Is your audience behaving? What online metrics really mean

Analyzing website data can be tricky, not to mention highly subjective. Although you can pull a lot of information about a single landing page, there are also many ways to slice it up. 

If knowing which metrics to pull for your analysis is complicated, determining what they’re telling you — and what your next step should be — can be even more complex. The interpretation of any given metric can vary based on your goals, the page or channel you’re reviewing, your audience, and many other factors. 

In this article, we’ll discuss how you can get deeper insights by bringing context to your data analysis. We’ll start with some common examples of how data can be interpreted — and misinterpreted. We’ll also cover some useful tips for setting up websites that may help you and your audience get the information they need.

Is time really money?
Many marketers pay close attention to the average session duration, i.e., the amount of time users spend on each of their webpages. It’s common to think: “The more time someone spends on my page, the better.” That might be true for a content-heavy page featuring a white paper or one providing detailed product information. 

But what if you have a category page designed to route customers between your homepage and a product-specific page? A long session there could mean a user is struggling to find what they’re seeking. If this happens often on a page where it shouldn’t, it could be a clue that you have a user experience (UX) issue. 

Similarly, users who spend too much time on the shipping step of the checkout process may have concerns about shipping costs, delivery times, etc.

Is your audience clicking with you? 
Another common metric for evaluating online marketing is your click-through rate (CTR). This metric measures the percentage of viewers who clicked on an ad out of all those who saw it (impressions). 

This is a great way to identify engagement, especially if you’re testing different options. Traditionally, the higher the CTR, the better (not to mention the easier it is to justify marketing expenses to your boss). This mindset has pitfalls, though. If you use your CTR as your only criterion for determining which marketing channels to use, you may very well throw out great programs that are helping you in other ways.

Paid search ads typically have higher CTRs than channels like display, social or video. Let’s say you’re running a campaign across several channels. Your Google Search ads are achieving a 4% CTR while your LinkedIn ads are achieving a 1% CTR. This is because users who come across search ads are actively searching for relevant links to click. They’re in the market for information and expecting to find at least one resource. 

Paid social ads, conversely, interrupt a user’s regular activity. We’re pushing content toward them when they’re not searching for us, so it’s only natural for the CTR to be lower. 

But that doesn’t mean social ads aren’t effective! Even if a user isn’t ready to buy from you today (and the chance of that is about 95%), your social presence still drives visibility for your brand. Over time, this results in more leads coming your way at a lower cost and other benefits.

Does it pay to pay per click?
Common sense suggests that you want to pay less for leads. But again, it depends on the context.

Let’s say that leads are really expensive for keyword X. But if deals closed with users who searched keyword X closed at triple the rates of those driven by less-expensive keywords, the cost could be worth it.

Similarly, keyword X might cost $500 per lead but deliver a return on ad spend (ROAS) of 20:1. If that’s the case, it could easily outperform a keyword Y that costs $100 per lead but only delivers a 3:1 ROAS.

Setting up your site for success
As these examples have shown, analyzing metrics without context can put you at risk of misinterpreting your data or, worse, missing out on some great opportunities. 

Start   by setting yourself up with a question or hypothesis about how visitors will interact with your site. The more specific you can make these, the easier it will be to assign the right metrics. But even a few loose speculations will help provide a framework for what you hope to learn.

Next, select the metrics that will best help you evaluate your theories about user behavior. To find the context, put yourself in a user’s shoes. What would you want or need? How did you find the site? What are you looking for?

Your site should be structured around fulfilling two sets of goals. Ask yourself the following questions.

What do my customers want from this site?
Possible answers could include, but aren’t limited to:

• Download a data sheet.
•  Watch a how-to video.
• Find a repair manual.
•  Find and contact a sales representative.
  
  

• Find and contact a sales representative.
• Compare product features.
• Order parts or related products. 
• Download a white paper.
• Watch a product-related video.
• Something else
  
  

• If you’re reviewing visual campaigns, also pull up your ad variations so you can see what the audience is seeing. Similarly, when you’re reviewing landing page data, visit the page itself.
• Have your audience demographics or targeting handy when you review click performance.
• Avoid looking at the same time periods every time you review your data. Back your window up 18 months and see if you can establish trends. Or narrow in on just the last three weeks to see if you can spot shifts in activity.
• Experiment with any segmentation options available in your platform to spot underlying factors or trends that might be skewing your overall data. Consider differences of geography, ad types, channels, etc.
  • If you’re reviewing email data, send yourself the same email so you can see exactly what your mailing list saw.