How to Create a Great Survey w/ YouGov’s Brian Reitz

• Focus on what sparks public opinion and conversation.
• Surveys with just 1–2 well-crafted questions can yield strong PR results—don’t overcomplicate it. Simplicity is key for headlines.
• Trending or overdone topics (like tipping) can still succeed with fresh, up-to-date data and emotional framing.
• Demographic splits (e.g., Gen Z, political affiliation) are essential for pitchable headlines and broader audience resonance.

After completing a deep analysis of over 2,500 YouGov surveys, I invited YouGov’s Senior Marketing Manager, Brian Reitz, to the BuzzStream Podcast to discuss my findings.

Brian was a wealth of information and provided us with insider insight into how the YouGov team identifies ideas, crafts surveys, and analyzes data to craft compelling stories.

As a bonus, he also provided new data findings from his Reality Checks podcast for our discussion.

This is our second deep dive with a survey company. If this is interesting, I highly recommend also checking out our episode with the team at Talker Research.

What does your internal process look like for deciding what topic is actually survey-worthy?

As for our process, we have a few teams regularly working on surveys and articles.

We kind of split it out two ways. So we have a B2B marketing team and then we have more of a consumer side.

And the consumer side focuses on a lot of the politics, a lot of the fun, like what’s your favorite ice cream flavor?

Could you beat a bear in a fight?

But also, do you think that the government shutdown is the fault of blank and things that people have opinions on?

People have opinions on both the fun things and the serious things.

Then, separately, we have a business side that focuses on conducting surveys relevant to insights and marketing leaders at brands such as automakers, CPG brands, and travel brands.

So I would say that everyone has a different process.

Still, ultimately, the same goal is to do something interesting. Find an insight or a question that leads toa  conversation happening.

Usually, the process for finding that is by reading what people are talking about, whether that is on social media, whether that’s in trade publications or talking to people in the industry.

So we’ll work with a lot of the people on our sales teams, client teams, see what topics are coming up regularly, and then use those topics, synthesize them, figure out what is the core theme here, run surveys about that.

But then separately, if you have a pardon, a presidential pardon, that’s newsworthy, that’s topical, that’s something that people have opinions on.

I would say a lot of the political stuff isn’t always predictable, but it’s pretty regularly divided along partisan lines.

At the same time, people do have opinions and want to share them.

So yeah, it’s basically kind of mixing what’s timely, what’s happening now, and then what people are talking about, what trends are emerging.

Do you have brainstorming sessions?

Yeah, I mean it really depends on what is happening that day.

If it’s a day when the White House is being reconstructed, that topic may not fit the bill.

There’s more pressing things to ask about.

But if it’s a slow news day, then yeah, I think the team published an article a couple weeks ago that was like, are the 12 biggest food fights in the country?

And it’s like, is a hot dog a sandwich?

Is coleslaw a casserole?

Is a hamburger a sandwich? Is cereal soup? Stuff like that.

And yes, there’s a team that focuses on that.

We also have several editorial staff.

We have a US News editorial staff; that’s their official title.

They’re the consumer forward team.

And they run that survey. So anyone can pitch them ideas.

We don’t have meetings to discuss it.

They may do so, but we have open lines of communication.

We have a pretty open Slack channel where people can share their ideas, people will collaborate on the ideas, toss in emojis when they like an idea, standard stuff.

But yeah, it’s very collaborative, and it’s something that really just depends on how much is going on in the world.

And then, as an aside to that, there’s stuff that has to be relevant to the American population.

So it’s something that you’re seeing on your social media. Like very often—and this happens to me, this happens to lot of people—you think that it’s the most important thing because you’re seeing it.

But realistically not everybody is talking about blank.

So it’s kind of having a sense for what is actually breaking through.

And maybe sometimes, even if you don’t know it’s breaking through, being able to prove that it’s worth.

You know people have an opinion on it: if you run a survey and 60% of people haven’t heard of it, the results aren’t that interesting.

How do you ensure a survey idea is picked up?

