In this episode of In-Ear Insights, Katie and Chris talk surveying: the good, the bad, the Buzzfeed. Instead of selecting random photos to tell you which Game of Thrones character you are, how do you build credible market research? What are the major biases that will sneak into your research if you’re not careful and ruin it? Listen or watch now to find out all this and more.
Subscribe To This Show!
If you're not already subscribed to In-Ear Insights, get set up now!
- In-Ear Insights on Apple Podcasts
- In-Ear Insights on Google Podcasts
- In-Ear Insights on all other podcasting software
Advertisement: Marketing AI Academy
This episode of In-Ear Insights is brought to you by the Marketing AI Academy.
AI Academy for Marketers is an online education platform designed to help marketers understand, pilot, and scale artificial intelligence. The AI Academy features deep-dive Certification Courses (3 - 5 hours each), along with dozens of Short Courses (30 - 60 minutes each) taught by leading AI and marketing experts.
Join Katie Robbert, CEO of Trust Insights, and me, Christopher Penn, Chief Data Scientist of Trust Insights, for three separate courses in the academy:
- 5 use cases of AI for content marketing
- Intelligent Attribution Modeling for Marketing
- Detecting and Mitigating BIAS in Marketing AI
The Academy is designed for manager-level and above marketers, and largely caters to non-technical audiences, meaning you do not need a background in analytics, data science or programming to understand and apply what you learn. One registration gives you unlimited access to all the courses, an invitation to a members-only Slack Instance, and access to new courses every quarter.
Sponsor This Show!Are you struggling to reach the right audiences? Trust Insights offers sponsorships in our newsletters, podcasts, and media properties to help your brand be seen and heard by the right people. Our media properties reach almost 100,000 people every week, from the In Ear Insights podcast to the Almost Timely and In the Headlights newsletters. Reach out to us today to learn more.
Watch the video here:
Can’t see anything? Watch it on YouTube here.
Listen to the audio here:
- Need help with your company’s data and analytics? Let us know!
- Join our free Slack group for marketers interested in analytics!
What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode.
Christopher Penn 0:02
This is In-Ear Insights the Trust Insights podcast. Do you want to use AI in your marketing but you’re not sure where to start? Take a class with Trust Insights and the marketing AI Institute. The AI Academy offers more than 25 classes and certification courses to start you on your AI journey including our intelligent attribution modeling for marketers certification. One membership gets you access to all 25 classes. Visit Trust insights.ai slash AI Academy to learn more and enroll today. That’s Trust insights.ai slash AI Academy to enroll today. Are you struggling to reach the right audiences? Trust Insights offers sponsorships in our newsletters, podcasts and media properties to help your brand be seen and heard by the right people. Our media properties with almost 100,000 people every week from the In-Ear Insights podcast to the most timely and in the headlights newsletters. Reach out to us today at Trust insights.ai slash contact to learn Again, that’s Trust insights.ai slash contact. In this week’s in ear insights, we are talking about surveying, market research and all the crazy stuff that people do. So Katie, what what’s going on with with surveying in your world?
Katie Robbert 1:18
You know, surveying is one of the more underutilized tools by marketers, but it’s also one of the more complex tools in a marketer’s arsenal of things because you need to be able to do it correctly. And there’s a lot of really bad surveys out there. And the thing that struck me over the weekend was that BuzzFeed, which is a, you know, online publisher, has these BuzzFeed quizzes, they have their own app, and I think that they’re just fun and entertaining. I think the one that I saw this morning was like, you know, pick a bunch of desserts and we’ll tell you where you should get married. So it’s like, you know, goofy stuff like that. There’s really no methodology. behind it, but my what’s what I started to wonder was are these pop culture accessible things, then misleading people as to how proper market research and surveying should be done? You know, obviously, we can’t see what’s behind, you know, the BuzzFeed quizzes and I’m not picking on BuzzFeed specifically as a hardcore market research, but they do publish a lot of these quizzes. And what I fear is that it sort of gives people this false understanding of how surveying should be done and how results are presented. Because there’s really no methodology. It’s, you know, pick a bunch of random images or rank something from gross to amazing and we’ll tell you how old we think you are. But
Christopher Penn 2:49
