Using Google Analytics to Generate B2B Content Ideas

Editor's Notes

• #2: Run-through of best GA features to inspire future content, then we’ll look at how to extrapolate that data from analytics and feed it into our editorial plan. “Unexpected”
• #3: Big part of my job and my editorial team is figuring out what kind of content we should be writing for each client. One size most certainly does not fit all and we have to put the right content ideas into practice, or we lose the client. Always looking for new tools to achieve this.
• #4: The answer is ‘content marketers’. It’s not just for SEOs. Google has given us a brief to “write quality content for the user”, which is a lot easier said than done. Fortunately, it’s also provided us with a lot of insight about how our users have responded to our content so far. GA is “the single biggest indicator of how the user really feels about your content.” As content marketers, we can use this info to generate ideas for new, improved, more targeted content.
• #5: Going to start with a few results from a recent survey we carried out amongst content marketers about content inspiration.
• #6: As I’m sure we all know, regular content is essential for SEO, so it’s great news that our content marketers overwhelmingly said …
• #7: “Yes”, they do add regular content to their sites.
• #8: However, when asked whether they produced as much content as they would like, they overwhelmingly said …
• #9: “No!”
• #11: I can’t create extra time for you, but hopefully I can show that by opening up a new avenue for inspiration, this can provide some motivation and shape your planning too. Let’s see if we can kill these 3 birds with one Google-shaped stone!
• #12: So, onto GA
• #13: Thankfully, most content marketers understand GA has a role to play, but
• #14: What role exactly?
• #15: 6 out of 10 marketers are not using their GA data to generate ideas for new content! This is a massive oversight.
• #16: I’ve created a custom Content Inspiration Dashboard within GA with 5 widgets linked to interesting reports. You’re welcome to download the link to import it to your own GA account. Dashboard is the birds eye view – enter each report to study what’s going on and find new angles for content. Ready for most shocking statement of the conference? This dashboard completely ignores the amount of traffic your content has generated! Not because it’s unimportant, but on the basis that it’s too easy to declare your content a success on the basis of how many hits it got, and to ignore every other indicator within GA of its true quality. In the long run this could come back to bite us, because as we all know, Google is working tirelessly to differentiate not just between good and bad content, but between excellent and average content as well. Having said all that, the examples I’ve used exclude any content with fewer than 50 views (– you can set your own number). I’d also advise looking at at least a few months’ worth of data within each report to gain a better sense of how your content has performed over time.
• #17: First widget may look unoriginal – everyone knows how to look at the split between new and returning visitors. From this, it’s pretty obvious that most of our visitors are new. But let’s make it a bit more interesting:
• #18: By comparing avg. session durations of returning and new visitors and benchmarking against the site average, it’s clear that while returning users make up only 20% of our readership, they’re by far the most engaged and loyal. They stick around twice as long to read our content! So how can we use this discovery in our content planning?
• #19: New vs returning content – mirrors the well-known marketing stat that it’s 7 x easier/ cheaper to sell again to an existing customer rather than recruit a new one. Most sites only speak to new visitors. Don’t ignore your returning visitors, reward them.
• #20: If you want to get even better at creating content for your returning visitors … (no widget for it yet, hence why it’s not on the dashboard). Recency not a real word! Your content publishing schedule should be in line with – or ideally, a step ahead of, the frequency with which users visit your site.
• #21: Next, let’s look at whether users in particular locations respond especially well to your content? Next widget is UK cities, you can change this to countries if you have global audiences. Again, a pretty standard pie chart with split of traffic across locations, but let’s dig deeper …
• #22: Have filtered results by average time on page by location, and benchmarked against site average. This shows us where are fans are, to help us position future content to them. Also shows us where we need to work harder.
• #23: Strong locations – find and curate cool content from these areas. Weak (target) locations – create case studies/ interviews, prove the relevance of our offer to potential clients in these locations.
• #25: Doesn’t tell us much on its own
• #26: Segmenting by new and returning visitors is much more revelatory.
• #28: Switch this on manually. Search depth – how many pages they view after performing the search.
• #29: Have sorted by time after search, compared again average search depth to look for the terms that were used by our most engaged visitors.
• #30: Switch this on within
• #31: Most important widget for future content planning – identifying our best performing content. B2B always worry their content isn’t interesting – this will tell you! (Interesting is relative. Right answer to the right question at the right time is always interesting!)
• #32: Have linked widget to All Pages Report, but you can run the following exercise on any report from Site Content, e.g. landing pages. You can create content groups within Analytics. Filtered by: blog pieces only – good idea to compare like with like, ordered by average time on page, again excluding any pages with fewer than 50 unique views.
• #33: Now we know which pieces were popular – but why? Our external content analysis gives us far more targeted, insightful ideas for future content. Export ‘most popular’ to Excel to look at how successful the pieces really were. The more pieces you analyse, the better. Think outside stuff that just has a ‘number’.
• #34: Added length and average time per 500 words to gain a more accurate insight into how compelling these pieces were. Look at items 2 and 3 on the list – they have near-identical average times on page, but one is nearly twice as long as the other. So really it performed only half as well.
• #35: Resorted the list according to most popular according to average time per 500 words.
• #36: Then examining other features (explain). We’ve only looked at a handful of pieces for this example, but the exercise will be more meaningful if you export several months worth of blog pieces. You can also run this exercise on your least popular pieces, to figure out what went wrong.
• #37: Don’t get caught in analysis paralysis – it’s not the numbers that matter, it’s the patterns they reveal. What are the common features across your best/ worst performing content? Use your GA data as a starting point: create a custom dashboard, or adapt mine, so you can quickly access the sort of data about your content that helps you plan even better content for the future.