Earlier this year, at the Nonprofit Technology Network’s annual conference, I introduced the concept of Rorschach Analytics as a way to describe a common and all too natural approach to analyzing metrics.
Often, especially in the nonprofit sector, the review of marketing data is left to non-technical staff who are rarely trained in data science and are left to interpret analytics using the assumptions and perspectives that they bring from their respective roles: a social media manager looks for a live-tweet, a policy analyst looks for a white paper, a fundraiser looks for a bump before the big gala…
We know the drill. You see a traffic curve, you think about the things you did during that time period, and you make up a story about how you feel your work probably fits into the chart in front of you.
End of story. Analytics accomplished.
And this might do well enough if you’re not particularly interested in knowing what actually happened and would rather just move on with your usual work (which is very often the case).
But what if you’re wrong, and your work didn’t have the impact you expected? What if there were missed opportunities to optimize the campaign that you weren’t aware of? What could you have learned from your last push that might make your next one more successful, if you only looked closer at the data?
With the data you have available, you can do better than imagining pictures in inkblots.
Move beyond Rorschach Analytics by studying the extra dimensions of your data to see what’s really there, and what you only imagined.
Ask better questions of your Google Analytics
Google Analytics helps you do this by collecting and reporting loads of information about your visitors. Start by drilling down into your data using the most common types of dimensions, things like traffic sources and your site’s content, to see if people came from and went to the places you expected.
Here are a few tips on how to get started and be a bit more methodical about your analysis:
- Browse the standard reports. Exploring your traffic from more than one perspective by using the left-hand menu to explore these reports. Popular starting places are the “Acquisition” report that shows where visitors to your site came from, and the “Behavior” report that shows your most viewed content. This can give you a sense of what was popular, but both give you a broad view of what happened across your entire site. Browse around the many options try not to feel overwhelmed—there are more reports available than you’ll need on a regular basis, so find a few favorites that matter most to you. Don’t give in to the temptation to stop at the overview reports; drill down to see things your visitors’ language or traffic channel.
- Secondary dimensions. Your best friend at this point will be the “Secondary dimensions” drop-down that you’ll find at the top of most detailed reports. For example, when you are looking at a report of your site’s content, try clicking on one page that interests you so it is shown alone, and then add “Source” as a secondary dimension. Rather than showing you where visitors came from across your entire site, this will now show you JUST how people who read this particular page came to your site. Right away, you may find surprises—you may put a lot of effort into social media, but some pages get most of their traffic from Google Search. What can you learn from this page to improve your SEO? Could you make better use of this page by adding a call-to-action that moves visitors elsewhere?
- Plot Rows. Sometimes when looking at these reports over a whole month, we’ll spot our most visited page at the top of the list, then look up at the trend line, see a spike, and say “There. That’s when this page got all its traffic.” Perfect time to check yourself. Click the box to the left of the page you suspect (or source, etc… if you’re looking at traffic sources or other reports), then press “plot rows” to see how that item stacks up against the whole site’s worth of traffic. Was it as big a driver as you thought, or might there be other factors at play?
- Introduce advanced features. Once you develop a sense for the information that really matters, start setting up custom reports to making drilling down into these questions easier and more consistent. Start with content, split by traffic channels, narrow further by specific traffic sources within those channels, then switch to device type to see if folks on mobile prefer one partner website over another. And when you’re ready to start collecting data that’s truly unique to your organization, consider how custom dimensions, content groupings, events, goals, and segments can be used to make your reports more informative.
When in doubt about where to take your analysis next, step back and think about what your team actually does on a daily basis and consider what they need to know to do it better. This exercise is at the heart of much of our analytics work, and makes all the difference between providing useful data or just a distraction.
✱ The extra special unexpected asterisk
But here’s the kicker, there is also immense value in Rorschach Analytics, so move beyond it, but don’t leave it behind! Subjective interpretations of what may have caused your traffic can tell us a lot about what’s important to your organization. These innocent assumptions, unbiased by the data, might not get the whole story right, but they can give you a new lens through which to view what actually happened and help you uncover hidden details about your audience.
The semi-educated guesses that come from untrained analysts are analytics gold. They reveal the thought processes and hypotheses that drive your organization’s communications decisions. Do not suppress them. Encourage them, collect them, mine them for questions you should ask of your analytics. These questions can show you what your next report should focus on to either validate assumptions or correct misconceptions in your organization. These questions can also point you to a corner of your analytics data so small that you never would have found it on your own, but so rich in value that it changes the way you market in the future. In other words, the proverbial “dumb questions” that analysts tend to sneer at can make your analysis smarter than it was without them.
Lastly, the stories we get from non-analysts will often point to flaws in our data, rather than flaws in their assumptions. If your analytics insist that a certain PDF has never been downloaded when your colleague sat next to a donor as they did so, perhaps it’s time to verify that your Google Analytics events are really firing the way you think they are.
Your audience and your internal stakeholders use your site on a daily basis—this is the best test your site will ever get, so listen to them when they report back the results! If you know how to use it properly, this feedback can make your analytics and your whole communications effort work better. This is the value you can get from Rorschach Analytics.