At NewsCred, one of our favorite things to say is that content marketing is not about being good at creating content – it’s being good at driving business results through content.
The definitive way of proving content marketing ROI is by using an analytics platform to measure hard conversions like leads generated or purchases completed. (You can read how to do so here.)
But if we know that consumers read an average of 11 pieces of content before making a purchasing decision, we need a way to track the efficacy of the content at the beginning of their journeys with softer metrics like engagement rate, time on site, and pageviews per visitor.
Below you can see NewsCred’s own model for aligning KPIs to the buyer journey. If we’re not tracking KPIs in the “Attract” column properly, we can’t drive users forward to capture and finally convert them.
There are plenty of analytics platforms to choose from, and most track similar metrics in much of the same ways. The issue is that these platforms were not primarily built to measure content and its impact on business goals – they were developed for e-commerce.
As a result, the metrics may not mean what we think they do. As content marketers, we need to truly understand what these metrics measure before we optimize towards them to meet our goals.
Let’s take bounce rate, for example.
How Bounce Rate is Calculated
How many marketers are held responsible for having low bounce rates on their content pages, without clearly understanding what bounce rate even means?
In tools like Google Analytics (GA), WebTrends, Yahoo! Web Analytics, CoreMetrics (and, to a certain extent, Omniture, although you have to set your own bounce rate), bounce rate is measured by single page visits divided by total entrants. This means the percentage of people who only looked at one page on your website, took no action, and then left.
Google Analytics guru, Avinash Kaushik, calls bounce rate a “puke rate” – meaning someone saw your page and was so uninterested that their immediate reaction was to leave as quickly as possible. Kaushik argues that bounce rate is “the sexiest metric ever” and that it’s the most important metric to use to understand a consumer’s experience of your site.
I would argue that bounce rate as it is typically understood and measured actually has no place in determining how effective your content marketing is.
Bounce rate is a great way to measure a homepage or a product page, in which a failure to click on anything is a failure of the page. When it comes to content marketing, however, a reader can be fully engaged without taking an action. (Obviously, we want to encourage action and we don’t want our content to be a dead end, but that is a blog post for another time.)
If a reader is being sent directly to an article page from social, email, or paid, they might only interact with the one piece of content they came to read. Depending on the design of your site, there might not be strong navigation options to keep them on the page, so they might bounce back to their Twitter feed, or the publisher they were reading when they discovered your content, even if they loved the article.
So while the below video shows me engaging fully with this article and demonstrating real interest in Red Bull’s Rising Force, because I didn’t click on anything else and in all likelihood went back to my Twitter feed, this is considered a bounce.
Yes, Red Bull missed an opportunity to follow up with me and bring me further towards an action. But they’ve also missed out on valuable metrics that will let them know how interesting their content is.
There are a lot of resources online to help you lower your bounce rate. Many of these focus on UI/UX changes that force readers to click to continue reading the content they want. How many times have you had to click “next page” to view what is really the “next paragraph” while reading an article? This serves to drive down bounce rates by forcing you to take an action to continue reading the same article. While we want to encourage readers to take action such as clicking to read more articles, opting in to receive email updates from us, or following us on social, we do not want to architect a poor user experience in the process.
Content marketing is a value exchange, so we can’t necessarily expect to earn these high value actions during someone’s first interaction with our content.
But it’s important know if someone is indeed “puking” and leaving our content at first sight, or reading the whole thing. Knowing which content is being read in its entirety should help inform our content creation strategy.
How to Effectively Measure Content Engagement
There are a few things you can do to alleviate the limitations bounce rate has for content marketers:
- Set up an adjusted bounce rate. This means you can customize your analytics platform to trigger an event after a certain amount of time to indicate someone is fully engaged. The industry standard for this is 30 seconds, but the best way to determine the right number for your site is to read a few articles and see how long each takes. For many sites, this will be closer to 15 or 20 seconds. Here is an article from Google that explains how to set up an adjusted bounce rate for GA, which is as simple as adding a snippet of code. (Note that you need to do this on every article page, or in your Google Tag Manager to implement the changes across your site.)
- Use NewsCred’s own CMC analytics tool, which uses scroll depth and mouse movement to measure engaged time and engagement rate instead of bounce rate. That means that if someone is reading through your article, we’re recording their engagement and measuring your content more effectively. These metrics were specifically designed to measure content, as opposed to applying a one-size-fits-all metric that was crafted to measure e-commerce.
- In Omniture, you can set up a campaign to specifically track “engagement,” which is a metric that blends pageviews, time on site, and capture score. This is not a perfect fix, however, which we’ll discuss more below.
How Time on Site is Calculated
At this point in this article, you might be thinking to yourself: “I already knew this information about bounce rate. I’m no patsy to big analytics. I use time on site to measure the efficacy of my content marketing, not bounce rate!”
In theory, time on site seems like the perfect fix to our bounce rate problem. Instead of measuring our content based on the next action, we should just measure how long someone spent on the site to tell if they’re engaged with our content, right?
