A Practical Guide to Content and Its Metrics

A small disclaimer:

Before you start reading, I want to say that I am not an analytics expert per se, but a strategic SEO and digital marketing consultant. On the other hand, in my daily work of auditing and designing holistic digital marketing strategies, I deal a lot with Analytics in order to understand my clients' gaps and opportunities.

For that reason, what you are going to read isn't an "ultimate guide," but instead my personal and practical guide to content and its metrics, filled with links to useful resources that helped me solving the big contents' metric mystery. I happily expect to see your ideas in the comments.

The difference between content and formats

One of the hardest things to measure is content effectiveness, mostly because there exists great confusion about its changing nature and purpose. One common problem is thinking of "content" and "formats" as synonyms, which leads to frustration and, with the wrong scaling processes present, may also lead to Google disasters.

What is the difference between content and formats?

1. Content is any message a brand/person delivers to an audience;
2. Formats are the specific ways a brand/person can deliver that message (e.g. data visualizations, written content, images/photos, video, etc.).

Just to be clear: We engage and eventually share the ideas and emotions that content represents, not its formats. Formats are just the clothing we choose for our content, and keeping the fashion metaphor, some ways of dressing are better than others for making a message more explicit.

Strategy, as in everything in marketing, also plays a very important role when it comes to content.

It is during the strategic phase that we attempt to understand (both thanks to our own site analysis and competitive analysis of others' sites) if our content is responding to our audience's interests and needs, and also to understand what metrics we must choose in order to assess its success or failure.

Paraphrasing an old Pirelli commercial tagline: Content without strategy is nothing.

Strategy: Starting with why/how/what

When we are building a content strategy, we should ask ourselves (and our clients and CMOs) these classic questions:

1. Why does the brand exist?
2. How does the brand solidify its "why?"
3. What specific tactics will the brand use for successfully developing the "how?"

Only when we have those answers can we understand the goals of our content, what metrics to consider, and how to calculate them.

Let use an example every Mozzer can understand.

Why does Moz exist?

The answer is in its tagline:

• Inbound marketing is complicated. Moz’s software makes it easy.

How does Moz solidify its "why?"

• Moz produces a series of tools, which help marketers in auditing, monitoring and taking insightful decisions about their web marketing projects.
• Moreover, Moz creates and publishes content, which aims to educate marketers to do their jobs better.

If you notice, we can already pick out a couple of generic goals here:

1. Leads > subscriptions;
2. Awareness (that may ultimately drive leads).

What specific tactics does Moz use for successfully achieving its main goals?

Considering the nature of the two main goals we clarified above, we can find content tactics covering all the areas of the so-called content matrix.

Some classic content matrix models are the ones developed by Distilled (in the image above) and Smart Insights and First 10, but it is a good idea to develop your own based on the insights you may have about your specific industry niche.

The things Moz does are many, so I am presenting an incomplete list here.

In the "Purchase" side and with conversion and persuasion as end goals:

• Home page and "Products" section of Moz.com (we can define them as "organic landing pages");
• Content about tools
  • Free tools;
  • Pro tools (which are substantially free for a 30-day trial period).
• CPC landing pages;
• Price page with testimonials;
• "About" section;
• Events sponsorship.

In the "Awareness" side and with educational and entertainment (or pure engagement) purposes:

• The blogs (both the main blog and UGC);
• The "Learn and Connect" section, which includes the Q&A;
• Guides;
• Games (The SEO Expert Quiz can surely be considered a game);
• Webinars;
• Social media publishing;
• Email marketing
• Live events (MozCon and LocalUp, but also the events where Moz Staff is present with one or more speakers).

Once we have the content inventory of our web site, we can relatively easily identify the specific goals for the different pieces of content, and of the single type of content we own and will create.

I will usually not consider content like tools, sponsorship, or live events, because even though content surely plays a role in their goals' achievement, there are also other factors like user satisfaction and serendipity involved which are not directly related to content itself or cannot be easily measured.

Measuring landing/conversion pages' content

This may be the easier kind of content to measure, because it is deeply related to the more general measures of leads and conversions, and it is also strongly related to everything CRO.

We can measure the effectiveness of our landing/conversion pages' content easily with Google Analytics, especially if we remember to implement content grouping (here's the official Google guide) and follow the suggestions Jeff Sauer offered in this post on Moz.

