonkey.com if you can be while we're doing this show, so you can actually see the area we're talking about, but there's actually a tab there that runs from left to right that says summary design survey. And then where we are right now is preview and score. Now when you're in your preview and score, you get to see the survey as it [00:03:24] would appear to a respondent, and you can actually do it as a respondent and you can do it on a PC tablet,

[00:03:31] or mobile and it's really good and right next to it, it has this survey score that we've talked about in the previous episode where they give you an estimated completion rate and your estimated completion rate is basically telling you, out of every hundred people that do this, here's the percentage of people we think will actually complete it.

[00:03:50] Because the thing to keep in mind, Chad, is the longer your survey gets, the [00:03:58] lower your completion rate is going to be because people simply bail on your survey, don't they?

[00:04:03] Chad: [00:04:03] Yeah. I mean, you start getting over five questions and that estimated completion rate can start going down pretty quickly.

[00:04:10] Mike: [00:04:10] Yeah. So, we have seen time and time again when a team is working on a survey, and particularly when you've got perhaps a broader team that includes maybe a client or a vendor. Before you know it, everyone's got a smart question and you're at 20, 22 questions. I would never, ever send out a survey with 22 [00:04:32] questions.

[00:04:32] If they were 22 amazing questions, I'd send out two different surveys a week apart because people don't want to fill out that many questions. I think that's been one of our biggest learnings, hasn't it? Chad.

[00:04:42] Chad: [00:04:42] Yeah, I mean, I'll say it again. Keep it simple, stupid. Find out exactly what you want to learn from the survey ahead of time, and then be sure that all of your questions are delivering on that promise of finding out what you want to learn.

[00:04:56] Mike: [00:04:56] If you're getting a really bad survey score, a prediction of your completion rate, I think the best advice we'd give you is, try and break your survey up into two [00:05:06] parts and do a second survey at a later time. Let's say you had 16 questions, which would be too many, okay? I would much rather do two surveys of eight questions,

[00:05:17] than risk sending a survey of 16. Eight questions. You'll have a really high completion rate. You'll learn some things from that. That maybe change some of the questions in the second eight questions. This would be a far wiser approach than that of just trying to cram as many questions into one go. There's plenty of chances to survey your customers and your users. Don't rush it.

[00:05:40] [00:05:40] Chad: [00:05:40] And one thing that we always love to do as well is send the survey to our colleagues first, before sending it out. Because every single time we've done it, someone has, you know, made a suggestion and how to improve it. Either adding or removing an answer, rephrasing a question to make it more clear what you're trying to learn and understand.

[00:05:58] It can only get better by sharing it with others and it just makes the user experience from the survey takers' point of view that much better.

[00:06:04] Mike: [00:06:04] One sort of last advance Ninja tip is that often advanced surveys, we'll have a form of logic in them, meaning that [00:06:14] if a respondent answers the question in a particular way. For example, if they choose answer A, you're going to send them to a different page or question than if they answer a question B.

[00:06:25] And what happens is in the preview and score, you can actually preview the logic side by side as you go through the survey. In a side panel, Survey Monkey will show you the logic that's underneath what you see, and this is like really, really helpful, a...