here’s the thing I wonder, you say, you know, misleading people into thinking account market research is being done. That presumes that people know how to do it correctly in the first place, which definitely knows about I think the lack of methodology is definitely a problem. The lack of understanding what the algorithm is, is is definitely a problem. But more than that, the thing that I know we see a ton, there’s there’s two types of bias that we see a ton of in, in data science and in market research in general, right? confirmation bias and selection bias. Confirmation bias. Yeah, for folks who are not familiar, is when you do data work of any kind with a pre determined outcome, you are, say a CDP company, right? And magically all of your research and studies say you know, cdp’s are essential to because you’re trying to sell more stuff. And we see that a ton it and in marketing. The other one is selection bias, where you are asking a population a sample of a population that is not representative of the whole so if we were to average to ask you know, for example, Our analytics for marketers group, which by the way, you can see the URL here if you want to join our free slack group. So how important is analytics? Well, of course, this group of people who have self selected to be part of an analytics for marketers slack group is going to think analytics is far more important. There’s this whole group of people out there who are not in the slack group who, frankly don’t give a rat’s butt about analytics. And so that is the second major kind of I see a ton of in in this type of research, and when you have something like BuzzFeed, you know, BuzzFeed audience is a very specific demographic. It’s a very specific age, gender, ethnicity. And if you don’t have any way of balancing that of doing the weighting properly, you’re going to get stuff that’s profoundly out of sync with the population as a whole as yet. I would say I would hope that Well, I was gonna say I would hope that marketers would realize that But to your point, I don’t think they do.
Katie Robbert 4:59
Well, I I think that there is the, you know, the bias built in of who’s getting the survey, but then there’s also the question structure. You know, and so you know, you gave the example of, you know, how important is analytics? So obviously the audience is the first problem. The second problem is, you know, the way in which those questions are structured, and are they leading questions? Are you giving, you know, answer options that allow people to say not at all, a lot of times what we see happen is someone will say, how important is analytics and it will be very important, extremely important, the most important thing and that right, there is a poorly designed question because the person the respondent has no opportunity to say it’s not important to me at all. And so the results are already crap.
Christopher Penn 5:53
I remember seeing a a well known politician sent out a survey like that asking how is this person doing their job good, great, amazing. Like, why is light yourself on fire? Not? Because for this particular individual, but yeah, imbalanced and that’s a professional unto itself is being able to design questions well and, and intelligently. So that’s the problem. How do we fix this? How do we help people fix this particular problem? Have a understanding market research and be doing it? Well,
Katie Robbert 6:25
we slap their hand and tell them to stop it. Stop it. Um, no, I mean, I think that honestly, there’s there’s a little bit of education involved, which is probably the answer that people don’t want. They want the easy answer right now. Marketers, you know, the general population is stressed out and overburdened and too many things to do and burnt out. And so saying, okay, you can’t do serving until you learn how to properly do it by taking courses or going to school. People are gonna say, Well, great, I’m not going to do it at all. You know, but there are websites out there that can kind of like help you. I don’t want to say cheat or shortcut, but that can assist with, you know, do I have the correct number of respondents to be a representative sample? Can I check to see if my questions are balanced or not, you know, is this written poorly? Or is it a leading question? Like there’s resources out there that already exist where you can literally copy and paste your stuff into it, and it will give you a little bit of feedback. So I would say definitely start there. It’s not the best solution. But it’s a start. It means that the person who’s developing the survey is at least willing to learn a little bit more.
Christopher Penn 7:49
What are you doing, Chris? There’s an easy answer. It’s just not a cheap one. Either. Be just hire somebody else to do it for you.
Katie Robbert 7:55
Well, sure, but not everybody has that luxury right now.
Christopher Penn 7:59
Exactly. No, I, I think your point that learning how to do it well, is important. I think also, one of the things we’ve run into, and we’ve run into this with a couple of our clients is that a survey is like a short story, right? When you’re writing a short story, what tends to happen for newer authors is you try and do too many things. You try and cram too much stuff into one story like, No, you’ve got two ideas, split that up, make each its own story. When you look at the way people design surveys, it’s exactly the same problem. They’re trying to do too much instead of say, Okay, let’s run a survey to learn, be very specific and very clear. We want to learn this one thing, this is the thing we want to know. And what inevitably happens particularly new survey by committee, is that every article I have a question I want to ask people and you know, you end up with this bloated survey is 500 questions long and you realize the completion rate will be zero percent because nobody wants to sit through that.
