Digital Marketing Analytics and Data Driven Decisions
As a marketer, there is one thing I rely on to guide my decisions regarding future strategy and...
One of the biggest differentiators between digital marketing and traditional marketing is the data and analytics. We can access so much data available with just a few keystrokes—literally at our fingertips. The sheer amount of information is almost overwhelming. And for many marketers, it is.
There are two types of analytics we should be monitoring: quantitative and qualitative.
Quantitative data are things we can measure with numbers and statistics. They’re indisputable facts. Google Analytics is the most popular quantitative reporting tool in the marketplace. Quantitative analytics let us track things like:
Qualitative analytics are soft metrics that aren’t so easy to measure and require deeper thought and analysis to make data-driven decisions. These are more opinion based, and oftentimes arguments could be made either way on which action(s) to take. Qualitative analytics let us measure user engagement metrics, such as:
We can also gather qualitative data via polls or surveys to collect actual feedback from users. Finally, there are user recordings that let us watch actual users during their entire visit to a site. All of this is anonymous, of course.
When qualitative and quantitative analytics are used in tandem, it becomes much easier to make educated decisions to choose a course of action. Now that you’re familiar with the two types of digital marketing analytics, here are 10 mistakes you should avoid at all costs. I’ve seen each and every one of these mistakes in my career, and they’re far more common than one might think.
This one may seem obvious, but I can’t tell you how many times I’ve inspected the code on a website only to see that Google Analytics is absent entirely. And these are relatively well built and designed sites, too.
Running a website without analytics is like flying a plane with no instruments.
Even experienced pilots dread the thought of flying blind. Don’t let your site fly blind. Include some sort of analytics tracking code at the very least.
Once you’ve started collecting data, you need to actually do something with it. Otherwise, what’s the point? Indecision is still a decision. Just do something. Anything at all. Something is better than nothing.
Data-driven decisions are the best kind of decisions because they remove opinion from the equation. No more “I feel like…” or “What if we tried…” or “I read somewhere that…” statements. Those become null and void if you have cold hard data to prove or disprove them.
This is where every business should start. With a measurement plan. Define what success will look like before you start. Determine which metrics you’ll use as key performance indicators (KPIs) and then define what actions need to happen if those are met or not met. Define different KPIs for different tactics or pages.
For example, blog posts always tend to have a high bounce rate—the percentage of visitors without a second interaction. They land on the post, read it, and then leave. This is problematic for sites whose majority of organic traffic is generated with their blog. When we approach this issue with our clients, we use the bounce rate metric on blog post pages as a key performance indicator. If it goes up, we know that changes made to the post template didn’t work. If it goes down, we know they did.
There are plenty of other metrics you might want to measure:
Pro tip: Take it a step further and segment your KPI by source, device, or gender (for example) to see how those specific users are impacting performance.
Filtering allows us to have clean data, and clean data is happy data. This particular data analytics mistake is multifaceted. The first thing to do is filter out any traffic from your company IP addresses. This will exclude employees’ data from the mix. Many companies set computer browsers to default to their homepage. Anytime an employee opens a web browser, they’ll be greeted with the company website. If you’re including those visits in your data, then homepage traffic will be skewed—the severity of which depends on the number of employees at the organization.
It’s also a best practice to set up filters to both include and exclude blog traffic. Having both options is a good idea because blog users tend to interact differently than users of the rest of the website.
Finally, if your website system doesn’t remove the trailing slash from the end of the URLs, or the web server doesn’t do it for you, then you’ll end up with two versions of that URL in the analytics, which is problematic because you have to add up numbers. It’s even more of an issue for metrics that calculate a percentage—like bounce or exit rate. That percentage is weighted by the number of visits, so we can’t simply average them out to see the actual number.
Here’s an example:
Notice how the first four URLs are actually the same page? One with and one without the trailing slash. If we wanted to calculate the Bounce Rate of just the Contests page, we can’t just add up the two percentages and average them out. Doing so would yield a bounce rate percentage of 3.58. But the version of the URL with the trailing slash has about twice the amount of
When we setup Google Analytics filters for our clients, we always start with these at the bare minimum:
Pro tip: Always keep one view without any filters applied. No IPs, no pages. Nothing. If you should ever need to look at the raw data for any reason, it’s there. Once a filter is created, any data coming through cannot be unfiltered so keep an unfiltered view just in case.
This is another cool feature of Google Analytics (GA) that take it from being a 90s hit counter on steroids into a powerful website analytics and user engagement platform. A Segment is just like it sounds—a segment of website visitors. GA comes with dozens of pre-built segments. Here is a handful of them:
You can also create custom segment unique for your business and measurement plan.
