Content Marketing Analytics: 11 Ways Data Can Inform Your Content Strategy

We all want to be data-driven marketers, but sometimes content marketers are left out of the loop and expected to run solely on creativity.

Content marketing is a creative endeavor – just as other types or marketing are – but data and analytics can inform content marketing just like they can paid, social, or affiliate.

Most current guides on analytics aren’t focused at content marketers. Rather, the examples and explanations are geared towards optimizers, paid acquisition marketers, and analysts.

So, since I’ve done a ton of content marketing as well as a ton of analytics work at both CXL and at HubSpot (as well as in my own nerdy free time), I thought it would be helpful to splice the two.

This article will cover some super tactical things content marketers can do to bring data into their daily work – and ultimately understanding how their content promotion and creation efforts are paying off.

Note: you should have a solid grasp of Google Analytics and Google Tag Manager. Just the basics will do (how data is collected, what basic reports means, etc.). If you don’t have any understanding of those, these two courses are great:

Asking Good Questions: The Core of Content Marketing Analytics

In my opinion, every content marketer should be able to answer the following business questions using analytics:

  • How effective are our current content marketing efforts according to analytics?
  • What are the content marketing opportunities we’re missing as shown by analytics?
  • How can we optimize our content marketing efforts in the near term?
  • At what points in the funnel are we dropping users? Can we do anything to plug these holes (or work with a team who can)?
  • Do we see trends or insights that tell us which types of content are most effective at attracting high value leads/customers?
  • What can analytics show me about new content marketing ideas and campaigns?

Generally, data should be able to guide your decisions as a content marketer just as it should if you’re a paid marketer or SEO.

There’s no reason for content to be distanced from business metrics, and using data does not negate your ability to use creativity.

This guide will cover what I think are the absolutely critical analytical functions a content marketer should be able to execute.

They span the qualitative and the quantitative, and I’ll even dip into specific reports in Google Analytics and specific survey questions to help you implement some of this. Here’s what we’ll cover:

These aren’t the only questions you’ll want to answer, but the span of these use cases for data should at least get you started and provide inspiration for further learning.

Each section may brush over the particular process fairly quickly, but I’ll link to resources to understand it more fully. Let’s get into it!

Content Marketing Analytics: The Quantitative Stuff

This section is mostly going to cover Google Analytics, but whatever digital analytics tool you use will suffice – Adobe Analytics, Woopra, Piwik, etc.

I have the most experience with Google Analytics, and it’s free and super common, so I’ll write with examples from GA.

Does this blog post actually drive conversions?

If you don’t have goals set up in Google Analytics, do that now.

If you haven’t set up goals, you can’t track whether your efforts are actually resulting in what matters for the business.

When you’ve done that, you can view which blog posts are actually leading to goal conversions. You can do so using advanced segments.

So, what you’ll do is click “+ Add Segment” and then click the red “+ New Segment” button. Find “Sequences” under the advanced menu. Then set up a sequence with the first step as “Page” contains “[your blog post URL,” and use “is followed by” and create a second step that includes sessions with goal conversions.

You could use “transactions” or a specific goal for this, but I just used “Goal completions > 0” here…

A common mistake here is that people will use “Conditions,” but since what you care about is the sequence of events, you need to use “Sequences.”

If you’re doing this at scale (for dozens or hundreds of posts), it might be better to do it in a tool like R or even in a Data Studio report so you don’t need to constantly query GA for the reports you need.

You can also find conversions and conversion rate by landing page with customer reports. This blog post does a really good job explain how.

Which blog posts have high traffic but low conversions (or engagement)?

A good way to find CRO opportunities is to locate pages with high traffic and low conversions or engagement.

There are a few ways to do this, but my favorite is to use the comparison feature on Google Analytics.

Just go to Behavior > Site Content > Landing Pages and use the comparison feature on the right hand side of the screen. Then select Goal Conversion Rate (or Bounce Rate) from the drop down menu:

This shows you all pages, but it’s likely that you’ll want to get more granular and focus solely on your blog.