Katie Robbert 8:58
Mm hmm. Yeah, no. No, I think that that’s a really good point. Um, you know, it’s such. It’s such a difficult thing to try to stay disciplined to say, I have a single question and I want to answer that single question. You’re absolutely right. People try to jam, you know, 20 different questions into a five question survey, and be like, well, if I ask them this, then I can maybe allude to what this other thing is here and make it complicated. And skip logic. And, you know, if I get open ended to me, they’ll get more detail, but I don’t have the resources to analyze properly open ended questions. And so honestly, your best course of action is to pick one thing you want to know. And then keep the responses simple, like keep it yes, no, it’s, but that is a very difficult thing to do, because it requires a lot of editing and discipline.
Christopher Penn 9:53
Yeah. And I think it’s a really important point you just made there too, that there’s two different types of surveys. There’s qualitative and quantitative of surveying. And you have to know which one you want to do first. So if your qualitative surveying generally comes first where you say to people, hey, when it comes to analytics, what’s important to you? And you might do this to, you know, a small sample like 50 people, and gather qualitative answers. You can also use things like social media. Social media is a great tool for qualitative surveying, because you can get a sense of what are all the topics and questions people have. And then you distill that down to your hypothesis following the scientific method. And then you ask the quantitative survey, you say, how important is Google Analytics to your marketing program? not important at all. No. unimportant, neither important nor important, very important. And the hypothesis behind the scenes the single provably true or false statement is Google Analytics is important to most marketers and you would then be able to to look at the results process but without that qualitative quantitative relationship. What happens is, you end up asking questions sometimes that are your point of view, but not what the audience cares about.
Katie Robbert 11:10
I think one of the mistakes that we see people make a lot is they already have like, the headline or the title of this survey formulated on there. Right? And so like, you know, so if we use that example, that Google Analytics is important to your marketing, you know, I think what ends up happening is okay, we want to design a survey and push it out to our customers that says, Google Analytics is an important tool, therefore, you need to hire us to help fix up your Google Analytics. And so in my head, I already know what I want the result to be. And so then I start to design a survey around getting that exact result, so that I can convince people to hire me and fix their Google Analytics. And that, unfortunately, is a huge problem because that’s not representative and it’s just Bad surveying methodology. It’s just it’s, your hypothesis is flawed. If you think you already know what the answer is going into it. And yes, you can see my dog dancing behind me.
Christopher Penn 12:13
We used to run into this a ton. Back in our, in the days, we worked at the PR agency where you know, you’d have a PR folks say, hey, I need I need a surgery to fit this headline. And boy, were they unhappy? We’re like, no, that’s that’s not how that works at all.
Katie Robbert 12:28
Well, they stopped asking us to help you that’s really what it came down to is they didn’t like the results. And so they stopped bringing up bringing the survey to us in the first place. Because they didn’t want to pay for it to be done properly. And if it was done properly, then they couldn’t get their, you know, sensationalized headline that they had already promised to their client. And so that was unfortunate because we had to watch sit by and watch a lot of really bad data. Leave the walls of the company and it’s just, you know, as people like us who really value data quality data integrity, proper methodology, you know, it’s just, it’s hard to watch knowing that there’s nothing that you can really do about it because your hand has been slapped, and they say, No, don’t touch it. This is the way we’re gonna do it.
Christopher Penn 13:18
And the interesting consequence of that was that within a year of Google consumer surveys being becoming broadly available, reporters and publications started saying we are not accepting survey data. Because it’s all you know, terribly flawed. The antidote for that that we came up with and we still use today is design a survey that will actually help your marketing. So if the survey results never aired anywhere, you would still get benefit out of it. What is something that you legitimately want to know? That would make your marketing better? And that tended to create better surveys tend to create people like oh, I actually want to know like, how, how important is Google Analytics to people? Am I barking up the wrong tree? Mm hmm.
