Data-driven marketing is a magnet for traffic, backlinks, and word of mouth. To produce it, you need to do some research and present it in an appealing way. In return, you’ll be able to boost the traffic to your site. Today we are talking to Susan Moeller, the business development manager at BuzzSumo. She’s going to tell us about how she finds, analyzes, and translates data for her readers.
Some of the highlights of the show include:
- A bit about BuzzSumo and what Susan does there, as well as her marketing background and what brought her to BuzzSumo.
- How Susan defines data-driven content and how she uses it in her position at BuzzSumo.
- Why it’s important for your content’s authority to research and have sources for things you know to be true.
- Why using data in your content can boost backlinks and shares.
- How Susan determines which questions to ask to find the data that her readers are looking for.
- Susan’s processes for gathering, analyzing, and translating the data that she finds.
- Susan’s best tips for starting to create data-driven content for someone who hasn’t done it before.
Nathan: Your audience has questions. That means they are looking for answers. You need to go to source for those answers and all it takes is some research. That’s why we’re talking with Susan Moeller today on the Actionable Marketing Podcast.
Susan is the business development manager at BuzzSumo. She’s been conducting a ton of her own research to create data driven content. This kind of marketing becomes a magnet for traffic, backlinks, and word of mouth. You’re about to learn how to find that data, analyze it, and translate your findings into amazing data driven content.
I’m Nathan from CoSchedule. Let’s check this out. Hey Susan, thank you so much for being on the podcast today.
Susan: Yeah, I’m really glad to be here. I’ve listened to some of the earlier podcasts and I’ve really enjoyed them and learned a lot from them so I was excited to talk with you more and be part of the podcast.
Nathan: Thank you. I’m really glad to have you. It’s fun to know that you’ve been listening. We’re big fans of BuzzSumo here at CoSchedule. We use the tool so we really like what you guys are doing. It makes it a lot more fun for us to have you on our podcast.
Susan: That’s great. We’re glad to have you as subscribers too. You guys are doing good work.
Nathan: Awesome. We are happy. I guess with that Susan, let’s kick it off. Fill me in on BuzzSumo.
Susan: We’re growing. We’ve added two people to our team. That brings us up to nine. We have just hosted our second event, Content SEO. This one was in New York. The first was in London. We partnered with Majestic who provides back links for the BuzzSumo tool for both of those. We’ve had some speakers like Shafqat Islam from NewsCred, Leoden, Jason Miller from LinkedIn to come and speak to groups of subscribers and their guests in those two cities. It’s been really fantastic for us to meet some of our subscribers and people who are interested in marketing, content marketing, SEO, in person.
We’ve also just kicked off a new BuzzSumo experts webinar series that’s going to showcase subject matter experts on various topics. That’s going to be offered free if you’re BuzzSumo. We’ll basically be supporting the speakers by providing research if they need it and adding to that series with research as the series progresses.
Nathan: Yeah, sounds like a lot of fun. I know that you guys are producing a lot of content over there. That eludes me to my next question. I was just wondering, what is your background and how did you end up in marketing at BuzzSumo?
Susan: Marketing is a new role for me. Talking to people, research, writing, and teaching really aren’t. I actually have a degree in education and a master’s degree in theology but it’s a joke in my family about how long it took me to settle even on those choices. I’ve worked in journalism and in public relations for a few very small non-profits and businesses.
BuzzSumo needed someone to run some of their webinars to help customers understand what the tool could do. That’s how I started with BuzzSumo. I guess it’s been about two years now. I was their first employee. I have that employee number one parking place. It’s a virtual one because we all work remotely.
One thing that I like about startups like BuzzSumo is the freedom and even the expectation there is for people to evolve and grow with the company into new and different areas. My title now is business development manager. That includes a lot of different things. Some of my favourites are working with new customers to help them learn to use BuzzSumo and to get them most of their subscriptions. Also, putting together these educational webinars with the experts who are interested in partnering with us.
