Supercharge That Holiday Campaign Part 2: Custom Variables

by Adam Clarkson

Last updated on November 29th, 2018

Black friday tips part 2 crazy egg

The holidays are upon us yet again, which means you’re likely coming up on the final stages of extensive social, search, email and paid campaigns to drive traffic in.

This two-part blog series is about a problem we see often with respect to email marketing campaigns: the gap in analytics between email performance (open rates, clicks) and your ultimate performance metric: conversions.

Few analytics tools fill in that all-important middle piece – visitor behavior on your site from the moment they arrive up to the point they convert, or exit.

This part of the story is vital.

Your successful email marketing campaign will fail if the experience you’re pushing customers to is sub-optimal.

There are multiple ways, in Crazy Egg, to track visitors coming from campaigns and measure their response to your promotional efforts. Part 1 was all about good ol’ UTM parameters which you are probably already appending to your links.

Crazy Egg assumes you are (cause why wouldn’t you), and has product features built in to help you understand how your various audience segments are experiencing your site — specifically, where else, outside of the primary goal conversion, those audiences are clicking.

But not all visitor classifications fit neatly into the taxonomy that UTM provides. And more importantly, there are an abundance of different ways that you want to classify visitors. For example:

  • By their actions
  • Behavior patterns
  • Browser cookies
  • Attributes from your data warehouse

Do you want to look at customer behavior by industry vertical?

By company size?

There could be stark differences in what these segments are interested in, where they’re clicking, and actionable learnings that could profoundly impact your conversion rate against these segments.

With the additional traffic and intense activity on your site, holiday campaigns are the perfect scenario to experiment with custom segments!

In this post we’ll take a look at a powerful feature of Crazy Egg that lets you pass those variables to your website behavior reports to measure with even more precision what those segments are doing once they reach your site.

Putting the “Custom” in Customer

How do we think about the different segments of visitors that come to our site?

Analytics tools handle a lot of this classification work for us. We are offered visitor segmentation by referrer (Google, Facebook, Pinterest), frequency (new vs returning), geography, browser type and so on.

But if we think more deeply about the real fundamental differences of our customers (i.e. differences worth analyzing) we want to see things more relevant to our business.

Let’s look at a few examples.

If I’m trying to understand what trial customers do and how they differ from paid customers, or guests I’d want hooks to see how each performs:

membership status
paid / on-trial / registered / guest

If I’m trying to gain insight on a particular social, email or search campaign:

campaign name
blackfriday_shoes / blackfriday_boots / black_fridayscarves

If I’m trying to build customer profiles out of demographic and behavioral data sources:

type of buyer
bargain hunter / wanderer / habitual buyer

If I’m trying to optimize the path to purchase on my site maybe the most important thing to know is something like:

Visitor scrolled to the bottom of the product page
true / false

Any of these individual segments can be tracked in Crazy Egg. Any segments that you can conjure up, measure and pass to Crazy Egg can be tracked – we’re talking non-scrollers, deep-scrollers, dog people, cat people, lefties, deadheads.

And they don’t have to be mutually exclusive segments either. With Crazy Egg’s custom variables, any visitor to your site can be assigned up to five values.

The Custom Variables Payoff

To get an idea of what this data looks like once it’s set up, let’s look at an actual Confetti report.

When you click on the dropdown menu in the filter palette,  you’ll see User Variables 1 through 5 toward the end of the menu. These are our customizable variables (the configuration of which we’ll get to in a minute).

In this instance, the site owner has configured all 5 User Variables. Let’s take a closer look at User Variable 4.

In this real example, the site owner has configured red dots to mean ‘false’ and green dots to mean ‘true’ (not intuitive to us, but meaningful to the site owner). Convenient of them to use these colors for my holiday season post!

There are over 67K Green clicks on the page, and 38K Red clicks, and a tiny percentage of customers who visited without getting tagged True or False (classified as ‘Unknown’, 37, in pink).

Now let’s look at the full report (the snapshot of the customer’s web page has been removed to respect confidentiality).

There’s a clear difference between the interest and behavior of the Reds (false) vs the Greens (true). Greens are all over the button in the top right and the area just under the top navigation. They also are more interested than the Reds in the items in the bottom/right part of the page. Reds are clicking on some of the same things, but clearly exhibit some tendencies of their own.

It can be beneficial to your analysis to isolate individual layers of confetti to confirm that there is a difference in click behavior, or to see the density and exact position of a particular color. You can do this by toggling the checkbox next to rows in the filter menu.

In the image below you can see areas where there appears to be disproportionate interest by a segment. The blue squares outline those areas.