I think that even if something is newsworthy, it might not be covered, you know, for variety of reasons.

Perhaps it’s pitched to the editors or journalists at the wrong time.

You know, they are writing a story and they miss it.

So, you know, there’s stuff that we publish that is like the most interesting thing to me and it doesn’t get picked up.

And then there’s other times where we publish something and it’s like going gangbusters.

So it really is kind of arbitrary sometimes.

One way we avoid that here is that, obviously, we run a bunch of surveys, and we’re going to have more media placements as a result.

But separately, YouGov owns its own research panel.

People are completing surveys for us.

We’re able to cross-reference the things that they tell us from one survey with another, which it’s all anonymized.

But the phrase we use internally is “connected data”, which is kind of like a buzz phrase.

That’s good, this “buzz stream buzz phrase.” 😉

But yeah,  connected data is like, if they told us in one survey that they think this and then in another survey, they told us this thing, then if in that second survey, we don’t see anything that interesting, we can kind of compare the results and say like, well, we know that they’re mobile gamers because of this thing.

So it turns out that mobile gamers are more likely to, you know, like advertising from Coca-Cola.

So maybe the U.S. general population in the survey doesn’t have as strong an opinion on Coca-Cola’s advertising, but we know they’re gamers.

And we can segment it through that.

So that is one of the ways that we’re able to sometimes circumvent some of that concern that we have, where you’re asking, “is this a worthwhile thing to ask about?”

But yes, I can understand why someone would be hesitant to invest in something if they don’t know they’re guaranteed a positive outcome.

How do you determine whether to publish a survey write-up on your blog or just display the results?

I think it’s a bonus effect we get because our goal is to do things that are interesting, publish the results, and be transparent about it to the fullest extent we can.

I think posting engaging content that gets coverage will drive more search volume.

Having more search volume brings more people to the site.

More people to the site means more customers to the site.

Ultimately, we’re trying to do interesting things because that’s how you get the attention of the kind of customers that we have.

On the consumer side, it’s the news coverage, but it’s also the transparency and sharing of data.

So I think there’s probably an opportunity there.

I’m curious what you’re kind of thinking about when you say like publishing something with the search volume. Is that in mind or like in tandem with it?

Is it like using keyword research to kind of guide survey questions?

Vince Nero

Yeah, exactly.

So, like, I do think there’s probably a way, you know, if you were a fashion company or something and you wanted to rank for a specific word, like there is kind of what we call these keywords with high intent to link, right? Like something like fashion statistics, right?

That would be a keyword that a researcher or a journalist is going to want to search when they’re doing, you know, research for a piece or something.

So if you can get your statistics ranking, then you’re more likely to get cited from these publishers, right?

These news outlets.

So you can improve these rankings by providing fresh, up-to-date data.

So, you would run a survey on, say, YouGov, gather as many statistics as you can, and publish them.

And eventually, you’d kind of, you could present them, but you could also start getting that ranking and start accruing.

links and eyeballs on the page. Does that make sense?

Brian Reitz

Yeah, no, that makes total sense. So yeah, I mean, think it’s like, we kind of have a bit more of a…know, demand generation approach, at least on the business side of things.

So it’s it’s tough because it’s like, that’s the data that you want to give away.

So that way you rank, but then it’s like,  content that is interesting to people.

So they’re willing to exchange information to get it.

And I think part of the idea behind that is, you know, if you’re willing to give your information to get the data, then that means that you’re very invested in it, you’re more willing.

I’m going too far down the rabbit hole, but yeah, I’d say as far as, you know, search goes, I’d say we’re more demand-focused than search, but obviously search is important.

And I think that’s something that we’re monitoring.

I think we’re in a good place to do that, given the volume of data and our regular links and quotes, which position us as authoritative.

I feel like I’m packing myself up like we’re viewed authoritatively. I generally sense that it’s something that, because of our approach, which was somewhat unrelated to search for the last couple of years, we’re in a good position to be pulled into a lot of the answer engines.

But it’s not something we’re actively considering, and I’m sure we will be thinking about it more and more.

