What’s the Best Content For Links and Shares? 5 Ways To Use Data to Win

Share This
Andy Crestodina

Content comes in all shapes and sizes. Helpful how-tos, news, guides, opinions, roundups and rants. But there is one type of content that stands above them all.


This format crushes almost anything else you can publish. And it’s not as hard to get into your content mix as you might think. Here’s a quick how and why on research as a content format…

Why Research Wins

Original research is the most powerful and effective format for content for many reasons.

  1. New research is original and owned by you, making your site the primary source
  2. Numbers are convincing and lend credibility
  3. Data can support your sales messaging
  4. The outreach involved in research grows your network
  5. Research gets more links and more shares than any other format for content

That last point is important since links and shares are the key to search and social, the two big sources of traffic.

The Data: Analysis of One Million Articles

This was discovered, through research, in a collaboration between BuzzSumo and Moz. They analyzed 1 million articles and found that most content doesn’t get traction in search or social. 75% of the articles had never been linked to, and 50% had fewer than two Facebook Interactions.


This chart shows the stark reality for marketers. The vast majority of content get no results.

But, as we all know, some content gets tons of traction. According to the research, this content tends to fall into two categories:

  1. Opinion forming journalism
  2. Well researched and evidence-based content

Creating opinion forming journalism on a regular basis is difficult for most of us, so let’s focus on that second group, research.

Most content gets almost no traction in search or social. Research-driven content is the exception.

The rest of this post is a guide to producing original research.

The 5 Types of Research

There are five types of research you can add to your content mix.


Research Type 1: Results from Experiments

test, analyze, publish, promote

You tried something. Did it work or not? Anything that you or your company tried is potentially publishable research, as long as it produced a datapoint. There’s an audience out there that wants to see your findings.

It’s a case study with a different angle. The key is to publish a conclusion that is relevant to a broader audience. It’s not about you. It’s about the truth.

Example: The A/B Test Results on the Mad Mimi Pricing Page

Copy Hackers has a data driven approach to helping clients. They run tests to measure the results of their work. Sometimes, they publish the results of those tests.

A great example is the Mad Mimi Pricing Page. The experiments they ran resulted in a 500% lift in conversions. It also produced a beautiful, visual piece of content.


This case study experiment and its results have appeared in many marketing presentations. It’s also earned 21 links and 107 shares.

Research Type 2: Observe and Analyze

pick your dataset, gather data, analyze

No need to produce new data. Just find it in the wild and run your analysis. Look for patterns, form a hypothesis, gather information and analyze. The Moz/BuzzSumo study cited above is a great example. Here’s another:

Example Article: Web Design Standards

A client once asked us if search tools are standard on websites, our instinct was to give a quick answer: it depends. But instead, we decided to find out.

  • Step 1: We needed a set of sites to check. So we went to Alexa and grabbed a list of the top 50 marketing websites. Alexa is a good source of lists of websites, but really any third-party source would have worked.
  • Step 2: Find a freelancer to review the sites to see which had search tools. We added nine other questions while we were at it.
  • Step 3: Turn the data into visual charts (10 pie charts and a roundup bar chart, see below).
  • Step 4: Write the article and share.

Web Design Standards: 10 Best Practices on the Top 50 Websites


The post was far more successful than a typical post on this blog, earning 178 links and 610 shares. Also, the post is optimized for the phrase “web design standards” so it gets a few hundred visits per week.

ga-gragh 1

Research Type 3: Aggregate Existing Research

new insights on a well-covered topic

When the research is already out there, you can combine and repackage it into new metrics and statistics. The new piece is often more credible since it’s an objective look at subjective data.

Example: Who was the best US president?

Tough question. A lot of people have tried to answer it by surveying scholars and historians. Rather than creating a new survey (time-consuming, expensive), an editor at Wikipedia aggregated past surveys and applied a simple methodology.

