What is visitor segmentation?Segmentation means to divide your audience data into parts (or segments), which are definable, accessible, actionable, profitable and have a growth potential. If you wanna get started with segmentation, you’ll have to dip your toes in web analytics. To make sure you don’t get lost on the way, let’s look at some definitions and vocabulary.
Dimensions and metricsYou’ll build segments with both dimensions and metrics, don’t get them mixed up!
- Dimension: A dimension is an attribute of your visitor (traffic source, age, name, …).
- Metric: A metric is a number, usually a count or a ratio (bounce rate, number of visits, …).
User, session and hitA visitor (user) browses your website (session) and look at several pages (hits).
- User: The user is your visitor
- Session: A session is the ensemble of actions a visitor takes on your site from the moment he lands until he leaves.
- Hit: Usually a pageview, event or custom variable.
My recommendation is that going forward all your segments should focus on users first and sessions second. Because if you focus on a relationship, rather than a connection, you will get better business results. You will, of course, create many session and hit level segments. They are also useful from a tactical perspective.
3 levels of complexityFrom experience with our clients, we see three levels of complexity in segmentation—depending on their expertise.
1. Standard segmentationWhen starting out, companies usually stick with basic targeting criteria such as new vs returning visitors, or desktop vs mobile devices, for example, sprinkled with simple behavioral targeting.
2. Advanced segmentationOnce comfortable with basic targeting, you can take your segmentation up a notch (this is where the biggest opportunities are by the way). You then build your segments with criteria such as:
- Acquisition sources (landing page URL, traffic type, referral site, …)
- Visitor characteristics (data from your DMP, CRM, IP, location, browser language, …)
- Contextual (weather, days, date range, time slot, …)
- Technical (device, cookies, operating system, …)
- Behavioral (page viewed, time spent, exit intent, …)
3. Predictive targetingSometimes, you can’t build the segments you want. At Kameleoon, we found 2 main reasons it can happen. 1. You know the attributes of the segment you want to build but you can’t identify the visitors in it. Example: One of our clients, Allopneus (the french leader of online tire retail) wanted to target “frequent drivers” with a special discount. They knew that such visitors have special characteristics, like vehicle type or size of tire used for example. But how do you know if an anonymous visitor meets these criteria or not? 2. You know the segment but can’t precisely define its attributes. Example: You run an e-commerce website, you want to target the indecisive visitors—those who are not sure they’ll buy or not, to try and influence them. But what are the characteristics of these visitors and how would you gather an exhaustive list of them? That’s where predictive targeting with machine learning algorithms kicks in. You tell it what you know, and it will crunch data, then calculate the probability of a visitor to belong in your segment. Granted it’s not perfect, but it allows you to get a LOT more than you could on your own. Aight, now that’s out of the way, let’s dive into the why you should absolutely segment your data.
Why segmenting your web analytics data is CRITICAL for your marketing effortsLet’s be honest, we’re quite lucky to be marketers today. We have a myriad of tools at our disposal, we can gather gigantic amounts of data on our audience, and it’s easier than ever to reach the masses. We also can have quantitative and qualitative insights as to which of our actions/campaigns work or not. Marketing where you throw money at something based on guts, or your boss decided to, or you always did it like that, or you have to spend your budget, is over (thank God). Now, you can test things, see if it works and for whom it does (or doesn’t). You can then be smart about where you invest your time and money. Segmentation is what allows us to do this. As Peep Laja would say, data segmentation is really about trying to answer questions or solve a problem. So here are 2 (voluntarily high-level, you’ll ask what’s relevant for you) questions—with examples to give an idea of what you could gain through segmentation.
Which of your marketing actions are the most effective?Do you want to see which channels brings the most traffic, thus where you should double your efforts? Create custom segments for traffic from search engines, email, social media, … . Or maybe you’re curious to see the traffic to your website from the specific keywords you targeted on AdWords? Create custom segments with these keywords. Or you’re running a promotional campaign for a specific page on your website, and you want to know how each of your actions is doing. Create an advanced segment using the page’s URL and dimension, you’ll then see in details the analytics for this page. Or maybe you’d wanna know if the average order is higher for traffic from specific sources? Or which referring websites send you the most valuable traffic? These examples are just a glimpse of what you can achieve with segmentation to get actionable feedback on your marketing efforts. It’s also extremely valuable to learn about your audience.
Who does what on your website?Whatever you do, don’t look at your audience as an undifferentiated whole. You would be leaving invaluable information on the table. We’re all different (no sh** Sherlock) and we use websites differently, have different interests, needs, and expectations. Quick example: new vs returning visitors (we’ll see how to build these 2 segments in Google Analytics later). You’ll see—and I bet you can easily wrap your head around it, they have very different behaviors and needs. They won’t land on the same pages, their focus won’t be on the same elements, they’ll look at different content… Or you could create a segment to study your highest converting visitors using goals, then refine it by location, traffic source, what content they look at, etc.. And with your learnings you can focus on your highest potential targets and invest in what gives you the most results. Curious about demographics and what people do depending on their age? Well, you can create segments for that too. Or maybe one of your current focus is SEO. How about creating segments with the keyword dimension to track people landing on your website from different keywords and see if you’re targeting the right ones? With these segments (which are just anecdotic examples, there thousands more you could play with) you’ll be able to understand who’s who, compare their behaviors and optimize your visitors experience consequently. Without segmentation, you’re shooting in the dark, AND you have a superficial knowledge of your audience. Resulting in failed campaigns, bad content, unhappy customers and money thrown away… Sounds like fun ? But let’s not be downers and be helpful instead. I’m going to help you build 3 segments in Google Analytics so you can get started now.