It’s possible you have a View specifically for this, but if not, you’ll need to search your category indicator (in my example below “drinkware” but in yours perhaps “blog”), and you’ll see only those pages and can nail down which ones need work:

The pages with high traffic and lower than average conversion rates or bounce rates are the biggest potential impact areas.

It’s these pages that you can optimize, or send to your web strategy or conversion optimization team for prioritization.

There’s a lot of room for improvement on these pages, and because they’re higher traffic, the impact is larger than the marginal increases in value you’d see with lower traffic pages.

How to find content gaps with user search data

You should know how to do keyword research.

That’s a given.

But after you’ve burned through the basics and the high volume keywords, it can get difficult to find ideas for what to write about. So, why not let your users tell you? Not directly, of course, but through their site search behavior.

If your GA is set up correctly, all of your site search data will be available in Behavior > Site Search > Search Terms.

This will give you a big list of everything users are searching for on your site, along with corresponding metrics like search exits, which is when someone searches and then leaves your site.

If your site has a small list of search terms, you can just look through all of these and find interesting terms – things you may not have even known people were looking for.

A better way, however, would be to do a time comparison and find trends.

This will give you the most popular terms, but even better than that is a little known feature where you can sort by “absolute change.”

This will let you see the biggest changes over time in your search data:

Anything interesting there? You can create pages for it.

For instance, when I was at CXL we found on our CXL Institute site an increase in search terms for “guarantee,” “about us,” and a variety of content topics that helped us to create internal pages for them.

Seriously good information to have as a marketer!

You should, of course, complement this with off-site and general keyword research using tools like Ahrefs and Keywords Everywhere.

This is just what people are searching for in your site-search bar, so it’s a narrow reflection of people’s intentions. It excludes all the people who don’t know who you are, haven’t landed on your site, or just don’t use site search to find their answers.

But this method does tell you what people are looking for, however small the sample. Can you find a way to answer their queries in the form of content?

How to Find SEO Opportunities

Content marketers need to be able to audit their current SEO efforts and results, and they also need to be able to plan out content campaigns based on search volume and intent.

So, we can basically split this step up between two areas:

  • How to plan SEO-driven content marketing.
  • How to audit your current approach and optimize and improve it.

To the first point, it’s largely a strategic process that starts with you having a discussion about your top business goals and conversion paths, and how you can work backwards to find appropriate keywords with enough volume and intent to help those goals.

For example, if you wanted to acquire users for a term like “free knowledge base software,” then you know you’d need to create a product page that captures some level of volume and intent there.

That’s your starting point, and then you can work backwards and think, “what kind of related things would people be searching for?”

That’s where your traditional keyword research process takes form, and this post isn’t going to be long enough to cover the complete process. There are great guides out there you can follow:

In our knowledge base software example, perhaps the results of this keyword research would tell us that “knowledge base” gets a ton of search. This is a high level, broad term, and can be tackled with a 10X Pillar Page.

But we also need “cluster” blog posts to complement this.

(In other words, we want to rank for the broad theme of “knowledge base,” and to do so we’ll need to write many, many related blog posts targeting long tail keywords).

Long tail keywords here could be things like “how to create a knowledge base,” “knowledge base examples,” and “self-service customer support benefits.”

To find search volume (estimated), you can use tools like Ahrefs, SEMRush, etc. Then use this information to plan and prioritize your content calendar.

With regards to content planning, I think the best marketers operate with a foot in the past as well as one in the future.

By that I mean that some of your strategy should come from historical data like search volume, but some of it should be based on looking forward and setting the trends yourself. You know what an only-SEO blog strategy looks like (in a word, “mediocre”).

Auditing Your Current SEO Results

The second part of your SEO skill set should include measuring your efforts, auditing it from time to time (from a content perspective at least), and improving or optimizing things based on your findings.

Of course, you can and should set up keyword rank tracking on something like Accuranker, but you can go even further and analyze broader trends in search data.

Step one: set up scheduled reports in Accuranker to monitor rankings

The big thing here is to be able to track drops in traffic as they are occurring.