Pro Tip: Use the demographics of your buyer persona to build out a custom segment. You’ll need to make sure Demographics & Interest Reports are enabled, though.
When I was a kid, my dad had an old brown electric guitar without any strings on it. One day when I was home alone, I cranked up the stereo in the basement and started shredding Linkin Park riffs on it. When I turned around, my brother was standing there laughing hysterically. My face turned bright red and I instantly stopped what I was doing. Had I known someone was watching, I never would have acted like that.
Surveys have the same effect on people. They behave differently when they know they’re being observed.
Writing good survey questions is truly an art form because people tend to answer as their ideal self, not their actual self. Meaning, consciously or not, they’ll try to give the answer you’re looking for, which defeats the whole purpose of the survey. It’s also tough to write a fair survey—especially if you’re writing it to collect information about your own website. The questions will naturally be worded in a way to lead the user to give the answer you want.
All of that said, surveys or polls can still work as long as you use them correctly.
Pro tip: Stay away from open-ended questions whenever possible. Use ranges, yes/no, or multiple choice options instead. Not only does this make it easier to segment the answers, but it takes opinion out of the equation.
To avoid this mistake, you’ll need some sort of marketing automation software that can collect information about website visitors and track their lifecycle from visitor, to lead, to customer. We prefer HubSpot, but there are dozens of tools that can accomplish the same thing.
Closing the loop is important because it shows where you’re seeing the best return on investment. When you can track ROI, you can do more of what works and less of what doesn’t, thus maximizing the return for every marketing dollar spent.
For example, if you notice that Social Media is going gangbusters at generating new customers, then you know where to focus more time, energy, and budget.
This really only applies to businesses running digital advertising campaigns. And I’m not just talking about using the “All Visitors” list in AdWords and calling it a day. That’s lazy remarketing. Did you know that you can build a custom audience from any Google Analytics Segment and export it to AdWords?
You’ll need to link AdWords with Analytics first, but that’s easy enough. Follow the steps here.
Once that’s done, it’s super easy to build a Custom Audience for remarketing.
From any report that shows a graph across the top, you can add a segment. The default is the All Users segment. Just click the + Icon to add a new segment. If you created one for your buyer persona, just add it to the view. Then, click the little icon in the top-right corner and choose Build Audience.
You’ll be taken to a screen to name the audience for AdWords and then export it out. Once the audience has processed, you can start using it in AdWords as a Remarketing List. It’s one thing to target keywords for services your company offers, but it’s even better to target those same keywords when your persona searches for them instead. What do you think the chances are of those users clicking an ad over the average Joe Searcher?
This is more than simply closing the loop (number 7 above), but closing the loop for specific marketing channels like AdWords, Bing Ads, Facebook, Instagram, or Twitter to name a handful. And conversions don’t need to be when a user fills out a lead gen form either. A conversion can be any action deemed valuable to the business, post ad click.
For example, one of our clients doesn’t sell their products directly to the end user. They have a network of dealers who sell products on their behalf. On their website, this client has a Dealer Finder Form for users to find a nearby dealer.
This form doesn’t collect any information. It’s just an easy way to enter a city or zip code to find dealers. Knowing that we’re tracking a conversion whenever an AdWords user interacts with the Dealer Finder form by either clicking on the Search Box or clicking one of the subsequent results that appear.
Pro tip: Use Google Tag Manager for easy conversion tracking. MeasureSchool.com has tons of valuable training videos about tag manager ranging from beginner to expert. I use it on a regular basis and tag manager has made conversion tracking much easier to implement.
One of the easiest ways to see which terms your site is ranking for in Google is with their free Search Console. This tool will let you see which terms your site appeared for in a Google search along with how many clicks, impressions, and the average position for those terms. Here are the top 10 non-branded search terms our website showed up for in the last three months:
This is sorted by number of clicks, but we can also see the impressions, CTR (click-through rate), and average position. This data is powerful because we gain insight into the minds of our visitors.
We then take this a step further and look at the pages these terms are raking for (hint: they’re all blog posts) to find
Here at Leighton Engage, we use these measurement tools for our clients:
There you have it. 10 digital marketing analytics mistakes you now know to avoid. As you can likely imagine, avoiding all of these is hard, and many require a more advanced technical understanding to even implement.
If you’re feeling overwhelmed at the thought of missing out on crucial digital marketing analytics, don’t fret. That’s what we’re here for. We’re happy to answer any questions and to point you in the right direction.
As a marketer, there is one thing I rely on to guide my decisions regarding future strategy and...
Creating and planning your upcoming marketing strategy? Is email marketing part of it? If not, it's...