How to segment (and where to start)Like for anything, you HAVE to start simple if want to develop a sustainable practice or habit. You might get excited and want to dive deep right from the bat. I get it, and I’m like that too but what will happen is you’ll get discouraged, or you’ll mess up somewhere, and it won’t work, or you’ll feel like it wasn’t worth the effort, … Whereas if you start simple, you won’t have the excuse “it’s too much work, I don’t have the resources.” You’ll slowly get used to the jargon, the nuts and bolts, understand and discover the opportunities lying ahead. I chose Google Analytics to build our example segments as it’s free and accessible to everyone. It’ll be largely similar in your preferred tools and the principles apply anyway. We’ll see how to add “system” (already created in Google Analytics), import segments created by others and create one from scratch.
Adding a system segment: new visitors and returning visitorsFire up Google Analytics, go to your website dashboard. Click on “Add Segment.” Then select “System,” and write “new” in the search field. Check “New users,” then type “returning” and check “Returning visitors.” Click “Apply.” You’ll notice your 2 segments appear next to “All Users,” and in all the graph in analytics. Well done! You have your first 2 segments ready to be analyzed and built upon. You can now compare what new and returning do on your website, where they come from, how long they stay, what they convert on…
Wait, how do I remove a segment?Easy. Click the down arrow on your segment, and select “Remove.”
Import segments created by othersWhat if you could “steal” segments created by analytics grand master Avinash Kaushik? Well, you can! Why reinventing the wheel when you can benefit from the brilliance of others ? It starts the same way: click “Add segment,” but now you’ll then click on “Import from gallery.” You can browse through on your own, but I advise you to take the most popular “Occam”s Razor Awesomeness”, it’s a bit more advanced, but you’ll get 6 segments, 9 custom reports, and a dashboard. Also get the “New Google Analytics User Starter Bundle“, there some nice SEO segments 2 dashboards and custom reports. Then select the “view” you want them on, then click “create.” You’ll end up in the “Admin” tab of Google Analytics, and you’ll see the list of everything you just imported. Click on anything in the list to see what’s inside and favorite or write down what might interest you. To add see these new segments on your graphs, it’s the same thing we did with new and returning visitors above. “Add segment” > either scroll or write down its name in the search bar > “Apply” > Learn ?
Create a new segment from scratch: New Visitors from ReferralsSame start as for the 2 previous ways: “Add segment.” Then click “New segment.” You’ll see in the next window the different attribute categories you’ll build your segments from:
- Date of first session
- Traffic Sources
Turn your segmented data into actionsSegments are knowledge, but they can become weapons for your business. Insight is interesting only if you act on it. Otherwise, it’s like spending 5 hours cooking and then throwing it all away… Web personalization is dynamically showing to your segments targeted content (messages, offers, banners, emails, …) on your website, to create a highly relevant and individualized experience. Depending on the segments you created, your website will change itself to correspond to the characteristics of the visitors you targeted in said segments.
Pop-up for new visitors with SumomeI don’t need to explain what Sumome is if you don’t know them—or how to install it, you can always check them out. If you didn’t have Sumome installed, I’ll assume it’s done by now (it takes about a minute to do). Let’s think about what we wanna do. Who are we targeting? Our “New Visitors” segment. What tailored experience are you going to create to answer the needs of your segment? A pop-in with downloadable content, to get their email address and have them come back later thanks to your newsletter. Where on your website will it be on? On the homepage and article pages of the blog When and how frequently are you going to show it? When exit-intent is detected, every visit until they are subscribed. Open up Sumome, click on “List Builder.” Caveat, if you want to have access to what I’m gonna show you, you need at least the “Pro” version of “List Builder.” I’ll let you create a new pop-up and design it, as it’s not what we’re interested in here. When you’re done designing your pop-in, click on “Behavior.” Make sure you have “Smart” selected (= exit-intent). Choose the frequency you want and make sure the 2 following toggles are “on.” Hit “Save.” Now go in “Campaigns,” then in “Popups & A/B Tests.” Click “Add existing Popup.” Select the one you just created. Hit “Save.” Now go in “Display Rules.” Click “Add a show rule.” Choose “When A Visitor’s Page Views Are…” then “Fewer Than” and “3”. Hit “Save.” I guess I should have called this segment “New-ish Visitors” here, you’re free to play with the display rule ? Then make sure you plugged in your mail service in, and you have a contact list with an automated email ready to send the content you’ll give out. And you just have to toggle on your campaign, you’re good to go! A nice pop-up for your new-ish visitors, without bothering people already subscribed or not interested! Oh, and if you like Sumome and segmentation, Sean Bestor wrote an epic article on their blog you should read!
In-page text change for returning visitors with Kameleoon
At Kameleoon, we have 2 products, an A/B Testing tool, and a personalization solution. We’ll use the latter here. If you’re interested, don’t hesitate to ask for a 1-on-1 demo, or just take a look at our product.