To get exact numbers of Red (‘false’) and Green (’true’) clicks on a specific link, let’s head over to the Overlay report (again, the snapshot of the page has been removed here).

The Overlay shows us the number of clicks on each link on the page, and for each link we can also see the exact percentage of clicks broken down by User Variable 4 – Trues and Falses. It’s through this report that we can determine exactly how much the two groups differ in their click behavior.

A closer look at some data in the Overlay report:

If we crunch a few numbers we can add another level of insight to our segment analysis, and reach some solid conclusions.

  1. The overall distribution of clicks on this page will establish a benchmark by which we can then evaluate each individual link.
    • Percentage of False/Red clicks overall on the page = 64% (67669 false clicks / 105803 total clicks)
    • Percentage of True/Green clicks overall on the page = 36% (38097 true clicks / 105803 total clicks)
  2. Any links on the page where the False group represents more than 64% of clicks represents a link that the False group has a higher-than-average interest in.
  3. Conversely, any links on the page where the True group clicks more than 36% of the time are links that are more attractive to Trues than the average.
  4. We can calculate exactly how much more popular each group is in a specific link by calculating the percentage increase vs the benchmark this way:
    • (Clicks by a Group on a link – Benchmark of that Group) / Benchmark of that group

If we calculate the percentage increase in popularity for a group, for each of the links on the page we can confirm what we saw at-a-glance in the Confetti report — but also know which links are popular with which audience segment, and exactly how popular they are.

Conclusions

The False group and the True group demonstrate very different preferences to the products (links) presented on this page. The preferences are clear and unambiguous. Reading the numbers in the graphic above we can see exactly what they are. If we were able to see the snapshot of the product page, we could clearly see the product preferences of each group.

Recommendations

What changes would we recommend out of these findings? The owner of this site needs to be talking to her team about dynamically presenting products based on User Variable data.

The Trues have a distinctly different set of preferences from the Falses, and a dynamic page that presents each group with custom product offers is going to perform much better than the current mix of appealing and unappealing offerings.

How to Set Up User Variables

Now that we’ve seen the value of User Variables in snapshot analysis, let’s look at how to configure them and get that data pouring in to your Snapshots.

Setting up User Variables does require a bit of custom coding, but is easily accomplished using the CE2.set Javascript function:

CE2.set(userVar, value)

userVar: The User Variable number you want to set, eg. User Variable 1, 2, 3, 4 or 5
value: Any string value (it may be truncated if it is over 100 characters)

You can use this function when the page loads to, like so:

function CE_READY() {
CE2.set(1,‘guest visitor');
}

In the above example the CE2.set function is nested in the CE_READY function, which is called as soon as Crazy Egg is initialized (thus ensuring you avoid errors by calling it too soon).

You could also set multiple values in one call, like this:

function CE_READY() {
CE2.set(1,'some string');
CE2.set(2,'another value');
CE2.set(3,'yet another value');
CE2.set(4,'more');
CE2.set(5,'and more');
}

The other way to use Custom Variables is to associate them with page events.

For example:

  • Clicks on buttons
  • Clicks on menus
  • When a site visitor scrolls to a certain point in the page
  • When a site visitor reaches a threshold of time spent on the page

Here is a link that will set a variable when clicked:

<a href="javascript:void(CE2 && CE2.set(1, 'clicked link'))"> link </a>

More information on Custom User Variables, CE_READY and CE2 are available in our help documentation.

Custom Variables in Crazy Egg are an extremely powerful tool for audience segmentation analysis. You can go way beyond the typical segments that common analytics tools offer and explore any measurable dimension of the visitors to your site, including gender, company size, revenue, product adoption….

Whatever you track can be appended to audience segments that you can quickly filter through and visualize in Crazy Egg.

If you want to supercharge your holiday campaigns and really understand the preferences and behavior of different audience segments, invest a little time in setting up Custom Variables.

As always, we’re here to help! If you have any questions, reach out to us at help@crazyegg.com.

No Comments

DON’T MISS OUT

Get updates on new articles, webinars and other opportunities:

NO COMMENTS

Comment Policy

Please join the conversation! We like long and thoughtful communication.
Abrupt comments and gibberish will not be approved. Please, only use your real name, not your business name or keywords. We rarely allow links in your comment.
Finally, please use your favorite personal social media profile for the website field.

SPEAK YOUR MIND

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Show Me My Heatmap

A3: @CrazyEgg was one of the best tools I have used to understand what and where users are engaging with a page. #seochat

Matthew Diehl

@matthewdiehl

What makes people leave your website?