Do you strategically target “best x” types of surveys?

I would say that oftentimes it comes more organically.

I don’t work with this client, but I know that we had one of our sports client services people in the office a couple of years ago.

He said he was visiting the Orlando Magic, which is a client of YouGov.

They argued at lunch over whether a hot dog is a sandwich. And it was something that had an emotional angle.

So he kind of brought that debate in and we’re like, yeah, let’s just ask about it.

So I think it’s the kind of thing that, you know, we’re not necessarily like optimizing for the machines.

That’s something that maybe works short term, but like ultimately, everybody who is, you know, we have multiple constituents, we have our clients, but we also have people we have 30 million people around the world  who are filling out surveys—so we have to have questions that are interesting to them too.

So it’s not just like, “Do you think that McDonald’s is a good value?”

That’s something that maybe more people care about than some emerging brands, but you have to keep people engaged, otherwise they won’t want to complete the surveys.

Would you agree that having emotion and a timely hook are key to a successful survey?

Yeah, I would say that those are…

The things that seem to resonate the most with the media.

That’s not necessarily what we’re always trying to accomplish.

So like we put out a media consumption state of report and it found that 67 % of people use Facebook, 70 % use Instagram, or that’s not right, it’s like probably 54 % use Instagram.

But you know, we published something like that in January because the idea behind doing something like that, while not necessarily newsworthy, is like you’ve got all these advertising agencies, PR agencies that have their standard decks and they need the data.

So that’s not something that like shows up in Google but it’s the you know kind of thing where we get our name out there.

Eight people want to access that data.

It’s one thing to know how to inform your strategy.

A lot of people use our data and they look at it and they’re like, okay, this is what I should do.

Others look for data to support what they’re trying to accomplish.

How many questions should you ask for a survey to be newsworthy?

Usually,y for pitching reporters, which is not something I’m directly involved with. I’ll help provide contacts. I’ll provide analysis.

I’m not the one handling interpersonal communications, but we usually send over a list of recent publications.

Or if it’s about one recent item or one topic, it’ll include a set of bullet points.

Sometimes that will be pulled out as a single item.

For anything we publish on our site, a clear, concise title makes it easier to digest.

You know, if it’s just U.S. media consumption trends or something, it doesn’t really capture attention in a lot of our outbound communications. Like the email, social media, blog posts, etc.

But as far as like these surveys we run at any given point, it’s a pretty direct question.

And part of the reason we’re able to do that is that we have that kind of connected data that I had mentioned previously.

Even if we are only asking one thing, it’s on the back end.

We can still see the answers to the one question, and then cross-reference them against a range of other metrics. So every day, I know like that US news team I mentioned that kind of runs the consumer side of things, the politics or the entertainment or the fun surveys.

Those are published on our website every day except for Saturdays, Sundays.

On those days, we’ll run a question over the weekend and publish it on Monday.

Still, they usually have about three per day that are related, so if it’s something like the government shutdown, which by the time this airs, we hope is back up and running.

Still, if not, this is topical, and just in case, I’ll do another.

We’ll have to hold this episode for another six months before the politics are added again.

But yeah, so it’ll be three questions that are related, and then it’ll publish automatically to our website, and then it has splits that are like region, gender, politics, age, race, and that’s all publicly available.

And then all those questions are added into one of the tools that we have, which is a syndicated data set that people can then digest further based on the other metrics, like mobile gaming, Coca-Cola, or I’m planning to paint my house in 12 months, or I think that I’d rather be named after I’d rather have a mountain named after me than a theory.

Just random items; not everyone needs everything, but someone needs something.

Vince Nero

Do you recommend they consistently aim for more than one question, or how many questions, in your view, would make a solid story?

Brian Reitz

Yeah, I it depends on the goal, because I know this audience is like digital PR professionals that are trying to link building.

I would say that, given my bias at YouGov, there are more questions, right?

But realistically, you only need one or two questions.

And ultimately, you want to be able to, you know, get the crux of those questions.