Nathan: It sounds like you’re doing a lot of different things there. You’ve mentioned webinars, conferences, and I know you guys are creating content too. We want to talk to you about data driven stuff because you are very smart with the projects that you take on. Just to talk about that just a little bit to define it for us. How would you define data driven content?
Susan: As I was thinking about it, I wanted to distinguish between data driven content marketing and data driven content. I would define data driven content marketing as the plans and strategies that we make based on information drawn from data sources. Things like customer profiles or preferences. Those come to you in a number of ways; landing pages, page views, shares, the number of backlinks for content and all of that in form is data driven content marketing.
And then I would see as a subset of that data driven content. I define data driven content as articles, infographics, anything that you would think of as content that either describe data or draw inferences from data.
Nathan: Give me some examples. How do you use data in your content at BuzzSumo?
Susan: It gives us things to write about is I think the simplest answer to that. In journalism, you think about finding the story. Maybe you go to an event or you do an interview or both. But for us, a new data set is both the event and in an odd way, the interview subject. We ask questions. We try and interact with the data to find the story that it’s telling us.
Nathan: You mentioned the definition of data driven content. Could you give me an example of a piece of content you’ve done at BuzzSumo that’s data driven?
Susan: I recently wrote a piece for Content Marketing Institute. In it, I analyzed 286,000 articles that had been published at LinkedIn Publisher. For me, it was a huge project. It was my first big jump into data analysis and using it in a post.
But for BuzzSumo, there have been a couple of others that I think are also good examples. One is my colleague and one of the founders of our company, Steve Rayson, wrote an amazing piece of analysis of 1 million posts. In it, he was able to determine that 50% of randomly selected posts actually gets 8 shares or less. That article got a ton of attention. It was published on Moss and then we published a summary on our site. I think the combined shares for that are approaching 10,000 right now, which for us was really good.
Neil Patel uses some of our data from BuzzSumo, an analysis of 1 billion Facebook posts and he wrote up an article describing what works on Facebook. Even more recently, BuzzFeed UK has used BuzzSumo data to create a social barometer that aims to describe the election as it’s playing out on social media.
What they’re doing is pulling our data in and helping people see in the UK’s general election which campaigns are winning on social media. Those are some of the examples of using BuzzSumo data.
I guess another more recent one for me would be, I looked at 3,000 videos from Facebook from the last year to determine what was an engaging video. Basically, I looked at the videos that had gotten the most interactions on Facebook and decided looking at the data, we already know, recipes, cat videos, things that draw your emotions. I guess that’s the dark side of data analysis.
Sometimes, you don’t find anything that is novel or new but I also feel like it’s valuable even in those cases, to have actually done the research. When we limit our understanding of what we see to only the information that we can take in as a person on social media, we are very vulnerable to blind spots and what we perceive. I like to actually do the research and see what comes out of that even if the result does tell you that yes, puppies and babies are popular on Facebook. It’s still valuable.
Nathan: I think that’s awesome and it’s really smart actually, Susan, because there are those ideas where it’s just like a stereotype that cat videos do well. But you did the research and now have the proof to say that that is a real thing, right?
Susan: Yeah, exactly. I feel like that makes me a more trustworthy source when I write about other things and I think it makes BuzzSumo a more trustworthy source as well if we’re honest about the things that we find. I still think it’s important when you look at data to look for the novel and the new but to be honest about what you find.
I think probably for me, especially because I’m not a data scientist is I need to be very honest and very clear about what I don’t know and make sure that that comes through in what I write about because for any data set, it’s important to describe what you’ve observed and how you’ve observed it and then not to go beyond that without changing the terms of the discussion to say this is what I think, based on the data. This is not what I saw in the data but this is a conclusion that I am drawing.
I think that that helps open up other research avenues for other people to go from where you’ve left off to the next step. I think it also keeps us honest and makes us better, I guess, filters for people. If we’re going to look at the data and describe what we’ve seen, then we need to be clear when we’re conjecturing and when we’re describing what the numbers actually said.
