SEO content measurement: you’re doing it wrong!
It’s fair to say that the arena of content marketing has evolved beyond the days of guest posts and surveys: our demand to consume novel, insightful and ‘lough out loud’ content has raised the marketing bar – and brands have responded with all manner of rich, interactive media.
What’s interesting, however, is that we’re still seemingly stuck in the Dark Ages when it comes to measuring the success of that content.
Digital evaluation hasn’t really embraced the fact that significant portions of web traffic are now driven to mobile versions of websites – with a correspondent shift in user behaviour.
Google got the memo a while back now, paying closer attention to ‘user signals’ as well as backlink profiles. This means that a significant part of its mobile ranking algorithm will look at how well or poorly a user interacts with a website, on both desktop and mobile.
Conventional wisdom would see a marketing manager go into Google Analytics and muse over metrics such a bounce rate, dwell time and pages visited per session. Whilst all of these metrics are important, poor optimisation choices can result from the way in which we interpret them in this multi-device world.
Let’s take each of these metrics in turn and examine how we might analyse them in a more rounded way:
A standard recommendation from an SEO audit is to reduce bounce rates across the site: put simply, this means to increase the time spent on a page by reducing those instances where a user visits and leaves immediately (thereby ‘bouncing’ off the page).
However, blanket recommendations can be a dangerous thing. Category pages on an e-commerce site, for example, would typically exhibit high bounce rates as users progress swiftly to product pages, where they spend time considering, comparing – and hopefully buying. To increase the content on the category page is likely to be counterintuitive, by protracting the conversion pathway, distracting the user and, ultimately, losing out on sales.
This is even more relevant to the mobile journey, where swift navigation through pages is likely to be indicative of readily accessible key content. In turn, this makes a high bounce rate a good thing.
Dwell time presents a parallel discussion to that of bounce rate: we’re just not meant to labour over certain web pages. Think of a contact page or a newsletter sign-up page: when executed effectively, these areas of a site will see us enter the necessaries within seconds, before continuing on our way.
This tendency is accelerated further by the data that we store in our browser, which populates capture fields in one click.
In the mobile scenario, rapid page load time with data auto-population will see users convert in exactly the way that the webmaster desires, at lightening speed. A resounding success; but not if you apply the wrong view of the metrics!
The lesson here, then, is that not all pages are created equal, so it’s necessary to look at the dwell time metric on a page-by-page basis, rather than on a site-wide basis.
Page Views Per Session
This metric is a bit trickier to dissect, since the expectation here is that a page of quality content will prompt the user to visit other web pages. There are several scenarios that buck this trend:
1. Returning to our e-commerce example, the category page ‘doorway’ to product pages will almost inevitably yield multiple page views per session – but this is a virtue of its design rather than great quality content.
2. A poor quality page can prompt the user to search a little further for the information they want, before abandoning the site altogether. This journey is clearly a poor show for the website, but if the metrics are applied erroneously it can turn into the fallacy of a success.
3. A content page or blog can be of such superb quality, with all of its important information fully integrated, leading to a social share but no further site exploration. There’s perhaps an argument here for the author of this page to consider internal linking as part of whetting the audience’s appetite, but at its core, this is a hugely successful content page which will be marred by an inflexible evaluation process.
In the mobile world, it’s still a sad fact that 4G isn’t the same as a solid broadband connection, which can impair page load time. Users adapt to this by keeping searches concise and website navigation to a minimum, often with the intent of returning later on their laptop or desktop.
In this case, the pages viewed per session may take a hit, and this analysis will not consider the overall pathway to conversion; one where the initial mobile visit results in a transaction.
These caveats demonstrate the need for a truly bespoke view of each webpage when evaluating the success of a given website. Which leads me nicely onto the next point: what’s the best way to evaluate web performance, whether on a desktop site or a mobile site?
I’d start with your HTML sitemap and plot the purpose of each page. You can then define what ‘good’ looks like for each metric that you plan to measure and evaluate. It’s a good idea to define the user intent of each page, rather than your own purpose as the marketing manager.
This maps directly onto the user-centric approach of search engines such a Google, whose mission is to serve relevant and accessible content to users across the device spectrum. If your content places the user first, you’ll reap the rewards of good search engine visibility.
There’s no doubt that analysing a website on a page-by-page basis won’t write your monthly or quarterly performance review report any faster – but it will reveal actionable insights, which will mean your next report is far more likely to demonstrate uplifts rather than stagnation or regression.
All that’s left now is to educate the C-Suite on the necessity to delve in to the detail, rather than producing the old ‘summary’ slide for the board meeting…
Want more information? Contact us here