The majority of my roles in the past have been with companies who heavily relied on organic traffic for acquisition. Therefore, it was always anxiety-inducing to see a drop in traffic and not be able to pinpoint its reason.

While you may have a digital analytics or SEO specialist working specifically on this stuff, it still helps to have the skill set as a content marketer.

Here’s a quick process for doing that (abridged for my and your sanity):

Run a Google Analytics report at the end of the every month to see which of your site’s top URLs have lost traffic.

To do this, go to GA, pick your highest traffic month from the past 6 months (if your traffic is relatively stable and you don’t have much seasonality), or pick a period of 6-12 months that you think represents your data (say Jan 2016 to Jan 2017).

Then go to the Behavior > Site Content > All Pages report. Set your date range (whatever you chose as your representative data set).

If you’re analyzing a property that doesn’t have a given segment, click “advanced” and apply this filter: “Include” “Page” “Containing” “[Property URL]”. Or just filter out pages based on URL structure (like /blog/) if that’s the way your site is set up. Point is: include only blog posts.

If you want all traffic, you can continue on. If you want only organic traffic, set up a filter for that.

To do that, apply a secondary filter of Default Channel Grouping and include only those Channel Groupings that contain “Organic Search,” like the following:

Make sure you include all rows, and then export it to CSV.

Open your spreadsheet and name the first tab whatever month and year it is (Nov 2017) then rename Column B something like “Pageviews Nov 2017” (or Pageviews 2017, depending on what date range you chose) and delete all the other tabs:

Go back to GA and change the date range to the most recent fully completed month (or similar time period to the one you’re comparing it to – if you chose a 12 month period, pick your most recent fully completed 12 month period).

Always compare apples to apples with time period comparisons.

Do the same thing with the filters and the rows and whatnot, and export to CSV. Bring it on over to your other spreadsheet in Tab 2, and name that tab something like Jan 2017. Delete all rows except for Column A and B again.

Now add another column to tab #1 and name it something like Pageviews Jan 2018. Now do a Vlookup, like the following (in column C) where ‘tab 2’ is the title of your tab:

=VLOOKUP(A2, ‘[tab 2]!A:B, 2, FALSE)

Should look like this:

Now we’re going to see if there has been a significant drop. You can do whatever percentage you think is significant, but this example we’ll flag anything that has dropped by 20%.

Add column D and title it “2%+ decline?” then insert this formula in D7:

=IF(C2<(B2-(B2*0.2)),TRUE,FALSE)

That’s essentially asking if the number in Column C is 20% or greater less than the number in Column B. Then you can do conditional formatting to highlight those where that is the case:

Note: My example data here probably doesn’t give the best example because my comparison month was a holiday period (lots of traffic!) and I compared to January (low traffic!) so of course there is a large drop (it’s ecommerce). But you can tweak the time periods and parameters, I’m just trying to show you how you can easily get some good trend data to explore.

From there you can explore why you’re losing traffic (is search volume declining in general for your terms? Are competitors stealing your rankings?)

One more way to do this is simply to create a custom report that only includes blog pages and organic traffic, and simply track trends. Are you down month-to-month, year-to-year? This report will show you the broad answer and then you can drill down and find out why.

Finally, the easiest but roughest way to notice trends in your organic data is to set up a simple custom alert.

You can do it either way, tracking spikes or falls in your data.

You’ll want to get a baseline level of your data, ideally at a yearly level since so much can effect your week-to-week and even month-to-month traffic (seasonality is real). So what would be alarming to you regarding yearly trends in organic search volume?

Set that up as a custom alert so you know right away. It’s super easy to do this in Google Analytics:

Then you can go dive deeper and find specifically why your traffic is in decline (which specific pages, etc.).

How far are people reading into our blog posts?

Google is making it easier and easier to implement robust user engagement tracking via Google Tag Manager. One cool way you can do that is with Scroll Depth Tracking.

Set up scroll depth tracking to see how far users read (Image Source)

Many marketers user bounce rate as a metric to gauge user engagement, but that’s actually not a great metric to track (for many reasons).

