SEO

How Not to Sully Infographics with Drunkenness: Lessons in Link Building

I ignored the opportunity to link build with infographics for far too long. As an engineer turned search engine professional, the idea of creating art scared me. I prefer complicated puzzles and problems over design, but after seeing so many others succeed I decided to go all in.

So, a few months back, we created a Beer in Colorado graphic. I’m by no means an infographic expert, which you’ll see as you read on, but some of the work we did will hopefully be useful for your next (or first) infographic project.

colorado beer facts 600px How Not to Sully Infographics with Drunkenness: Lessons in Link Building

Promotion Strategy

We chose Colorado beer as our topic, because we wanted to piggyback onto Colorado’s annual Great American Beer Festival (GABF). In the beer blogging world GABF is akin to the “Superbowl of beer.” The number of beer bloggers that cover this event is astonishing, which is what got me salivating at the link prospects. However, the toughest part would be reaching those bloggers. We chose to focus on four different promotion strategies.

1 – Contacting Beer Bloggers and Beer Associations Directly

We launched the graphic a few weeks before the event by contacting individual beer bloggers and beer-related sites to ask them for feedback on our graphic. Every email was personalized, and we reminded them of the upcoming beer festival with the hope that they would post our graphic on a page or post dedicated to the GABF.

2 – Contacting Infographic Bloggers

Here is a list of a few of the infographic sites that we submitted our graphic to and/or picked our graphic up spontaneously. If you’re like me, you’ll bookmark these for your next infographic promotion:

3 – Press Release

We used Google Trends to look at the most popular GABF press coverage from 2009 in order to determine the best timing for our press release and press contact. The most significant stories about the event occurred the Saturday before and then the day of, so we timed our 2010 press outreach and release accordingly.

4 – Social Media

We ran an ad on Facebook targeting beer fans in Colorado with an ad headline designed to challenge viewers. It read: “Think you know beer? Check out this Colorado beer trivia.” We had limited success, but it did help our page get some spontaneous “likes.”

We used Twitter to shout out to every blogger that included our graphic on their site, as well as SponsoredTweets to get the word out to bloggers early in the process. Finally, we promoted the infographic on our own blog and on our Facebook page the week of the GABF.

Lesson: Infographics Should Contain Hearty Data

The most stinging criticism from the infographic bloggers we contacted had to be that our infographic was light on data. Unlike some of the most successful infographics we bench-marked, we didn’t include a data-intensive graph or diagram. The reason? I didn’t think any of the numerical data I found was very interesting.

Of course, after the fact, I received half-dozen suggestions on data possibilities, and I struck the palm of my hand against my head with every piece of advice I received. Why didn’t I think of that?

Lesson: An infographic needs some strong numerical data to meet the standards of infographic critics, as well as to have some value as a reference material.

Lesson: Get Community Buy-In Before You Begin

I did not bring beer experts into the project early; my second regret. We contacted a knowledgeable beer blogger as well as a local brewing advocacy group for feedback after our graphic was done, and in both cases I learned of a missed opportunity.

The blogger would have helped us promote the graphic had we done something specific to her state (which would have been easy to create side-by-side with our Colorado graphic), and the brewing advocacy group would have been on-board had we not included the beer fact “you get drunk faster in Colorado.” They said they would have been happy to help us if we hadn’t sullied the image of beer drinking with drunkenness (my words, not theirs).

The point: Had I contacted these people earlier and gauged their interest in participating, we might have had a much more successful product.

Lesson: Design For The Web

When we conceived our plan, I wanted to be sure that the final graphic would be available in both web and print resolutions. As a result, our designer started his design with print resolution in mind and then scaled things down to web resolution, which caused a problem: the text from the print version was too small to be legible in our web version when scaled down. As a result, we had to create two distinct graphics – one designed for web, and one designed for print. I would have saved some time and money had we planned for this originally.

In terms of the best size, I found that our 600px wide graphic had the best embed rate. The 500px wide graph, despite being fairly legible, wasn’t quite big enough for most bloggers that used our graphic.