Nathan: Something that you mentioned is that data driven content tends to get a lot of back links, a lot of shares. I was just wondering if you could explain your perspective on why is using data in your content important or helpful for your audience then?
Susan: I think that it’s helpful and I think that it gets a lot of shares comparatively because it’s a scarce good. There’s a lot of data available to people and there are a lot of posts that are not about data. But there isn’t nearly as much content out there that is developed based on research. I think it’s valuable because it’s scarce. It’s not something that you could find everywhere.
I think it’s also valuable because we all want to know what the trends are. We want to know what people are interested in and so these records that we have people’s interactions online are really one of the best ways of determining those things.
But I do think the value really lies in the fact that it’s not the easiest content to create. When I wrote the LinkedIn post, I was a mess, the last week. I was looking at spreadsheets and shaking my head and drinking coffee and crying. I’m sure other people would be able to do it with less drama but for me, it was definitely an experience so I think that that kind of energy and effort that you put into a post like that doesn’t get duplicated as often as say a list of seven ideas about how to do a content marketing project. That’s much easier to produce I think.
Nathan: It’s almost like it stands out of the noise. You cut through the clutter, right?
Susan: Yeah, exactly. I think people are grateful. If they trust the source who’s analyzed the data, then it’s a time saver for them. They don’t have to go and do the research themselves. If you are a trustworthy source, your audience will appreciate what you do with data because it’s helpful to them.
Nathan: That brings me back to the old marketing clip that people buy from people they know, like, and trust.
Susan: Yeah, exactly. I think that that applies to the data. We have to figure out what people are interested in. We have to figure out how to meet them and how to build an audience so they know us, they like us because they’re interested in it. The trustworthy part I think is just in being careful in how you describe what you’ve done.
It’s very easy especially if like me, you’re writing a data post and it’s been a big event in your life with blood, sweat, and tears lost over it. It can be tempting to be more grandiose about what you’ve found and so I think that if you do that, you run the risk of losing trust with your audience. You just have to be careful in the midst of all the effort to earn their trust.
Nathan: Something that we talked about CoSchedule is that finding good data often starts by asking a good question. I was wondering from your perspective, how do you determine the questions you want to answer with data or how do you even start doing that?
Susan: Part of it is I have always been a curious person. I was the kid in class who got in trouble all the time for asking too many questions. They would just not pay attention to me after a while. Questions are part of my personality. But that’s a non-data driven answer. I also have a team of people who are very talented. They’re very analytical. They’re very curious. We will sometimes talk about our answers too.
I’m aware that choosing questions just from the pool of what one person or even a few people can think of is limiting, no matter how curious they are. There are limits to what you’ll come up with. As a team, we read a lot about the content marketing industry so that also helps. We’re hearing ideas and questions from other people.
We’re actually using a new tool that was developed by one of our founders called [00:15:18]. It’s an e-learning platform that aggregates information for teams by pulling it into shared dashboards. If one of us reads something, we can quickly alert someone else on our team to read it as well.
The engineers at BuzzSumo also developed a new tool called BloomBerry. It’s a question analyzer. It allows you to find the questions that people are asking. They’re doing that by looking all over the web for questions about a topic or a keyword. It includes forums like Amazon question and answer. It’s a great tool for those days when maybe you’re not feeling as curious to find out what people are asking about a topic or a keyword.
That feature is actually going to come into BuzzSumo at some point. Right now it’s available for free so if people want to start it out and start asking some questions by finding out what other people are asking, they can check bloomberry.com.
Nathan: Alright, Susan, let’s just say I’ve got that really good question. I know what I want to find the answer around. What does your process look like then for gathering that data or research?
Susan: In the ideal case, I’ll talk to our engineering team. They’ll go into the BuzzSumo database and pull out information in SESP or JSON format. I have no idea what that looks like on their side. The ideal for me is I put in Slack, “Hey, I need this.” In 20 minutes, someone hands me back what I need. That has happened but it doesn’t always. If I can’t get the data from the engineers who have access to the whole database, I will pull data from the BuzzSumo tool myself.
