The promise of content marketing is the astounding ability to provide value to your audience, help them find great advice and entertain them. In order to do so successfully, we need to adjust and react to insights provided by data. But in order to make those adjustments, we need to ensure that we're actually measuring the right indicators and not creating strategy based on arbitrary numbers.
Looking at site engagement in aggregate
The most common mistake in measurement is comparing traffic sources to a site by primarily looking at the Google Analytics' dashboard high level comparison table of traffic sources, which shows the number of page views, average session length, bounce rates, and page views per session. Your site has very different types of pages: a homepage, landing pages for campaigns, content pages, e-commerce pages, etc. The nature of engagement with each type of page is very different. A product page or an e-commerce section has a much higher "time on site" than a blog post. So, if one traffic source drives traffic to the e-commerce section and another traffic source drives traffic to the blog post, the results of that engagement will look very different if we simply compare numbers like time -- even if they are achieving their optimal result.
Solution: Resist the urge to look solely at the default high level view of your site. In order to understand your performance, you need to compare the traffic sources' engagement parameters for each section or each page type in isolation. This practice enables you to create a benchmark for each area of your site and draw the right conclusions to determine which traffic sources drive the best engagement result. One more piece of advice: When comparing traffic sources, remove all pages that score less than 10 page views. This will help you focus your comparison on those areas of the site that actually achieved minimum engagement, which means focusing on the pages that actually matter.
Comparing click-through rates between platforms
Another fatal mistake that marketers fall victim to is assessing the perceived effectiveness of different paid media platforms by comparing the ratio of impressions to clicks between them (aka CTR or click through rates). Why does this yield inaccurate insight? Because this is not an apples to apples measurement. Can you really compare the CTR on a YouTube video advertisement to a paid search campaign on Google to a paid content discovery campaign on Outbrain? The minimum requirement for comparing platforms is to make sure they function similarly and that you are promoting the same kind of advertising or content on both. So should you be able to compare CTR between BING and Google for the same adwords copy? Well, that depends. You need to make sure both Microsoft and Google count "a page view" in the same way before you can compare that data. One search engine may count a page view as "a user reached the page" and one may measure a page view as "a user actually saw the ad." In this case, the definition and result of CTR would be very different.
Solution: Ask each paid media platform to provide a benchmark CTR for your category on their platform. Then judge how you fare versus the competition on each specific platform. Compare between platforms that drive traffic to the same area of the site based on engagement or conversion results relative to the cost of each platform.
Looking at bounce rates in isolation
Bounce rates are probably the least understood KPI. Many marketers think it means "someone came to my site and left immediately." This is an oversimplification. What it actually means is the percentage of people that visited one page and left. Is that good or bad? What if I have a blog with a loyal fan base that reads every post I'm uploading? My bounce rates would be through the roof, because this audience would come daily to my blog, read the latest post and leave. But using the criteria of bounce rates, my efforts would be considered failures.
Solution: Bounce rates should be considered relative to time on site. You can change the Google Analytics default of bounce definition and make it "visited only one page AND stayed on the page they landed on less than 20 seconds." This will help separate between a "good bounce," which is actually an indication of high engagement, and a "bad bounce," which is fly-by traffic that comes to your site and leaves quickly.
Measuring for conversions and not engagement
Too often, I hear marketers flirting with implementing content marketing strategies who try to impose the same tactics of direct response measurement on their content marketing strategy as they use in their direct response efforts. This is a mistake. With time, your team will focus on the bottom of the funnel and on the kind of content that drives conversions. When related to content marketing, it's a bit like going on a first date and asking them to marry you. It's probably better to go out a few times, get to know the other person before you get down on one knee, no? Good chance you'll improve the odds of getting a "yes."
Solution: Different content is warranted for each conversion's purposes (getting married) versus for audiences' delight (dating). Think about what your audience needs (as opposed to what you need), and how you can help them. Then measure whether or not it's working: are they discovering your content (visit), do they enjoy it (bounce rate+time on site, PV per session), do they love it (become fans, follow, favorite) or engage with it (share it with their friends, "like," comment)? Optimize for "delight" first, and only then, when you have enough traffic and engagement with your content, start optimizing some of that engagement to conversions.
Defining social engagement goals only
Don't get me wrong: it's vital to measure all social signals like shares, "likes," etc. It's a great way to measure success IF your content was really made for social. Studies show we often share very different stories than those we actually found most interesting. For example, if I just learned I have a medical condition there's a high chance I'll be deeply engaged with content related to that subject, but a very low chance I'll share it with anyone on my social networks.
Solution: If social sharing is important for your business goals, make sure to create content that is "made for social." It would be shorter, more visual, and have a catchy title. Longer form "deep" content tends to get shared less, so measure traditional engagement for deep content.
We're still at the nascent stages of content discovery being wielded most effectively, and there are no hard and fast rules. Embarking on the path before there's a given formula can be stressful, but it's also exciting. As we point out throughout the piece, a great content discovery strategy requires its executors to think critically, accept that there are no givens or hard rules, and constantly tinker like a true entrepreneur.
Gilad de Vries is senior vice president at Outbrain.
On Twitter? Follow iMedia Connection at @iMediaTweet.