Results

Our graphic garnered dozens of good quality links from beer bloggers, exposed Spork Marketing to local breweries, and helped me develop a process for our next infographic. Our cost-per-link was incredibly low when compared to some other projects we’ve done—low enough, in fact, to be cost-effective for some of our small business clients.

 How Not to Sully Infographics with Drunkenness: Lessons in Link Building

Jason Lancaster

Jason Lancaster is President of Spork Marketing, a Denver Internet marketing company specializing in search engine optimization, marketing, and web design.

Comments are closed.

20 thoughts on “How Not to Sully Infographics with Drunkenness: Lessons in Link Building

  1. Nice tips. Attempting an info graphic seems easy enough when you look at already created works but when you decide to attempt one yourself you soon realize how daunting it can be.

  2. Well-done post that I’ll refer to in the future, thank you!

    Regarding the lesson about being data-heavy, do you have any suggestions for brainstorming what kind of stats to look for, besides asking other people? Search queries you might use, for example?

    1. If I had to do it over again, I would have found two dissimilar concepts that were both relevant to Colorado (say skiing and John Denver) and then try to find some data that related the two.

      I also found some data on some beer industry sites, but it just seemed too boring…but looking back it would have been good to toss something in.

      Hope that helps.

    2. If I had to do it over again, I would have found two dissimilar concepts that were both relevant to Colorado (say skiing and John Denver) and then try to find some data that related the two.

      I also found some data on some beer industry sites, but it just seemed too boring…but looking back it would have been good to toss something in.

      Hope that helps.

  3. Well-done post that I’ll refer to in the future, thank you!

    Regarding the lesson about being data-heavy, do you have any suggestions for brainstorming what kind of stats to look for, besides asking other people? Search queries you might use, for example?

  4. Researching your target audience, the dos and don’ts would have matter a lot. The drunken word could have been avoided. A lot of campaigns failed because of this, the target audience were turned off at once upon seeing the word not appropriate for their group.

    1. Steve – I thought long and hard about using the fact “You get drunk faster in Denver”, but after reading the beer blogs that we were targeting for placement, I didn’t get the sense it would be offensive.

      If we hadn’t used that fact, a beer advocacy group might have helped us promote it. Of course, it might not have been as entertaining without that fact…not sure if it would have helped or not.

    2. Steve – I thought long and hard about using the fact “You get drunk faster in Denver”, but after reading the beer blogs that we were targeting for placement, I didn’t get the sense it would be offensive.

      If we hadn’t used that fact, a beer advocacy group might have helped us promote it. Of course, it might not have been as entertaining without that fact…not sure if it would have helped or not.

  5. Researching your target audience, the dos and don’ts would have matter a lot. The drunken word could have been avoided. A lot of campaigns failed because of this, the target audience were turned off at once upon seeing the word not appropriate for their group.

  6. Never would have considered the heavy data part you mentioned – great tips! Added the infographic bloggers to my feedreader as well, thanks for the info.

  7. Yup, great post Jason….as you did, we too noticed the use of infographics for SEO serp success…and have blogged about it quite a bit over the past year, eh!

    Works. It’s that simple!

    :)

    Jim

  8. when you are dealing with ‘connoisseurs’ of anything – they get testy over small things that might not even hit your radar. ever meet someone who is an “audio enthusiast” and claims to be able to hear frequencies and noises that would probably even elude Spiderman’s power senses?

    i wonder if you would have more success if you went after a more mainstream market/event and less populated by highly opinionated experts who would take offense to seemingly innocuous and obviously good-natured statements like that one.

    maybe an event where the topic is not taken so seriously, or humor is encouraged with data infographics. i do know plenty of brew-meisters and they take it very seriously – sometimes more than wine experts.

    this was a very good post. nice idea and appreciate the sharing of your test and techniques.

    1. Thank you! I think you’re right about the risks of diving into such a niche community. We felt pretty confident tackling beer, but I can see now that we were a little too confident.

      Still, as I write this the infographic page is one of the most linked to on our site, and it continues to generate new links, so mark me down as ‘Very Happy.’ :-)

  9. Researching your target audience, the dos and don’ts would have matter a lot. The drunken word could have been avoided. A lot of campaigns failed because of this, the target audience were turned off at once upon seeing the word not appropriate for their group.