Paid subscribers to BuzzSumo can follow much the same process that I use for a year’s worth of data. I do have access to the entire database which goes back about five years. Let me stop and just go back a step because I don’t think I described this well. When I say I’ll pull the data from the BuzzSumo tool itself, what I mean is I’ll go into the most shared section of the site and enter my topic or a domain or a series of topics and domains and let BuzzSumo do those searches for me.
And then because it’s hard to analyze those things just by looking at them on a page, especially if you’re going to do a large number of them, I’ll go ahead and use the export function to get that information into either an Excel file or a CSV spreadsheet format. When I say I pull data from the BuzzSumo tool, that’s when I mean I do the searches there and then I export the information.
I can do that to get about five year’s worth of content. If you’re looking at BuzzSumo and trying to figure out how you could do this, one thing to keep in mind is that the spreadsheets that we export are 10,000 lines long. If you’re looking at a year’s worth of content in a spreadsheet, you’re only going to have 10,000 lines of content. The work around or the data hack would be to go ahead and search for shorter periods of time so that instead of having one spreadsheet with 10,000 lines, you’ll have 12 spreadsheets with 10,000 lines.
If you wanted to go really crazy, you could even export a week’s worth at a time and get more data that way. That’s how we get the data. And then I either use R or Excel to do a little bit more analysis of it. I would say even before the R or Excel analysis, there’s an interim step that is maybe half meditation, half imagination, and half staring into space. That’s basically just looking at it.
I sit and I commune with the spreadsheet until I feel like I got a little bit of an orientation to what’s actually in there. If something does jump out at me as I just look at those first 100, 200 lines, I’ll add that to my list of questions. Fairly quickly though, other tools are going to be necessary. That’s when I begin to use R. I do it in just a very, very rudimentary level.
Steve is actually better at using the R tool. I will use Excel to analyze the data by setting up formulas to determine what’s actually in the spreadsheet. There’s also a free online text analyzer that I use. Sometimes, I’ll use a word cloud too because that helps me visualize what I’m seeing in the language of the headlines and things that are in the spreadsheet.
If my data set is 50,000 lines long or more, R is a better fit but if it’s shorter than that, I tend to use excel more so than R.
Nathan: Alright, Susan. Let’s just say I have this monster spreadsheet. I’m looking for trends. How do you actually translate that data in the meaningful takeaways that you would write into the content that you’re going to publish for your readers?
Susan: The best way to define it for me would be to say that I look for patterns. I look for authors who repeatedly get high numbers of shares. Obviously, I’m talking about data that’s specifically from BuzzSumo. But I think in any data set, you’re going to have common factors like the names of authors or the names of headlines or the number of words in the article.
I try and look at each of those factors and determine if there are any patterns that emerge from that. For example, I might look for words that occur frequently in the most popular headlines for a topic or in the 1 billion Facebook post analysis that Neil Patel wrote up, we could see that Sundays were the days that got the highest number of shares on average. I think that the process is really looking for patterns.
And then I think another part of it is once you have the pattern, once you have the data, the final step that really pulls it together is where I think the teaching degree comes into play because you need to take all of that and package it in a way that someone who’s never seen the data can understand it. I think that it’s tempting to maybe not put as much time into that step but packaging conclusion so that they’re understandable is really important.
I also think asking the so what question is important as well. I mentioned earlier the Facebook analysis that I did with the 3,000 videos. As you look at it and you realize, “Okay, animal stories are really popular.” Well, so what? What does that mean to my particular audience? For BuzzSumo, we have a lot of B2B clients. I was able actually to find a couple of examples of companies using animals to help tell their stories.
One of them was Oracle. They had an employee who had nine Labrador Retriever puppies for whatever reason and they used that in a video. Helping your audience connect this data with their particular use case is I think really an important part of data driven content marketing too. You can’t just list it all out and say now go and do what you will. I think you have to help people connect it to what they’re doing.