In my Red Bull example video above, we’d be able to see that I spent 45 seconds on site, and therefore fully engaged with the article.
Except that’s actually not how time on site is calculated at all.
Time on site is much like Indiana Jones’ time in temple in this famous scene from “Raiders of the Lost Ark.” Time doesn’t actually start ticking for our hero until he takes a definitive action by swapping out the weighted bag for the idol.
Time on page or time on site is calculated by measuring the seconds between two actions taken directly on a site.
For example: Time on Page 1 = (Time Stamp of Page 2) – (Time Stamp of Page 1). So if there is no Page 2, there is no time on page for Page 1.
This means if a user bounces, pukes, doesn’t initiate a Page 2, or comes in from social and leaves after Page 2, there’s no time on site for the visit at all.
Even if you aren’t using bounce rate as a KPI and even if you’re not optimizing towards it, having a high bounce rate means your time on site is completely unreliable. If your site has a bounce rate of 85%, you are only measuring time on site for arguably already the most engaged 15% of your traffic. This offers you no way to measure how interesting or effective your content is, or to use these insights to inform your content creation and distribution strategy.
Another issue with time on site is the actual time of day. Many analytics platforms, including GA, automatically stop recording time on site at midnight depending on how you set your timezone. This might not seem like an issue for your target audience, but if you’re selling herbal supplements to insomniacs or regular supplements to Australians, you might want to know what they’re up to in the middle of the night.
Patagonia has a beautiful corporate social responsibility hub called The Cleanest Line, filled with long-form, image-heavy content about the the outdoors, activism, and sustainability.
Patagonia drives traffic directly to article pages from Twitter and Facebook, and does not feature heavy CTAs to subscribe to newsletters or donate money. The goals of this site seem to be focused on building awareness of Patagonia’s social mission, increasing brand equity, and nurturing a community of activists through social sharing and commenting. They employ many best practices for content hubs including social sharing icons, floating right-rail article recommendations, and a related articles widget at the bottom of the page. No one can accuse their content of being a dead end. There are plenty of user-friendly navigation options on the page to deter users from bouncing right away.
Below you can see me coming across an interesting article on Facebook. (Note: I’m arriving from Facebook and not clicking into the article from a Patagonia page, so this doesn’t count as a click on the site.) Then, I read the entire thing (quickly, so as not to bore you, reader, who is watching me read it). When I finish reading the article, I am so engrossed in the beautiful and engaging content that I read two more articles in their entirety.
In total, I spent 01:08 on The Cleanest Line, consuming three full articles. The breakdown is as follows:
- I land on “When I First Fell for the Arctic” and read it from 00:04-00:25. (Remember, there is no Time Stamp for this article since I clicked to it from Facebook.)
- I then click on “Do What You Love to Protect What You Love” at 00:26 (Time Stamp 1) and read until 00:45.
- And finally, I click on “A Letter Writing Party to Protect Bear Ears” (yes, bear ears) at 00:46 (Time Stamp 2) and read until 01:07, before navigating back to Facebook.
My total time on site for this visit will only be calculated as 00:20 (i.e. Time Stamp 2 – Time Stamp 1) because only the middle article is sandwiched between events, which in this case, are clicks on the hub. Patagonia will have no idea how I first fell for the Arctic, or how invested I am now in the increasing threats facing bears and their ears.
How to More Effectively Measure Time on Site
There are a few things you can do to better understand the real time your readers are spending on your site, which are very similar to the mitigators of bounce rate issues:
- Set up a custom event in your Google Tag Manager to trigger when someone starts scrolling. This will start the clock on your time on site. This still doesn’t solve for your last page, but will capture your first page.
- For your exit page, consider a non-invasive lightbox to capture email addresses, NPS, or even thumbs up or thumbs down to ensure you’re capturing an event as a user leaves (and therefore getting credit for that last page).
- Use NewsCred’s CMC analytics tool, which tracks scroll depth and mouse movement to measure engaged time and engagement rate – thus ensuring people are actually consuming the content.
- Again, in Omniture, you can set up a campaign to specifically track “engagement,” which is a metric that blends pageviews, time on site, and capture score. Time on site remains the same issue, but coupled with the capture score, can be more valuable.
- Ensure your site implementation is configured to capture data that is meaningful for you. For example, if you have an “expand menu” button, a “click to watch a video” CTA, or “click to share on social” icons, make sure these are configured as events in your analytics.
Web analytics can feel like a black box, but understanding your data correctly will allow you to track the right metrics.
As Albert Einstein says, you can’t judge a fish by its ability to climb a tree. Similarly, we shouldn’t be holding our content to unrealistic standards designed with the advent of the Internet to measure product landing pages.
By adjusting our metrics to accurately measure consumer engagement, you’ll have data that will better enable you to build a compelling case for content marketing, as well as continue creating content that is delighting your readers and driving them towards your business goals.
Esti Frischling is a Senior Partner, Customer Success, at NewsCred.