I won’t go too in-depth on setting this up (otherwise we’d have to start at square one and talk about setting up Google Tag Manager), so I encourage you to read this guide if you’re interested in implementing something similar.

Once you’ve set up scroll depth tracking, you can use that event to see how far people read your blog posts. You could also combine with with time on page, bounce rate, etc., metrics that are usually used to define engagement, and you can see what the relationships are between the variables.

What is the best time of the day or day of the week to publish?

The time of day and day of the week you publish and promote content matter. And it doesn’t have to be a guessing game. Google Analytics can help answer this (and give you a sweet heatmap to show your team).

The best way to do this is with custom reports in GA.

Note: Google Analytics’ dashboard gives you a heatmap now, but it still helps to do this on your own so you can customize the variables.

Heat maps tell you a quick story on when users are coming to your site

Go into customer reports, and choose the “flat table” option. Add dimensions “Day of Week Name” and “Hour,” and then add relevant metrics. For content, you might care about sessions. For an ecommerce site, you could care about conversion rate by hour of day and time of week.

You can also add things like Source / Medium if you want to analyze, say, only organic sessions.

Then export your data to Excel. Make sure that you show all rows in the “shows rows” in the bottom right-hand corner of your screen.

Once you’re in Excel, you can whip up a quick pivot table based on the metrics you’d like to analyze by time of day and day of week.

To do a heat map, it’s as simple as using conditional formatting based on cell value:

You can also visualize this data in other ways, but I think this is the easiest way. I actually learned how to do this type of heat map in R first, and then later went back and figured it out in GA/Excel. The GA/Excel way is easier.

How to find out if something you changed on site mattered to SEO or it was just noise

The gold standard of online optimization is the controlled experiment: the A/B test.

However, with content and SEO, we’re often fighting to prove causality because it’s hard to control all variables (especially with Google’s constantly changing algorithms).

However, you can estimate the significance of the impact without a team of data scientists using Mark Edmondson’s GA Effect dashboard.

You just plug in your variables and the time horizon that you want to analyze (as well as when the change took place) and it calculates the probability of the change happening by chance.

You can read up more on the stats behind the tool here.

Qualitative Analytics for Content Marketing

There’s a whole qualitative side to this game, too.

Qualitative data can help you come up with topic ideas, gauge visitor and customer sentiment, and chase down usability errors.

In addition, it can help you feed critical business insights to other parts of the organization, too.

How to set up qualitative feedback loops

You can and should use any qualitative surveying or feedback tool you can to collect this data.

However, I find communities are even better places to source insights for new content, as well as feedback on your current content (as long as the community doesn’t suck).

At CXL, one of the highest leverage moves we made was to create a Facebook group to facilitate conversations on CRO, analytics, and other data-driven topics.

Granted, it may have been easier for us to do because we already had a large email list and many previous course students. But you don’t actually need to create a community to get insights; you can become a member of a relevant community and get the same value.

We’d start a lot of conversations ourselves, but we also invited tons of people we knew would be high quality contributors. The conversations were (and still are) super high quality. It’s pretty self-sufficient at this point, and I check back all the time for new updates.

You don’t need to invent the community to get insights and value

When I was at CXL, I would use our group 1) to eavesdrop on conversations and 2) to help form and iterate on content.

Here’s an example of me straight up asking for feedback on a new style of content we were testing:

And here’s an example of me fishing for insights for the second point:

Those posts both got published:

How to get real time visitor insights

Here we want to gauge the intent of individual and anonymous visitors on specific pages. To do this, we’ll strap an on-site polling tool (like Qualaroo) on our site and come up with some questions.

When setting up an on-site poll, you should think strategically and carefully about the question you’re hoping to answer.

Do you need to know what type of an audience is visiting your page? This one came from a VWO blog post:

Do you want to source content suggestions and feedback?

Want to use your survey as a tool to get leads and hand raisers?

You can ask any number of questions here, just make sure to approach it with clear intentions in mind. After all, a poll is a UX hurdle for your users, so it better have a pragmatic tradeoff in terms of the value of the insights you’re getting.

