Question your assumptions about your data to be sure you’re getting the right message.
I recently did a show on how analytics can deceive if you don’t have the proper context. I ran that idea past my friend Michelle Drewek from Lessiter Media, who is a data gal – she’s very good at digging into the data and finding the useful nuggets – and she’s agreed to join me today and discuss the topic. The following is a slightly edited transcript of our interview.
***
Welcome Michelle.
[Michelle] Thank you for having me, Greg.
[Greg] You looked at what I had to say about “data without customer feedback,” and you had three very interesting observations. The first had to do with average session duration. Can you unpack that a bit?
[Michelle] So when I came more into the data role within my position here, I was looking at some of the legacy stats that our upper management team is looking at. And one of them that I questioned a bit was how they were looking at the “average time on page.”
(This would be a universal analytics metric for those who are working in the GA4 world.)
It would tell you the average time on page as an average for your site, but you could see the page by page breakdown, and I questioned why we were looking at that because average time on page could vary drastically for a video versus a few paragraphs or an entire feature, and the number isn’t telling you much.
So we switched that up and I said let’s look at the average session duration, trying to keep people on the site for a longer period of time – be that through multiple articles or through one extensive article. That helped us make some decisions about content – whether long or short form was working better for us.
But now that we’re in the GA4 era, we have a new “average engagement time” metric, which is a lot more helpful in getting to a more narrative-based number.
But again, you still need to have some of that context. Context is everything when it comes to data. If you’re going off straight numbers, you’re not really going to learn much about how your content is performing.
[Greg] So you were able to use that to adjust how you were doing content on the site.
[Michelle] Yes, absolutely. Over time you’ll see shorter form worked great at one point, but then all of a sudden people wanted longer form content.
Of course SEO plays into that too. Now you want to have your short video but also the long form content to help with Google. So it really helps to evolve your strategy along a bunch of different parameters.
[Greg] You also have an interesting way of qualifying stats on page views. What is it that you do?
[Michelle] I’m always quantifying page views month over month and year over year, but we could have one article that performs really well one year, and that could make the metrics skyrocket and make it look like this year we’re knocking it out of the park.
That’s where I take the data and say, hold back, let’s take this one out and look at our average page views without the anomaly.
Then we can look at things like seasonality. There might be certain topics or certain industries that play well. Then you identify an article that’s strong seasonally, and you can plug that into next year’s strategy and find a couple more topics like that while the energy is hot behind that topic.
That can help you in terms of context and knowing how the numbers are fluctuating, but also why they’re fluctuating, and then making decisions off of that.
[Greg] That reminds me of a statistics class where sometimes you would knock off the top and bottom 2 percent of the data and look at the more typical numbers.
[Michell] Yes, absolutely. Anomalies are great to have because those can help drive content decisions, but you’ve really got to look at them as outliers from what you’re doing on a day to day basis.
[Greg] In my article I mentioned the need to get actual customer feedback. You guys do that. What are your chief sources of customer feedback, and how has that helped you with your job?
[Michelle] We use multi-channel sources. We have emails, survey, polls, social media …. One thing I’ve been doing is using the comment section on Facebook. That’s helped to drive content decisions. We have a lot of different audience members who have different stakes within a particular industry. It’s interesting watching them go toe to toe on certain blogs or articles that we’re posting.
I’ll dig in and try to find other content related to that to plug into the next daily email update, or into social media, to keep driving engagement on that content.
Sometimes we find discussions around content that was only in print first, so we don’t have traffic figures on those articles. We bring them into the digital space and get traffic and eyeballs that never saw them before.
It creates a cycle where we can bring a print piece onto the website all because of what people were saying on Facebook.
[Greg] As I recall, a lot of your editors have personal contact with a lot of the readers and call them, so you get feedback that way.
[Michelle] Yes, absolutely. LinkedIn is a big one for us, where our editors will put some of their content on LinkedIn and we’ll get some industry feedback from there. We’ll also post some of our polls and surveys, or the editors will send out personal emails to get direct feedback.
We find that we get a lot more narrative that way, when they feel like they’re talking to an actual person. So even some of our marketing and audience development efforts come from the editor or brand manager.
[Greg] So, there you have it, folks. Some practical tips on how to make sure you’re getting the right message from your data.
Michelle, thanks so much for lending a hand today.
[Michelle] Thank you for having me on, Greg.