We can find another great resource and practical suggestions in this older (but still valid) post by Justin Cutroni: How to Use Google Analytics Content Grouping: 4 Business Examples. The example Justin offers about Patagonia.com is particularly interesting, because it is explicitly about product pages.

On the other hand, we should always remember that the default conversion rate metric should not be taken as the only metric to incorporate into decision-making; the same is true when it comes to content performance and optimization. In fact, as Dan Barker said once, the better we segment our analysis the better we can understand the performance of our money pages, give a better meaning to the conversion rate value and, therefore, correct and improve our sales and leads.

Good examples of segmentation are:

• Conversions per returning visitor vs new visitor;
• Conversions per type of visitor based on demographic data;
• Conversions per channel/device.

These segmented metrics are fundamental for developing A/B tests with our content.

Here are some examples of A/B tests for landing/conversion pages' content:

• Title tags and meta description A/B tests (yes, title tags and meta descriptions are content too, and they have a fundamental role in CTR and "first impressions");
• Prominent presence of testimonials vs. a more discreet one;
• Tone of voice used in the product description (copywriting experiment);
• Product slideshow vs. video.

Here are a few additional sources about CRO and content, surely better than me for inspiring you in this specific field:

• Unbounce;
• Optimizely;
• ConversionXL.

Measuring on-site "editorial" content

Here is where things start getting a little more complicated.

Blog posts, guides, white papers, and similar content usually do not have a conversion/lead nature, at leastnot directly. Usually their goals are more intangible ones, such as creating awareness, likability, trust, and authority.

In other cases, then, this kind of content also serves the objective of creating and maintaining an active community, as it does in the case of Moz. I tend to consider this a subset, though, because in many niches creating a community is not a top priority. Or, even if it is, it does not offer a reliable flux of "signals" so as to appropriately measure the effectiveness of our content because of pure lack of statistical evidence.

A good starting place is measuring the so-called consumption metrics.

Again, the ideal is to implement content grouping in Google Analytics (see the video above), because that way we can segment every different kind of editorial content.

For instance, if we have a blog, not only we can create a group for it, but we can also create

• As many groups as there are categories and tags on our blog;
• groups by average length of the posts;
• groups per the kind of prominent formats used (video posts like Moz's Whiteboard Fridays, infographics, long-form, etc.).

This are just three examples; think about your own measuring needs and the nature of your content, and you will come out with other ideas for content groupings.

The following are basic metrics that you'll need to consider when measuring your editorial content:

1. Pageviews / Unique Pageviews
2. Pages / Session
3. Time on Page

The ideal is to analyze these metrics at least with these secondary levels:

• Medium / Sources, so you can understand what channel contributed the most to your content visibility. Remember, though, that dark search/social is a reality that can screw up your metrics (check out Marshall Simmonds' deck from MozCon 2015);
• User Type, so to see what percent of the Pageviews is due to returning visitors (a good indicator of the level of trust and authority our content has) and new ones (which indicates the ability our content has to attract new potentially long-lasting readers);
• Mobile, which is useful in understanding the environments in which our users mostly interact with our content, and how we have to optimize its experience depending on the device used, hence helping making our content more memorable.

You surely can have fun also analyzing your content's performance by segmenting them per demographic indicators. For instance, it may be interesting to see what affinity categories of your readers there are, depending on the categorization used in your blog and that you have replicated in your content grouping. This, in fact, can help us in better understanding the personas composing our audience, and so refining the targeting of our content.

As you can see, I did not mention bounce rate as a metric to consider, and there is a reason for that: Bounce rate is tricky, and its misinterpretation can lead to bad decisions.

Instead of bounce rate, when it comes to editorial content (and blog posts in particular), I prefer to consider scroll completion, a metric we can retrieve using Tag Manager (see this post by Optimize Smart).

Finally, especially if you also grouped content for outstanding format used (video, embedded SlideShare, etc.), you will need to retrieve users' interactions through Tag Manager. However, if you really want to dig into the analysis of how that content is consumed by users, you will need to export your Analytics data and then combine it with data from external sources, like YouTube Analytics, SlideShare Analytics, etc.

The more we share, the more we have. This is also true in Marketing.

Consumption metrics, though, are not enough in order to understand the performance of your content, especially if you strongly rely on a community and one of the content objectives is creating and growing a community around your brand.

I am talking of the so-called sharing metrics:

1. Social shares (Likes, Tweets, Pins, etc.);
2. Inbound links;
3. Un-linked mentions
4. Email forwarding.

All this can be tracked and measured (e.g.: social shares, mentions on web sites or on social).