It’s not new research, but it’s new data, and it gives a highly credible answer to the question and creates a list of historical rankings of the best presidents (Lincoln, FDR, Washington are the top three).


It’s also gotten lots of traction. This Wikipedia page has 214 links and 1000+ shares. That’s a lot of shares for a Wikipedia page. Imagine how aggregating existing research could be a huge hit in your marketing mix.

Research Type 4: Online Surveys

massive outreach, big numbers

This is the classic approach. Get a group of people to answer a few questions in an online survey. Analyze the data. Publish.

There are many benefits to this kind of research:

  • The final results can be shared with the respondents.
    When people take the survey, they can enter their address to get the results. You are building a list as you are gathering the data. Once you publish the results, email the respondents.
  • The massive outreach is good for brand awareness
    Any project that involves a lot of outreach will cause a spike in attention to you and your brand.


  • The data supports your sales pitch
    The statistic that you discover will help you sell. Imagine a sales rep for cost tracking software explaining how 40% of companies lose track of their expenses. Or the mouthwash company showing that 76% of people say bad breath is a “dating dealbreaker.”

Note: Since this was one, multiple choice question, it’s really a poll, not a survey.


Example: Annual Blogger Survey

We first discovered the power of research two years ago when we asked 1000 bloggers eleven questions about how they work. This produced new statistics, new soundbites and new answers to old questions.


It was a hit, earning 644 links and 1700 shares. The results were so big that we’ve turned it into an annual survey. If anyone writes an article about how much time it takes to blog, it’s only natural to reference our research, showing the average post takes 2.5 hours to write.

Repeat your research to show trends. Trends are fantastic soundbites, great for sharing and linking.  If anyone writes about blogging trends, it’s only natural to reference the article showing that the typical post is now 900 words, 100 words longer than last year.

ProTip: To create data that is statistically significant, put the sample size and total populations into a statistical significance calculator, to measure the margin of error.

Getting ideas for how an online survey could work for you? This next idea is even more powerful…

Research Type 5: Phone Surveys

smaller dataset, higher touchpoint

Online surveys are about quantity; phone surveys are about quality. You gather less data, but in a more direct way. You pick up the phone and call people.

The hidden benefit to this is obvious. This research gives you an excuse to reach out to high-value connections, like sales prospects and influencers.

The call starts with you asking questions about the industry. It’s a brief interview. It may or may not lead to other topics. It ends with an offer to share the research once complete.

Here are examples of how phone surveys, subjects and content outcomes.

  • Do you sell social media services?
    Call 50 marketing directors and ask about their biggest social challenges.
    Possible content: 68% of marketing directors have trouble measuring social media ROI.
  • Do you sell supply chain software?
    Call 50 logistics professionals and ask them what tools they use.
    Possible content: Only 16% of logistics companies have a transportation management system in place.
  • Do you do technology staffing?
    Call 50 IT directors and ask them about skills gap on their teams.
    Possible content: 4 out of 5 tech hiring managers are looking for QA/testing experience.

Getting a few more ideas?

This is a powerful way for B2B marketers to get right in front of decision makers. The key to success is to be very considerate and follow up.

Example: The Seven Skills You Need to Thrive in the C-Suite

What are the most important skills for succeeding at C-level jobs? It’s a subjective question. But Boris Groyberg produced a high quality, research-based article on the topic by surveying “several dozen top senior search consultants.” It was published in the Harvard Business Review.


This page won 35 links and 351 shares. The networking benefits were probably priceless.

But What Should I Reasearch? Find the “Missing Stat.”

In every industry there are common statements. But they aren’t backed up with data. We call these gaps “missing stats” and they are ripe for research. These statements meet two criteria. They are:

  1. Frequently asserted
  2. Rarely supported

If you can find this gap in your industry and produce the research that fills it, you’ll have something truly share-worthy, link-worthy and even press worthy. Seriously, you may want to reach out to the media.