What is something that actually matters to people, and what is something that people have an opinion on?

And chances are you’ll pick out something from that, especially if you’re able to kind of segment it by different audiences, even if it’s not in your questionnaire.

But more often than not, I feel like a common challenge is when we assume this demographic in this region has this job. Then, based on that, this is the percentage.

It’s like, nobody cares.

People want to know that the headline reads something like “30 % of Americans are concerned about tipping.”

Nobody wants it to be too specific, because then it gets too niche.

It’s like you spend more time complaining, or not complaining.

I mean, I’m complaining now, but you spend more time explaining than you do actually just getting to the crux of what the topic is.

I would say don’t overthink it.

Trust your gut with the emotion, but also make sure that, especially if you’re investing in it, you’re doing something people actually care about.

Even if there’s research out there about it, like we will see stuff every week, there’s something new about tipping.

Emotional topics are often written about because so many people write about them.

Even if a topic has been covered, is it always acceptable to add new data?

Yeah, again, easy for me to say because it’s so simple for us to run surveys.

But if there’s a topic that is relevant to what you’re doing and there’s anything new about it—even if it’s like a month later— and people seem to want to talk about, I’ll just say that again just in case, there might new people who will want to talk about it again.

Should you always include demographics in your results (e.g., gender, age, political leanings)?

Yeah, I mean, I’d say that it’s a combination of those two, where that is in every survey.

So everybody has that data.

It’s very easily accessible, I think, from a pitching standpoint, the headline of like Gen Z thinks blank is usually attention grabbing.

Or I guess I don’t even know if I’ve seen any Generation Alpha stories yet.

I feel like we’re due.

We had the millennials are killing blank.

Do the Generation Alpha survey.

But I say liberals versus conservatives, people have opinions on.

Men versus women. People have opinions on.

It’s very standard, and I guess it’s easier to pitch those.

It’s harder to pitch when your niche a lot tighter. (And I think one of your takeaways was like, if you’re going to do a niche, only pitch out the five reporters.)

So if you’re going wide with it, you want to do like a gen-Z blank, but if you’re really going to a hyper-targeted industry, then that’s when you might want to cut it deeper.

And again, that is what’s useful to a lot of the people who work with YouGov in a you know commercial capacity  is they want to know what the clients or customers are thinking.

They want to know what perspective customers are thinking.

That’s not what the press wants to know.

They don’t think their readers care because it’s hard to package that and distill it.

No one has the attention span to figure out the methodology. Hence, they need the method to be as simple as possible.

It needs to be accurate in the headline, so they don’t get called fake.

Do negative headlines sell better than positive ones?

I think it depends on what the negative thing is and where it is positioned.

You might get more coverage for negative stuff, but in the long run, it does not lead to more business.

And this is vague, know, vague answer.

I don’t have hard data on this, but I think negative emotions are easier.

It’s very, it’s a lot, it’s a little lazy.

And I’m not saying that I haven’t done it or that you haven’t done it, because people are positive and negative.

So, that is a full spectrum of things.

It’s easier to say Gen Z is killing the diamond industry than it is to say, Gen Z is saving for a future opportunity to build a home.

That’s not that interesting. S

I think it’s easier to hate someone or something than to find the positive in it.

But that doesn’t mean that it’s not possible to get a positive story in it.

I may not have answered that question. I’m not sure I raised enough.

Vince Nero

It can feel like every bit of news is negative, and it’s easier to feed into that, as you said.

But I do feel like a lot of the stuff that has done well for you guys, for YouGov, specifically has seemed to be kind of in the middle. There were fun topics, the pie at Thanksgiving for example. But in order to have success with that, I think you need to be more like a YouGov, where you can lean on your authority that you’ve built over time.

Whereas I may not be able to launch a new brand and just say, “Here’s what everybody’s favorite pie is at Thanksgiving.” You have to put in your dues.

Brian Reitz

Well, and also you probably wouldn’t want to unless you’re selling like flour or apples.

I don’t know what value I would add if I worked at an agency.

