Analytics: Do the Math

Check your analytics on a typical day and you’ll see numbers like 5.87, as in traffic is up 5.78% and 62 as in 62% of traffic came from search. You might see numbers smaller than that — like visitors spent an average of 1.85 minutes at your site — or larger, as in 32,564 pageviews last month.

In other words, you’ll see all kinds of numbers. Some numbers are easier to grasp and deal with than others, though.

Chances are, if you pay attention to your analytics but you’re not finding much actionable data, you are being stymied by one of two things: small traffic or large traffic.

Now, I’m not going to define either of those numbers for you. We know that something like 90% of all websites have fewer than 10 visits a day. Among our clients, the most popular site had five million visitors last year, and we know there are much more popular sites around. For the purposes of math, though, numbers between 50 and 500 a day are the easiest to analyze at a glance. So if your numbers are in the ones column or in the thousands column, it’s going to be a bit harder to see what’s going on during your daily analytics visits, whether that’s more or less traffic than you want — or just right.

Small amounts of traffic

When you have a new site, or your site just hasn’t developed much traffic, it’s hard to get much out of your analytics. If you had 3 visits last month and 11 this month, does that mean anything? If you usually get 2 visits on most days but on Thursday you got 4, does that say something about Thursday? Is that 3,700% increase worth celebrating if the previous number was 1?

You can never tell as much from a small sample size as you can from a larger one, it’s true. But you can still get some value from analytics if your numbers are small by focusing on two things:

  • change over time
  • surprises

That is, going from 3 in April to 11 in May doesn’t necessarily mean much, especially if you had 11 in March. But a series of visitor counts like 3, 11, 19, 27, 35… that shows steady improvement. It’s slow improvement, and you might want to increase its speed, but it’s not random variation.

What’s more, if you have a series like 3, 2, 2, 3, 5, 4, 3, 11, 19, 27, 35… then something happened. It’s worth tracking it down and figuring it out. and a sequence like 3, 2, 2, 3, 35, 3, 4, 2… is really worth figuring out. You want to know what you did right that month so you can do it again.

Large amounts of traffic

When you have small traffic, you sometimes don’t get enough data to see patterns. If you have lots of traffic, it’s easy to miss patterns. At FreshPlans, our lab site, we hadn’t realized that we were getting an increase of traffic at our rainforest pages because they are overshadowed by our most popular pages. They’ve had a 52% increase recently, but they get hundreds rather than thousands of visits, so my quick glance didn’t catch that.

With FreshPlans, we like to catch waves of interest as they arise, and we could very easily have been caught without our surfboards in this case.

If you look at your large numbers and see nothing much going on, look at smaller lengths of time: a few days instead of a week or month. Also check Google Intelligence. Google, being a computer, doesn’t overlook things like a sudden increase in time on site among British visitors. That stuff doesn’t stand out as much to human eyes.

With large data, you can have trouble seeing the forest for the trees. Going in with a specific question in mind can help you identify trends. Then you can narrow your scope and look more closely.

Big or small

Of course, we didn’t really do much math here. Google Analytics does the math for you — you just need to be able to interpret the graphs. The important part is getting from all those numbers to the actions you need to take. We can help. Contact us to discuss your data and how it can lead you to strategic action.


Posted

in

,

by

Tags:

Comments

2 responses to “Analytics: Do the Math”

  1. Nick Stamoulis Avatar

    Analytics is an important piece of the puzzle when it comes to SEO. Without it, there’s no real way to assess if the SEO program is effective or not.

    1. Rebecca Haden Avatar
      Rebecca Haden

      Exactly! Without that data, every strategic decision is just a guess.

Leave a Reply