If Research Is So Powerful, Why Doesn’t Everyone Do It?

Because it’s work. It requires forethought and planning. It’s not more difficult, it’s just more time-consuming. It will take ten times as long to create, but you’ll to get 100x the results.

Don’t take shortcuts; they take too long.  –  Sonia Simone, Copyblogger

If you’re ready for a huge upgrade to the quality and results of your content marketing, here’s where you can begin:

  • Plan at least one research project per year
  • Answer a question that is unanswered in your field
  • Gather data in a way that connects you with people, either quantity or quality
  • Package it into soundbites, stats and visuals
  • Share the results with your audience, your subjects and industry publications
  • Repeat the research next year

The world is not waiting for another medium-quality blog post. Publish something truly original. Make a contribution to your field. Find the truth and share it with the world.

Share This

What are your thoughts?

By signing up you are agreeing to our Privacy Policy.

Comments (12)
  • Oh yeah, I’m bookmarking this one. If this doesn’t inspire people to go out and create research based content, I don’t know what will. I love the web design standards idea btw.

    • Thanks, George. Orbit isn’t a web marketing company, but if we were, I’d call it Orbit Research & Marketing. And I’d create and promote original research for all of our clients. The results are so much better, I wonder why I do any other content!

  • Andy–

    I totally agree that research is a winner no matter how you use it in your content marketing. The top bloggers (like you) include data points in every major point of their body text.

    Original research is a great piece of foundational or pillar content that can be updated on an annual basis. One suggestion: Use a PR release to help it have wings.

    Alternatively, round up other people’s research and update it annually.

    Happy marketing,

    Heidi Cohen

    • Thanks, Heidi. I agree that press releases are definitely appropriate in this case. This is one example when the media may perk up and notice!

      You also mentioned another important point: just using data in general. I know you do this too, every post can include data points, even if it’s not original research. If you don’t support you points with facts, the content is weaker. Weak content doesn’t win.

      Thanks for the comment, HC! Talk to you soon. 🙂

  • Andy, as always, insightful thought leadership here. Thanks so much for sharing.

  • This is absolutely our experience. For example, when promoting the work of our wonderful market research client, Zeldis Research Associates, we cannot reveal the results of the proprietary custom research they conduct for their clients, such as customer satisfaction studies. But Zeldis recently conducted the same study for themselves to find out the criteria on which clients rank market research firms. And we are sending out these results to their clients and prospects. As you have always preached, Andy, content needs to be relevant and useful to the recipient, and nothing is more useful to them than research they can use and quote.

    • You bring up a good point, Nancy. A lot of research is done on behalf of clients, and can’t be used as content. So your example was perfect. Research needs to be created specifically as content for marketing, not for internal purposes or for clients!

      It would be nice if those clients let you take their names off and publish those studies, wouldn’t it?

      Thanks for the comment, Nancy!

  • Thank you, very useful article!

  • a fantastic inspiration for the researchers, this information can move any person with interest in search engine optimization to great level.

  • Any suggestions on the best places to hire freelance researchers?

    • Hi Mark. Just seeing this (fabulous) article. At the risk of sounding self-serving, we recently launched a new consultancy focused on helping marketers create and amplify the type of research Andy talks in this post. Reach out anytime to @michelelinn if you need help — or just want to brainstorm.

  • Wow, Andy you’ve really got me thinking. And here I come from a research background! Yeah, what a great way to spice up an article. I also think collecting feedback/survey date etc. creates a nice interactive element and particularly so, when you can give the results back to the respondent.

    What I appreciate about your article – and I don’t think I got the full impact of it before – is how research can help the customer better understand how a product might be useful (Like the Listerine example). It paints a picture for them. I mean we create products to fill a need. Why not make sure our potential customers see it from that perspective.

Join over 16,000 people who receive bi-weekly web marketing tips.

By signing up you are agreeing to our Privacy Policy.

Share This