I guess you just get your name out there. Like, so what? People will say that some PR firm published the pie results.

And it’s like, OK, good to know.

It’s like it doesn’t really do anything like for us.

Just because it keeps our survey panelists engaged.

But then also it is the kind of thing where, because there’s so much negative news, it kind of counteracts that a bit, because you know, nobody wants that.

You have to do it because it’s a topic people are discussing.

But like, I feel like it also is dependent on the group too.

So it’s like, we published a report, I want to say two years ago and it basically  asked people what they thought of the economy and then we looked at it by their ideology.

And two years ago, when Joe Biden was the president, liberals were like, ” The economy is great.”

And conservatives like, “this is the worst it’s ever been. We’re really going down the tubes.”

And now we’ll ask that same question today when you have a Republican in office, and it’s the exact opposite.

Liberals are like, “oh, man, I don’t have a future. We’re really, really doing bad.” And conservatives are like, “yeah, let’s go.”

So I really do think it’s dependent on the target group that you’re looking at.

What are common mistakes and pitfalls that people fall into when creating surveys?

Yeah, I mean, I think that is similar to what I was saying earlier, where overthinking it is probably the big thing.

Specifically for a PR situation, if you are going out and you’re like, “64 % of mobile gamers in Alabama think that playing on the toilet is gross.” You’re explaining too much.

There’s too much going on.

If you’re able to say, as two-thirds of gamers play on the toilet.

It’s like when you’re kind of trying to distill it, you really have to simplify it as much as possible.

Otherwise, it’s too convoluted.

And, obviously, people who want more information can request it, especially reporters.

Giving them too much is nice, but it’s overwhelming.

So it’s like, you want to lead with, you what are the headlines?

You can provide more information if helpful.

At least that’s what’s worked for us.

Again, could be different for everyone, because, you know, it’s easy for me to say, because people know who YouGov is.

We’re not fighting that uphill battle of like proving the credibility as much.

But then, yeah, the secondary point is, that you either have to be very transparent with the methodology or the data or be very conscious of the fact that people might think you have a bias when you run a survey.

You can look at another example, I think probably two, three years ago, we ran some research that asked people if they were feminists, and I think like 33 % of people said they were.

But then we also asked people like, you think men and women should be equal? And like 70 % did.

And then we asked people like, the definition of feminism is this. Do you think that men and women should be equal, and it was like 50%.

So it’s like, depending on how you word the question, it’s like people have totally different answers.

Something that people get in trouble with a little bit is that, they sometimes let their biases affect the questioning that they do.

And then when they pitch it out, they don’t realize that maybe it’s not as objective as they as they think it is.

And again, I’m not saying that I’m immune from that.

Not saying that everyone at YouGov is immune from that.

But it’s something that you’ll see that be a pitfall that people have when they’re doing surveys.

What is the Reality Checks Podcast that you host?

Basically the concept for that is we’ll partner with somebody on a survey, they don’t see the results, and then they are asked to predict what the results are live and then give an answer.

So it’s like a two-part.

It’s like, do think this?

Now that you’ve seen it, why do you think you were wrong or why do think you were right?

And I don’t try to fool people but I don’t want everyone to get everything right because that is boring.

But if you get everything wrong, it’s like there are Instagram Live shows where guests invite someone on, and the whole goal is to embarrass them.

There’s some like there’s some like social dynamics at play where it’s like the person that’s being invited on maybe needs to be embarrassed.

So that’s not what we’re doing. But anyway, you know, that’s how that is structured.

So I have a few surveys here.

Guess I could ask you a couple of questions.

You haven’t seen these, I think.

Well, I mean, you’ve probably seen a lot of these because you went through 2,350 YouGov surveys.

So I’ll have to pick something recent.

So all right.

So following the NBA federal government illegal poker game kind of match up that came out recently, you have asked how often do you think current professional athletes alter the way they play to help sports gamblers win bets?

So yeah, we asked 18,000 actually is 18,900 so close to 19,000 Americans that question and 13 % said often 32 % said sometimes.