Katie Robbert 14:03
You know, I think that one of the things that we’ve done that’s been really, really helpful and to your point, the results never see the light of day except for us. And it helps inform us is a one question survey. And I think that that for people is like, Well, no. Why would I just ask one question? I have 100 million different questions. Well, yeah, so do we, however, by asking one question at a time, once a quarter, you don’t burden your audience with, you know, tell me everything so that I can give you exactly what you want personalized all the time. You know, you’re really focusing in on one thing at a time. And then taking those results from that one question and applying it and seeing how that impacts your marketing from the inside out. And so our most recent question, which you can answer on our website, you can shoot us an email whatever you want is about live shows and We’re recording, you know, a video podcast right now. There’ll be audio available all of that good stuff, but with live events, taking, you know, a hit with the pandemic, speakers are looking to do live streaming and, you know, live videos, just to sort of like, you know, keep their audience engaged and you know, do the talking head thing. And so, we thought for us at Trust Insights, we’re like, Well, why aren’t we doing this? But before we started to do it, we wanted to know what, what makes a good live show what would make someone continue to come back. So we went in with the hypothesis that people will check you out once, what keeps them coming back over and over and over again, and we’ve gotten some really interesting answers that, you know, we may not ever share publicly, we’re, you know, they’re not proprietary. We’re happy to share the information but the value for us is that we’re getting real heat Back from people who are likely to become loyal viewers of our show, and how can we provide a show tailored around the things that people say they want?
Christopher Penn 16:10
Yep. Yeah. And this is, I mean, even still, this is still the qualitative stage where you’re asking an open ended question you can quantify some of it if you do some text mining and you know, cluster based on the different, you know, most frequent words and phrases, but best practice would be okay, great. We’ve gathered the information. Now, if we wanted to, we could put together a follow up or even better to avoid the selection bias. commission at to a panel that is not our audience, to say like in a live streaming show. These are the rank which of these five things is most important to you important, intelligent speakers, you know, relevant, timely topics, better graphics, and that quantitative thing would then help us reorder the priority. But one of the other things that’s so helpful About the one question survey is that it puts you into a a mindset of, Okay, what do I need to prioritize what’s really important? Because if I only get one shot a quarter, what do I need to know for the next quarter? And that I think has been super helpful to us, as opposed to saying, I’m going to email my list every week.
Katie Robbert 17:21
I agree with that. And like I said, that one question survey takes a lot of focus and discipline, because obviously, we are naturally curious, and we have a lot of questions for people. But to your point, Chris, we have to prioritize what is the most important thing right then in there? So we’ve asked questions about Google Analytics, and it’s given us a lot of information. We’ve asked questions about general, you know, analysis and analytics. As I mentioned, we recently asked this question about live streams, you’ve asked questions about, you know, what makes an engaging newsletter. And I think the other side of that is if you’re going to take the time To ask your audience or any audience to answer your answer your question, you’d best have a plan to do something with that information, because then otherwise, someone’s going to say, Well, I gave feedback on that thing. And I didn’t see anything changed. So I’m never going to answer them again. And that’s frustrating as a user, especially when you take the time to put together some sort of a thoughtful response. And then you don’t see anything come of it.
Christopher Penn 18:24
Yep, another hack that I love, especially in the current age of you know, more webinars you can shake a stick at is attending webinars, ignoring the speakers. And just listen to the q&a. Listen to the ads questions people ask. Because you can get a you know, on any topic given topic, you can take that webinar recording stuff into a service like you know, auto that will auto transcribe it, and pull out the questions that the audiences have. If you attend five Google Analytics webinars, and you keep hearing the same question over and over and over again, like how do I fix my UTM tag? It’s not right. You get Very good qualitative sense, this is something that might be important to people. Can we refine that and turn that into a quantitative survey? So there’s a lot of different ways to pull in that qualitative data. I would encourage folks, if you’re trying to figure out how do I get started with some of this? The audience is there. They’re asking questions all the time, go to Korra. Go to other people’s webinars, go to your slack communities and your discord communities and ask folks there’s there’s no shortage of information, you don’t have to resort to, you know, pictures of cats with cakes, to find out what’s of interest to your audience. So to wrap up, surveying and market research, are absolutely essential if you do them right. And there are so many ways to not do them, right. But it comes down to making sure your surveys are free of selection and confirmation bias that you are following the proper qualitative and quantitative processes and that you’re analyzing the data correctly. If you got questions about this particular topic, hop on over to our mo x for marketers community the URLs below at trust insights.ai slash analytics for markers for those who are listening to the audio, and ask your questions there about this or any other topic you want to know about. We’re more than happy to answer them. While you’re at it, hop on over to the website TrustInsights.ai dot AI slash newsletter and subscribe to our weekly newsletter we’ll talk to you soon take care want help solving your company’s data analytics and digital marketing problems. This is Trust insights.ai today and let us know how we can help you
Need help with your marketing data and analytics?
You might also enjoy:
Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, Data in the Headlights. Subscribe now for free; new issues every Wednesday!
Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new 10-minute or less episodes every week.