Every year around this time, people tend to make the same mistake. No, we’re not talking about falling for April Fool’s Day jokes (or publishing them in the first place… ha ha ha.) Rather, the mistake I’m referring to is about date comparisons and one of my most hated units of time… the month.
In case you were unaware, not all months have the same number of days. In fact, if that first sentence shocked you, you’ll be positively gobsmacked to learn that the month of March typically has 11% more days than the month of February.
So, whether you’re thinking about website and marketing performance with Google Analytics or clicks and impressions with Google AdWords, before you send off your monthly dashboard with your shiny green arrow comparing month over month performance, let’s take a second to consider what this means.
Each year, in companies all over the world, we can recognize a familiar pattern that occurs.
1. A Strong Start to the Quarter
A new year begins. January, with its 31 days, exceeds our December numbers. The holiday seasons in November and December took their toll, but now we’re back and running full steam ahead! (Of course, some industries will do better during holiday seasons…)
2. A Slump In February
Oh no, February traffic numbers dip pretty low. We might panic, we might explain it away by bringing up our ‘number-of-days’ issue or just attribute it to January being a great start.
3. Marching Forward!
Bam! And just like that, we’re back! Our numbers are strong, and this time, while explaining our reports, we may fail to mention those pesky ‘number-of-days’ and instead claim full credit for the strong rebound.
Just Give Me An Arrow
Digital marketers, and really anyone who uses data of some sort, often have to find ways to report on how things are going. Whether it’s the results of a specific campaign or just the current status of their website or project, everyone wants to know “How are things going?”
This has led to all kinds of innovations over time in how we communicate this information. Not only that, the number of data points that we’re expected to follow has increased exponentially, with every new tool that we purchase, or each new metric we decide is critical to our business.
Our methods for conveying the current status may be getting more technically advanced, with whizzbang dashboards built in Tableau or Google Data Studio, broadcast to corporate TVs or delivered via automated emails. Yet, too often, the amount of information we ultimately convey is no more than a thumbs up while passing someone in a hallway.
People love arrows. Green arrows, red arrows. They want the high-level, at-a-glance, give-it-to-me-straight-and-make-it-snappy version. Not only are we vying for budgets, job safety, political or corporate capital – too often we’re simply vying for someone’s attention.
This has led to the ultimate state of the Month over Month dashboard. It comes at an expected time, the end of the month, the time when things get reported on. Contracts, calendars, and fiscal years often align with ends of months, why shouldn’t our reporting?
Again, I’ll stop to remind everyone that months are terribly inconsistent units of measurement.
From Pageviews to Contact Form Submissions and possibly even to Opportunities Won to Revenue, the same problem exists. When we compare metrics that count up over time, some metrics will simply do better when we have a longer date range.
Well Fix It, Dear Henry
We know it’s an issue, but let’s talk about how we can address this. We can’t fix the calendar and realistically, we can’t get rid of our dashboards, monthly checkins, or green/red arrows. So what’s left? Here are a few options – take what you can!
This first step to fixing this issue is realizing that it exists. Just because there are a different number of days, doesn’t necessarily mean you’ll see it so plainly. Any number of factors may affect your actual numbers, which may hide or even amplify the differences. A change in your Google AdWords strategy in February may completely make up for the lack of days, or timely factors may contribute to spikes or dips in traffic. Are you publishing blogs, holding events, or attending trade shows? What type of seasonality does your site or industry have?
Most likely, you’ve realized how many days are in each month. Perhaps it might even seem, dare I say, obvious. But – is it obvious to the people reading your reports? Glancing at your dashboards over coffee or while in the middle of an unrelated meeting? The responsibility lies with the report creator.
Explain and Add Context
Reports are better with context, with explanations. That might mean that you present the data and have a chance to explain everything, or you find some way to attach the WHY next to the WHAT. Look for ways to include the context or the story to help make it easy for the reader.
Google’s Data Studio (and almost every reporting/visualization software out there) has the ability to add text boxes and descriptive information into the report before distributing.
This is your chance to show off, explaining the factors that went into the numbers and adding your analysis – is it a big deal? Do we need to change something? This is what sets you apart from being a simple reporter vs an investigative journalist. You’re not just repeating the stats, you’re explaining the stats.
Choose a Different Date Range
This might not fly in your organization, but consider different date ranges when you’re comparing a month’s performance. Google Analytics offers the ability to compare to the “Previous Period,” which will take your selected number of days and compare against the same number of days going backwards. So for March, it will look at the previous 31 days, grabbing a few January days to mix with the month of February
As another alternative, you can take the smallest month (February) and always use that number of days for your comparison. You could even change the distribution of your reports so that you begin reporting every 4 weeks, rather than every month. This ultimately gets you away from the month model, which may or may not be realistic.
Compare Year over Year
Don’t just look back at March compared to February, but instead look at March 2018 back to March 2017. Same number of days to compare, and especially for months like December with lots of holidays, this will be a much more consistent comparison.
Use Different Metrics
The biggest issue here is when you’re looking at the straight numbers for things like Sessions or Pageviews. If you’re looking at rates or averages, Bounce Rate, Session Duration, Sessions/Day, etc. the effects of the too-many-days problem are largely minimized. (Though, if we really want to get into it, perhaps we should be looking at ratio of business days to non-business days per month…) Create calculated metrics if needed, or choose metrics related to specific campaign rather than a month.
Ideally, numbers about traffic to your site or engagement with your ads are shown side by side with meaningful numbers about your bottom line. Google Analytics does a lot out of the box, but it requires customization to set up important pieces like Goal Tracking and Ecommerce reporting inside of Google Analytics. Sure you may have had more traffic, simply because there were more days, but tell me if it affected your conversion rate.
Use a Different Report
We’ve written about this particular challenge before, and Hannah wrote about a better way to look at trends over time using something called moving averages. Take a look at how to use moving averages with your Google Analytics data to help spot trends over time and minimize the date issues you’ll get by using months.
Published: September 26, 2017
And Now You Know
Whether you’ve always considered the number of days, or whether it makes a big difference for your particular type of report, the takeaways are the same. Make sure your company and team are looking deeper than the thumbs up/thumbs down to understand why certain numbers went up or down. Explain the differences in simple to understand statements, and make sure you’re tracking the right information to begin with. Happy reporting!
The post Digital Marketers: Don’t Be A Reporting April Fool appeared first on LunaMetrics.