I usually add comments into these Metrics, because of the social nature comments have. Again, thanks to Tag Manager, you can easily tag when someone clicks on the "add comment" button.

A final metric we should always consider is the page value. As Google itself explains in that Help Page:

Page value is a measure of influence. It’s a single number that can help you better understand which pages on your site drive conversions and revenue. Pages with a high Page Value are more influential than pages with a low Page Value [Page Value is also shown for groups of content].

The combined analysis of consumption and social metrics can offer us a very granular understanding of how our content is performing, therefore how to optimize our strategy and/or how to start conducting A/B tests.

On the other hand, such a granular vision is not the ideal for reporting, especially if we have to report to a board of directors and not to our in-house or in-agency counterpart.

In that case being able to resume all these metrics (or the most relevant ones) in just one metric is very useful.

How to do it? My suggestion is to follow (and adapt to your own needs) the methodology used by the Moz editorial team and described in this post by Trevor Klein.

What about the ROI of editorial content? Don't give up; I'll talk about it below.

Measuring the ROI of content marketing and content-based link building campaigns

Theoretically measuring the ROI of something is relatively easy:

(Return - Investment) / Investment = ROI.

However the difficulty is not in that formula itself, but in the values used in that formula.

How to calculate the investment value?

Usually we have a given budget assigned for our content marketing and/or content-based campaigns. If that is the case, perfect! We have a figure to use for the investment value.

A complete different situation is when we must present a budget proposal and/or assign part of the budget to each campaign in a balanced and considered way.

In this post by Caroline Gilbert for Siege Media you can find great suggestions about how to calculate a content marketing budget, but I would like to present mine, too, which is based on competitive analysis.

Here's what I do:

1. Identify the distinct competitors which created content related to what we will target with our campaign. I rely on both SERP analysis (i.e.: using the Keyword Difficulty Tool by Moz) and information we can retrieve with a "keyword search" on Buzzsumo.
2. Retrieve all meaningful content metrics:
   • Links (another reason why I use the Keyword Difficulty Tool);
   • Social shares per kind of social network (these are available from BuzzSumo). Remember that some of these social shares can be tallied by sponsored content (check this Social Media Explorer post about how to do Facebook competitive analysis).
   • Estimated traffic to the content's URL (data retrieved via SimilarWeb).
3. Assign a monetary value to the metrics retrieved.
4. Calculate the competitors' potential investment value.
5. Calculate the median investment value of all the competitors.
6. Consider the delta between what the client/company invested in content marketing (or link building, if it is moving from classic old link building to modern link earning) before, as well as the median investment value of the competitors.
7. Calculate and propose the content marketing / content-based campaign's value in a range which goes from "minimum viable budget" to "ideal."

Reality teaches us that the proposed investment is not the same than the real investment, but at least we then have some data for proposing it and not just a gut feeling. However, we must be prepared to work with budgets that are more on the "minimum viable" side than on the ideal one.

How to calculate revenue?

You can find a good number of ROI calculators, but I particularly like the Fractl one, because it is very easy to understand and use.

Their general philosophy is to calculate ROI in terms of how much traffic, links, and social shares the content itself has generated organically, hence how much it helped saving in paid promotion.

If you look at it, it reminds the methodology I described above (points 1 to 7).

However, when it comes to social shares, you should avoid the classic mistake of considering only the social shares directly generated by the page your content has been published.

For instance, let's take the Idioms of the World campaigns Verve Search did for HotelClub.com and which won the European Search Awards.

If we we look only at its own social share metrics, we will have just a partial picture:

Instead, if we see what are the social shares metrics of the pages that linked and talked about it, we will have the complete picture.

We can use (again) BuzzSumo for retrieving this data (also using its Content Analysis feature), or using URL Profiler.

As you can imagine, you can calculate the ROI of your editorial content using the same methodology.

Obviously the Fractl ROI calculator is far from being perfect, as it does not consider the offline repercussion a content campaign may have (the Idioms of the World campaign was organically published in a outstanding placement on The Guardian's paper version, for instance), but it is a solid base for crafting your own ROI calculation.

Conclusions

So, we have arrived at the end of this personal guide about content and its metrics.

Remember these important things:

• Don't be data driven, be data informed;
• Think strategically, act tactically;
• Content's metrics vary depending on the goals of content itself.