Nathan: That’s awesome advice, Susan. I want to pick on this because the last time we chatted, you mentioned you weren’t always an Excel or spreadsheet wiz but a lot of this is like involving analyzing this data with a tool like that. How did you learn that skill?
Susan: Let’s back away from the wiz part. I would not say that I am an Excel or an R wiz. That should be good news for people who aren’t either because even if you’re not an expert, those tools are so great that you can tease insights out of using those tools out of a data set even if it’s not your background and your history to have done a lot of analysis. That’s the caveat or the disclaimer.
For me, I guess I did it the old fashioned way. I took a class. I did a Udemy class for the Excel spreadsheets. I think I’ve done two of those. I finished it. I’ve signed up for Udemy classes before and done three lessons but this one, I knew that I needed to get better at Excel. I had done some of the free tutorials. For me, before I took the Excel class via Udemy, I was even lacking some of the basic vocabularies. The tutorials weren’t helpful because they would jump in at a higher level than I was. I think it was a six hour class. It went back to the very beginning. This is what a row is. This is what a column is. I sat through all that so I could really understand the program.
Now that I’ve done that, if there’s something I don’t know how to do, I find that it’s much easier to use a YouTube video or something like that just to find out more. It’s almost like the class laid the framework that has enabled me to access a whole other set of resources that are not quite as structured.
For R, I took an edX class and worked through that. I should actually say that I also tried one at MIT, one of their online ones. Totally failed, bombed, I was like, “Okay, I can’t do this.” And then I realized that part of the reason I couldn’t do it is because I didn’t have the basics again. I found a simpler class with lower prerequisites.
The short answer is I took a class. The long answer is I tried online tutorials and a really hard class and failed in both of those and just kept at it until I found one that was basic enough to bring me up to a level where I could actually do the projects I needed to do.
Nathan: I think that’s great because I think a lot of people are in those beginner shoes. That leads me to my last question for you, Susan. Let’s just say I’m a complete beginner at this. How would you recommend that I start generating data driven content?
Susan: I think the first question is do you have data. If you don’t have data, you’ll need to either find some pre packaged data groups online. You could just Google data sets and they’re out there. I think there are some ones that are weather data, things like that. If you want to make it specific to content marketing, you can use tools like BuzzSumo to find data that’s relevant to the projects that you’re doing.
I think I would begin with Excel. The reason is I think that the platform is familiar enough and it’s a little bit clearer how the logic works in Excel. What I found was once I understood the Excel logic a little bit better, I was able to understand the R logic. I think of the two, I would begin with Excel, get a grip on formulas that allow you to do averages and even do some nested formulas so that you can understand if, then statements within Excel.
Once you’ve done that, then I would suggest the edX course on R. I also am enrolled in DataCamp, which is I think a Microsoft project. That allows me to tap into ongoing R classes and it also allows me to get a refresher if there is an analysis I want to do in R and I don’t remember how to do it.
I think for me, if you are someone who like to learn in a group, if you can identify people who are interested in this type of stuff, that can also be a huge help. I’ve benefited a lot from having Steve Rayson who is one of our founders who does this and has done it more than me. I can bounce ideas off of him or if I get stuck, I can send something to him and he’ll look at it. I would say community is a great thing if you can get it.
Nathan: Yes. Susan, I think all of that is really great advice. I just want to say thanks for being on the podcast today and thanks for sharing everything that you know about data driven content.
Susan: Nathan, thank you so much. I really appreciate talking with you today. I had a great time on the podcast. Thank you so much.
Nathan: At CoSchedule, research is one of the standards of performance in our marketing strategy. Some of the best performing content we’ve ever written is 100% based on answering our audience’s questions through data driven research. Susan, thank you so much for sharing your take on data driven content. This is awesome advice and I know for a fact it drives traffic, backlinks and conversations.