Vince Nero

Wow. Yeah, mean, it’s okay. More is coming out. There’s I’m a much bigger baseball fan. Still, there is a big thing with the pitcher on the Guardians who I think he’s been suspended indefinitely because he was there was one specific pitch that, you know, however many people bet on. Draftkings or something and that they flagged it because it was a terrible pitch and he ended up there were a lot more of those types of pitches. Yeah, pretty wild. Pretty wild stuff.

Brian Reitz

Yeah, yeah, that’s a crazy kind of world.

So I won’t weigh in on that, but I know that this recently came out and is something that we surveyed about.

And there’s a lot of skepticism from average Americans. So again, this isn’t cut by sports fans, it’s not cut by betters, just average Americans.

Vince Nero (40:41)
So that’s a great point, Brian, too.

So that’s kind of where your mind would go, right?

OK, if I’m going to present this to a journalist or something or write a post about it, it’s going to be like, now let’s see what angles can make this more interesting as well.

Like, we know the overall takeaway. Let’s dig in.

Brian Reitz 
Yeah, mean, it’s like, yeah, we know 35 % of Americans think athletes often or sometimes alter the way they play to help sports gamblers win bets.

But you’ll see clips of like.basketball players who are playing the game and then you’ll hear somebody yell like, get one more rebound, Paolo, to reference the Orlando Magic.

There’s a clip and he’s like, you’re betting on preseason, bro.

You got a problem.

So it’s like, you know, with something like basketball, there is a dialogue.

You might not get that in football because the field’s farther from the court or hockey because it’s so fast paced.

And, you know, maybe you can still hear it, but it’s like you’re not like just standing around as much as you might be in basketball or baseball.

You know, you could probably yell at the players and they would hear it, now that’s not like socially acceptable at a baseball game like it might be at a basketball game.

But there’s different ways that some, and again, I don’t know how anyone answered that.

There’s also something too where like we asked this question the day that the FBI came out and put this out.

I’m a big basketball fan, I kinda had, I hadn’t heard about Chauncey Billups, but I had heard about Terry Roger before that had been an open and shut, put it in quotes, case, so to speak. that wasn’t new information. for anybody who’s not watching this sport,

We also asked on the same day how many people follow these sports or follow the NBA.

And I think it was like 52 % said they don’t follow it at all. And then like 48 % had a level of like a little bit interested, somewhat or very. So we would cut it too by like, are you an NBA fan? Are you a sports bettor? Are you like men versus women who are within those categories? Or like you could even be like people who bet thousands of dollars or people who do like a little $5 parlay.

Vince Nero

Right.

Yeah.

Wow, so you have that much, that degree of kind of, what do you, whatever you it, data accuracy or, I don’t know what the word I’m looking for, but you can dig in that closely to some of these people to slice it that way. That’s pretty incredible.

Brian Reitz

Yeah, I mean, not too far, because you want to have enough sample size to be statistically significant.

But yeah, that’s where we would explore things.

And usually, again, we’ll publish the baseline information because that is the most succinct way for anyone to interpret it. But that’s stuff that people, a lot of the clients are looking at is they want to know the very specific. They have their eight different customer groups, and they want to know which of them thinks about everything.

Vince Nero

Let’s let’s let’s leave it here.

Brian is maybe hesitant to sell this, but like YouGov has a self-service platform and they have a full service platform, I believe for brands.

So like you can go on you gov and run your own surveys the same way that you would you know, SurveyMonkey and Pollfish and these other the other ones.

And I mean, I remember doing it myself. It’s a great service. And so, yeah, definitely check it out.

Vince Nero

Vince is the Director of Content Marketing at Buzzstream. He thinks content marketers should solve for users, not just Google. He also loves finding creative content online. His previous work includes content marketing agency Siege Media for six years, Homebuyer.com, and The Grit Group. Outside of work, you can catch Vince running, playing with his 2 kids, enjoying some video games, or watching Phillies baseball.

More Posts by Vince Nero
Website: https://www.buzzstream.com