Buying Awareness to Build Links

It's a common dilemma. You work hard to publish a great piece of content, but no one notices it. How can you build an audience when you don't have one? One solution is to buy the initial traction, which I explored in a recent presentation at a meetup in Seattle.

I apologize for the poor sound quality... and my pauses. I wanted to share some additional details on why I'm talking about this, how I did it, and what the results looked like.

Traditional Paid Advertising vs. Content Advertising

In paid advertising, we worry a lot about the cost of paid advertising per sale (or lead). Accordingly, category and product pages are by far the most common landing page. While a novel, interesting, or remarkably superior product might get some natural links, they almost never go viral. Building links is not what category and product pages are designed to do; they're built to convert.

But what about using paid channels to build links? What happens when we design a paid advertising campaign to get links by pointing ads to content that is more likely to go viral? In the video above, I tried to make the point that you can turn two knobs to increase user interest. First, you can make sure that your content is interesting to the places you want to advertise to. Second, you tweak and adjust your ad targeting to bring in those who would be most interested.

With this in mind, we can begin to use channels that have traditionally been used for sales to build links, branding, and overall awareness.

Which Ads Worked Best?

It's usually impossible to know which channel to attribute a link to. To really know which advertising channel builds the most value - meaning links, sales, awareness, email subscribers, etc. - we would have to repeatedly test advertising the same piece of content through various channels. Even then, inconsistent variables like a share from a large blog can skew the data.  Most of us have neither the time nor the budget to run a test that would be considered statistically valid.

That said, it would be foolish to simply throw our money to the wind and hope that links come back. Instead, we can run a test like the one I ran above, and then analyze user metrics to determine the degree to which the audience was engaged. More interest (insofar as we can determine it) is more likely to lead to more sharing and thus more profit. Here are my findings for this particular campaign:

  • Advertising the page on Facebook was one of my favorite campaigns, and I didn't even target interests like I should have.
    Facebook ad results for PPC SEOThat CPC might look high, but I ultimately ended up with about 1600 Facebook visits and around 150 new followers. Given that 90% of the Facebook traffic happened while this was the only campaign running, I'm fairly confident this is a result of the ad. 
  • This post did best on industry-specific display ads, with an average time on site of 2 minutes and CPC under 10 cents. Availability, demographic interest, and pricing will vary wildly here.
  • Reddit provided the lowest CPC by far (less than a penny), but also the lowest time on site with a pathetic average of just 8 seconds - perhaps just long enough to see that there weren't any cats. Then again, the high bounce rate means this number is probably artificially low.
  • Adwords was by far the worst channel for the money. I bid on low-competition keywords with an "estimated CPC" of $.10, but only received impressions at higher rates. With a final average CPC of $.40 - $.70 average CPC for purely informational phrase-match keywords, I feel like Google doesn't have room for ads that aren't focused on leads or sales.

Using Ad Targeting to Influence Content Targeting

I also tried StumbleUpon paid discovery, as Ian Lurie's suggested. These weren't included in my original figures.

What I love about StumbleUpon is that you can easily see whether you're actually appealing to the people you thought you were appealing to. My post clearly appealed to the philosophy crowd, while also connecting with people following health-related topics. If I thought this was a great piece for science blogs (hint: I did), I might not be right on target.

Alas, the organic stumbles never amounted to much, and I didn't see many new links. That may be a result of the post now being a little dated. Still, the figures here tell me that lots of people were engaged for longer periods of time. So how can I use that data?

This is a fantastic process to follow when deciding what a blog or other content hub should be about, and what topics they should cover. Clients often have questions about what they should do to make a blog more interesting. While, "you need to write more shareable content" is almost always true, it's not very useful advice. Instead, can we get people to think, "where would I advertise this content?" If you're coming up empty, you might need to think about writing a less narrow blog.

A blog about birds is probably too narrow. Can we talk about all pets and animals, including more popular topics like dogs and cats? If you were to actually run this campaign, you'd come out with some decent evidence that a given piece of content was or was not working with your target audience. 

Who Should Run Content Ads?

We really can't be sure whether my paid advertising campaign was more effective at building links than a similarly-priced pure outreach campaign. I also can't say (yet) whether my paid campaign was abnormally effective or ineffective for the topic. What we do know is this:

  1. Using ads to build links can work. I built links from about 140 legitimate linking root domains for under $300, plus a handful with the StumbleUpon ads. That doesn't include the many hours invested in creating the content, of course.
  2. Targeted ads can help you test content and audience interest in a unique way. You can make sure you're on track before investing in months worth of content that doesn't interest your audience.
  3. Ads take very little time. For campaigns where time is more limiting than budget, or where outreach targets are limited, there may be value in selective content advertising.

Most companies aren't writing just for the joy of writing. Thinking about outreach and content generation from a cost/benefit perspective creates goal and return-oriented content generation. In other words, paying for visitors helps you create more shareable content. Companies who already have shareable content may be able to use paid advertising to increase the profit it generates, both directly and indirectly.

Get blog posts via email