Of course, you can also gauge the intent and goals of visitors to landing pages, too. This helps you align your conversion paths as well as helps with landing page optimization.

How to find usability issues

While I don’t think it’s necessary to become a usability expert (unless you’re really early stage and wearing tons of hats), it’s really helpful to occasionally gauge the user experience of your website and audit your blog for usability issues.

You can send these to a front end developer for prioritization, or maybe you can solve them yourself, depending on the complexity.

There are two general ways I like to find usability issues as a content marketer (there are a few others if CRO or UX is your job role):

  • General feedback forms
  • User testing

General feedback forms, like that of Usabilla or HotJar, help you crowdsource frustrations from your users at scale. If you have 10,000 visitors, maybe 20 will leave feedback for you and maybe 5 of those responses will be useful.

But again, it’s at scale, so you can usually find some glaring patterns and solve those issues.

The second method, user testing, is a whole issue of its own.

If you can get help conducting user testing from someone who has done it before – a CRO person or user experience research – that will be much better, as there is some nuance and there are common mistakes people make.

But I do think user testing can be beneficial for those outside of traditional user experience roles, and I think content marketing is one of those roles it can help.

Essentially, the way to run a user test is to:

  • Recruit target users who have never used your site (5-7 users)
  • Create tasks for them to accomplish (usually a few specific and a few general)
  • Watch them interact with your site
  • Conduct post user test surveys to get attitudinal insights (optional)

There’s more to it than that rudimentary summary, so don’t just jump into it if you don’t know what you’re doing. Rather, read real resources like this one.

You can also just talk to your customers and readers. It’s more manual work, but can be a massive source of insight. I got so much value from being at conferences and interacting with CXL readers and fans. You really can’t overestimate that value.

Proving Business Value

In a larger organization, one of the most impactful things you can do is measure, improve, and evangelize the business impact you’re responsible for.

It can help you improve your own results, and it can also help executives in allocating resources and giving you the props due to you.

What’s ROI anyway?

Content marketing ROI is tricky because there are a lot of long-term and untrackable benefits to doing it. For example, Tim Urban can measure the number of subscribers he gets via his blog (and if his tracking is more advanced, how many per blog post).

But no Google Analytics attribution model will attribute his Ted Talk and numerous podcast appearances and whatever other opportunities he’s had to speak, consult, etc. to his blog – even though it’s quite obvious that was the reason and the catalyst for these opportunities.

That’s an extreme example, but even running a B2B blog, you’ll run into the same issue.

You can track leads and subscribers by blog, category, and post, but you can’t track how your authority built through content has helped close deals, build partnerships, or even raise capital. Those data points are more anecdotal if you ever hear them at all.

So, when you seek to prove out your ROI, you separate that which you can track from that which you can’t, and realize that your measurement strategy for content marketing is always going to be conservative by nature.

Still, though, it’s beneficial both from an optimization perspective and a resource allocation perspective to track what you can. And here we have a few models…

A range of approaches

Calculating the ROI of content, then, isn’t simple. It depends on what your goals are. Are you collecting email addresses? Building site authority to rank ecommerce product pages? Creating thought leadership because your category barely exists in the public knowledge? Building funnels to collect freemium user signups?

All of these goals require a different framework and methodology for measurement.

Therefore, I’m going to cop out here and simply point to resources that can help you structure your ROI measurement:

Conclusion

There’s only so much you can cover in a blog post, and this one is already long. But you get the point: content marketing and analytics don’t need to be antonyms. Rather, you can become a data-driven content marketers, and it can actually help complement the right side of your brain.

Knowing this stuff not only helps you do your job, but it also helps you communicate your value and evangelize your efforts. I think any business function that is serious about garnering resources should learn to speak in terms of business value, and like it or not, that necessitates a conversation centered around data.

Learn the lingua franca and reap the rewards.

Feature image source: Instagram.com/trevoratx

Leave a Reply

Your email address will not be published. Required fields are marked *