Fractl has created thousands of content marketing campaigns on every topic. For the past seven years, we've kept a close eye on every campaign to refine and improve the content we create on behalf of our customers.

In my last post for Moz, I explained how you can set realistic digital PR expectations for your content based on your niche. In this topic, I want to dive a little deeper into the data and share insights on how the source of your content can be just as important in determining the performance of your content.

In this analysis, I examined 1,474 customer content campaigns in six different data source categories:

  • Customer data
  • Social media
  • Participatory methods
  • Publicly available data
  • survey
  • Germ swab

It is important to note that Fractl has countless other data sources that we use daily for content campaigns that are not mentioned in this article. In this analysis, each category has at least 20 campaigns, while some categories have several hundred campaigns.

It is also important to note that averages were determined by excluding the top outliers. For campaigns that went "viral" and were far above the norm, we excluded them in the calculation so as not to distort the average values.

In addition to sharing link and press averages, I will also explain how printable, shared content can be created from each data source and provide examples.

Manage expectations across different content types

In the total sample of 1,474 campaigns, one project received an average of 24 dofollow links and a total of 89 press releases.

A press mention is defined as each time the content campaign is mentioned on a publisher's website.

There were some averages for individual data source categories that matched the sample mean, while other categories differed significantly from the sample mean.

Publicly available data

For almost every niche out there, you can bet that a publicly available record is available. Some examples are data from the CDC, the United States Census, colleges and universities, the WHO and TSA. The possibilities of using publicly available data as a method for your content are really unlimited.

Free records can be a treasure trove of information for your content. However, keep in mind that they are not always the easiest to work with. They require a lot of analysis to understand the enormous amount of information in them and to make the findings digestible for your audience.

For example, take a campaign we created for a customer called Neighborhood Names. The data was free from the U.S. Census, but to make sense of it, our researchers had to use QGIS, Python, Text Mining, and Phrasemachine (a text analysis API) to restrict them to what we were looking for.

And what were we looking for? Looking at neighborhood names across America may seem boring at first until you find certain words match prosperity.

I was the outreach specialist for this project, and using the wealth angle I was able to get two notable placements at CNBC and a press mention at MSN. The project quickly took off on the Internet and received 76 dofollow links and 202 press releases by the end of our reporting period.


Unlike searching the internet for free data, using a survey as a method can be more expensive. Apart from that, using a survey to design your content has one big advantage: you can find out everything you want.

While publicly available data tells a story, that's not always the story you want to tell, and that's where surveys come in.

When it comes to surveys, of course, anyone can create one without looking at best practices for research methods. That is one of the problems we have to address. Given that "false news" is on everyone's lips in 2020, building trust among journalists and editors is paramount.

As content creators, we are responsible for ensuring that content is not only eye-catching and entertaining, but also correct and informative.

In the case of survey campaigns in particular, you must analyze the answers under strict methodological criteria. When collecting data for surveys, pay particular attention to ethical preservation, data validity and fair visual representations.

Germ swab

In my personal experience, spotting campaigns are the most fun and often the most disruptive. Fractl did some research on the emotions that cause content to go viral some time ago, and spotting campaigns often hit all the right emotions in the virus equation.

Negative emotions such as disgust are often evoked when the results of spotting campaigns are checked. Our study found that negative emotions, when paired with emotions such as anticipation or surprise, can still achieve viral success (internet virus, not germ virus). What is more surprising than finding out that the airplane table is dirtier than a toilet seat?

Publishers around the world also seemed to find the content surprising. This campaign exceeded the norm for a typical content campaign with 38 dofollows and 195 press reports – and that before the COVID-19 pandemic.

Participatory methods

Participatory methods are campaigns that require active participation in the methodology. These are unique ideas – no two are the same. Some examples of campaigns that fall under the category of participatory methods are when team members conduct a 30-day squat challenge, respondents are asked to draw brand logos from memory, or when we literally use a dash cam from DC drove to NYC to record traffic violations.

There is a certain risk associated with these campaigns. They require a lot of preparatory work and planning without the promise of a return – and that is scary for customers and our team who have made enormous efforts to pull them off.

However, as you can see from the table above, these ideas were implemented along with other campaign types and even better than the survey methods for both the number of dofollow links and the mentions in the press. To get big benefits, you have to be willing to take a big risk.

Social media

Social media as a data source is almost child's play with survey methods and publicly available data sets. Unlike campaigns with participatory methods, you don't have to leave your computer to create a campaign based on social media data.

In the seven years of content creation, Fractl has created campaigns based on data from Twitter, Instagram, Facebook, LinkedIn, Reddit and others. From this experience, we know firsthand which types of social campaigns work and which ones fall flat.

The best thing about using social media as a source of content is that it can be applied to all industries.

The biggest lesson we learned from creating content based on social media data is that the methodology is usually subjective. So you need to keep the project light-headed to get full coverage.

For example, we created a campaign for a customer in which we looked at Instagram posts with the hashtag #sexy and a geolocation. From this, we were able to determine the "sexiest" countries in the world as well as the US states.

While it would be impossible to know what the sexiest places in the world were (what does that mean anyway?), We were able to create a fun campaign that used bait to bring light-hearted publishers like Glamor, E! Online, Women's Health and Elite Daily.

Make sure that whatever you produce contributes to an ongoing conversation. Statistics that don't point to anything are irrelevant to authors who are actually trying to add something to the conversation.

Customer data

Customer data is often the most underrated data source for content marketers. You may be sitting on a wealth of actionable industry insights and you don't even know it.

You can only think of internal data as useful for improving your internal work processes, but it can also be useful outside of your company.

Unlike publicly available data, internal data is never seen before and is 100% unique. Journalists eat this up because they only provide exclusive resources.

For example, think of this article. This article contains data and insights that Fractl has gained after creating thousands of content marketing campaigns.

An additional advantage of using internal data to create your content is that, according to our analysis, it is comparable to surveys. Unlike surveys, however, it's completely free.


Regardless of the methodology or vertical for which you create content, it is important to know that we as content creators have an ethical and moral responsibility to create it for an audience.

Given that "false news" is on everyone's lips, it is of the utmost importance to build and maintain trust in writers and editors.

All content that you produce and promote must be assessed through a strict methodological lens to ensure that the content is correct and informative, as well as eye-catching and entertaining.

Regardless of your methodology, contributing to the fake news epidemic if you don't take the right steps to make sure your data sources are correct.


Please enter your comment!
Please enter your name here