The Fractl team has worked on hundreds of content marketing projects. Along the way, we’ve kept track of a lot of data, including everywhere our client campaigns have been featured, what types of links each campaign attracted, and how many times each placement was shared.

While we regularly look back on our data to evaluate performance per campaign and client, until now we’d never analyzed all of these data in aggregate. After combing through 31,000 media mentions and 26,000 links, here’s what we found.

Most high-authority links don’t receive a lot of social shares.

Most marketers assume that if they build links on high-authority sites, the shares will come. In a Whiteboard Friday from last year, Rand talks about this trend. BuzzSumo and Moz analyzed 1 million articles and found that over 75 percent received no social shares at all. When they looked at all links – not just articles – this number rose to around 90 percent.

We (wrongfully) assumed this wouldn’t be the case with high-quality links we’ve earned. It turns out, even the majority of our links on sites with a high Domain Authority (DA) didn’t get any social shares:

• 52 percent of links with a DA over 89 received zero shares.
• 50 percent of links with a DA over 79 received zero shares.
• 54 percent of links with a DA over 59 received zero shares.

On average, our campaigns get 110 placements and 11,000 social shares, yet a single link accounts for about 63 percent of total shares. This means that if you exclude the top-performing link from every campaign, our average project would only get 4,100 social shares.

Since most links don’t yield social shares, marketers with goals of both link building and social engagement should consider a strategy for gaining social traction in addition to a strategy for building a diverse link portfolio.

The social strategy can be as simple as targeting a few key websites that routinely yield high social shares. It’s also helpful to look at target sites’ social media accounts. When they post their own articles, what kind of engagement do they get?

Of all the sites that covered our campaigns, the following five sites had the highest average social shares for our content. We know we could depend on these sites in the future for high social engagement.

Exceptions to the rule

Some content can definitely accomplish both high engagement and social shares. The BuzzSumo and Moz study found that the best types of content for attracting links and social shares are research-backed content or opinion pieces. Long-form content (more than 1,000 words) also tends to attract more links and shares than shorter content. At Fractl, we’ve found the same factors – an emotional hook, a ranking or comparison, and a pop culture reference – tend to encourage both social sharing and linking.

Few sites will always link to you the same way.

To ensure you’re building a natural link portfolio, it’s important to keep track of how sites link to your content. You’ll learn if you’re earning a mix of dofollow links, nofollow links, cocitation links, and brand mentions for each campaign. We pay close attention to which types of links our campaigns earn. Looking back at these data, we noticed that publishers don’t consistently link the same way.

The chart below shows a sample of how 15 high-authority news sites have linked to our campaigns. As you can see, few sites have given dofollow links 100 percent of the time. Based on this, we can assume that a lot of top sites don’t have a set editorial standard for link types (although plenty of sites will only give nofollow links).

While getting a site to cover your content is something to be celebrated, not every placement will result in a dofollow link. And just because you get a dofollow link from a site once doesn’t mean you should always expect that type of link from that publisher.

Creating a lot of visual assets is a waste of time in certain verticals.

There’s an ongoing debate within Fractl’s walls over whether or not creating a lot of visual assets positively impacts a campaign’s reach enough to justify the additional production time. To settle this debate, we looked at our 1,300 top placements to better understand how publishers covered our campaigns’ visual assets (including both static image and video). This sample was limited to articles on websites with a DA of 70 or higher that covered our work at least four times.

We found that publishers in different verticals had divergent tendencies regarding visual asset coverage. The most image-heavy vertical was entertainment, and the least was education.

Some of the variation in asset counts is based on how many assets were included in the campaign. Although this does skew our data, we do receive useful information from this analysis. The fact that top entertainment publishers used an average of nine assets when they cover our campaigns indicates a high tolerance for visual content from outside sources. Verticals with lower asset averages may be wary of external content or simply prefer to use a few key visuals to flesh out an article.

Keeping these publisher vertical preferences in mind when developing content can help your team better allocate resources. Rather than spending a lot of effort designing a large set of visual assets for a campaign you want to be placed on a finance site, your time may be better spent creating one or two awesome visualizations. Similarly, it’s worthwhile to invest in creating a variety of visual assets if you’re pitching entertainment and health sites.

Analyzing our entire link portfolio taught us a few new things that challenged our previous assumptions:

• High DA sites don’t necessarily attract a lot of social engagement. Just because a site that linked to you has a huge audience doesn’t mean that audience will share your content.
• Most sites don’t consistently use the same types of links. Got a dofollow link from a site one time? Don’t expect it to be the norm.
• Certain publisher verticals are more likely to feature a lot of visual assets. Depending on which verticals you’re targeting, you might be wasting time on designing lots of visuals.

While I hope you’ve learned something from Fractl’s internal study, I want you to see the broader lesson: the value of measuring and analyzing your own content campaign data as a means to improve your process. If you’ve done a similar analysis of links earned from content marketing, I’d love to hear from you in the comments.