Segment or Die: The Art Of Data Clustering

Table of Contents


If you’re a marketer, chances are you’ve heard of data clustering. It’s one of those things that people either love or hate – a bit like Brussels sprouts. But whether you’re a data-crunching aficionado or a marketing newbie, there’s no denying that data clustering can be a game-changer for your marketing reports.

Now, I know my marketing peers have a bit of a love-hate relationship with this technique. Some swear by it, while others roll their eyes and mutter something about “overcomplicating things.” But trust me, if you’re looking to take your marketing strategy to the next level, data clustering is worth exploring.

So sit back, relax, and grab a cup of Earl Grey tea, because, in this blog, we’re going to dive deep into data clustering. We’ll explore what it is, how it can benefit your marketing strategy, and how your favorite tool, Maya, can help you implement it quickly and easily.

What is data clustering?

Data clustering is like having a personal assistant who helps you tidy up the mess of data that accumulates over time. It’s like the Marie Kondo of data management – sorting through the overwhelming pile of information into smaller, more organized groups. And let’s be real, who doesn’t need a little decluttering in their life?

In technical terms, clustering is a classic data mining technique based on machine learning that divides ​groups of abstract objects into classes of similar objects. This technique can help make sense of large, complex data sets and reveal insights that may be difficult to see otherwise.

Once the data has been clustered, it can be analyzed and visualized in a variety of ways to reveal patterns and insights that may not have been apparent before. For example, in marketing, clustering can be used to group customers based on their behavior, allowing marketers to tailor their messaging and campaigns to specific groups of customers. Or it can even be used to group search queries based on the users’ intent, to track the performance of SEM and SEO campaigns more effectively.

But, why is data clustering important to marketers?

The Problem

Imagine you’re a marketer, spending hours poring over reports and data sheets. You’re trying to make sense of all the numbers and metrics, hoping to find some insight that will help your company grow. But all you see is data in aggregate: total revenue, average page views, overall customer satisfaction, and more. It’s like looking at a giant, blurry picture with no clear details.

This is the problem that many marketers face. They have access to tons of data, but they struggle to turn it into actionable insights. And the achingly tiny percent of the time they do segmentation, it seems to stop at basic segments like new vs. returning visitors. While these basic segments can provide some initial insights, they often fail to capture the nuances and complexities of customer behavior, making it difficult to develop targeted marketing strategies.

The Solution

That’s why clustering is super important for marketing teams. When you segment your data, you’re essentially breaking it down into smaller, more manageable pieces. Instead of looking at the big picture, you can zoom in on specific groups of people, behaviors, or actions, and then drill down further. This allows you to see patterns, trends, and insights that you might have missed otherwise.

For example, let’s say you’re looking at your revenue or goal performance report. It takes just a trivial amount of effort to add some default segments like Paid Search Traffic and Non-Paid Search Traffic. This simple step can give you deeper insights into where your revenue is coming from and which channels are driving the most conversions.

But the real glory comes from going beyond the default segments. Because let’s face it: default segments are created to appeal to everyone, and there’s no such thing as “everyone”. You are unique, and so is your business. You need a segmentation strategy that’s tailored to your specific goals, challenges, and target audience.

By clustering your data, you can gain a more nuanced understanding of your customers and their needs. You can identify segments with a higher likelihood of conversion or customer loyalty. You can tailor your marketing messages and campaigns to specific groups, increasing your chances of success.

How can Maya help you segment your data efficiently?

Let’s face it: data clustering is like a jigsaw puzzle that’s missing a few pieces. You know the potential insights are there, but it takes forever to piece everything together. And when you’re dealing with massive amounts of data that are constantly changing, it’s like trying to solve that puzzle while riding a unicycle.

But don’t panic just yet. Maya is here to save the day! With Maya’s help, you can segment your data quickly and efficiently, giving you an analytical picture of your data that makes sense. Whether you’re a seasoned data expert or a newbie, Maya offers a range of tools and features to help you quickly and easily segment your data based on any relevant combination of data points…

And when it comes to setting up your Maya account, creating clusters is one of the most important steps you’ll take. Clusters are like your trusty sidekick in the world of data, helping you break down your information into bite-sized chunks that are easier to digest.

So, what exactly can you cluster within Maya? Well, the answer is pretty much anything. Maya’s clustering functionality is flexible and customizable, which means you can create clusters based on any combination of data points that is relevant to your business. For example, you could cluster campaigns to see which ones are driving the most conversions. Or you can cluster landing pages to identify which ones are performing the best and which ones need improvement. The possibilities are endless, and the more you customize your clusters to your specific needs, the more insights you can gain from your data. 

So, let’s dive into what elements you can cluster within Maya and how can you do it using your own custom business rules.

Campaigns Clustering

When it comes to Campaigns Clustering in Maya, the possibilities are pretty much endless. You can group them based on a variety of factors, such as campaign or ad group name, type, or channel (think SEM vs. Facebook). Or even by key metrics like impressions, clicks, and cost. And don’t worry, Maya won’t judge you for wanting to group your campaigns by which ones are the biggest money pits (we’ve all been there). In this way, you can actually track which of them consume most of your budget and ensure they will have a good return in results.

A quick note here: that is essential for your marketing analysis in general. To get the most out of clustering, it’s essential to maintain consistency in how you name your elements. In this case, your campaigns. This means keeping things organized within the advertising platforms themselves, as well as in your tracking templates (aka UTMs). That way, you can easily identify which campaigns belong in which cluster without having to decipher a series of random letters and numbers.

For example, let’s say you want to create a cluster of campaigns that are all aimed at achieving a similar marketing objective, such as driving new customer acquisition (aka “Reach”) or retargeting existing customers (aka “Remarketing”). 

You can apply a name rule to your cluster with just a few clicks, depending on the naming template your team uses, to distinguish reach from remarketing campaigns:

Create new campaign cluster, based on scope (reach vs remarketing)

It’s important to note that when you create a new cluster in Maya, you’ll see a preview table on the right side of your screen. This table uses a random subset of your data to quickly show you the outcome of your clustering rules. This feature is incredibly useful for debugging and optimizing your cluster before deploying it for the entire dataset. Keep in mind that the ELT process takes a few hours to complete for the full dataset, but with the preview table, you can apply your changes and immediately see the results.

Clusters Preview UI

Budget Clustering

Budget Clustering refers to an upcoming feature in Maya that will allow you to segment your available marketing budget, make predictions per budget group, and allocate it more efficiently. You can use your campaign clusters in your budget clustering rules to make them even more accurate. This will allow you to make better decisions about how much money you want to spend on each campaign cluster and make adjustments. Please note that this feature is currently under development and will be available in Maya soon.

Conversions Clustering

Continuing, you can also cluster every possible conversion you need to analyze for your marketing strategy. With Conversions Clustering in Maya, you can group conversions based on various factors, such as conversion name, ID, or service. Maya extracts the default conversions from the Sources you have connected, as well as the custom ones your team has created on any advertising platform. This means you can group all conversions from different platforms and group them in one cluster for a comprehensive analysis.

But that’s not all. Conversions Clustering can also help you debug your website and event tracking. By grouping conversions based on your custom conditions, you can easily identify any issues with your website tracking and optimize your events for better accuracy.

Furthermore, you can create a custom conversion funnel that applies to any customer journey you may need to monitor. Whether it’s for a simple one-step checkout process or a multi-step lead generation form, you can create a conversion cluster that reflects your desired customer journey and track the performance of each step.

And the best part is, you can use different conversion clusters in your reports to serve different analysis needs. Whether you want to track overall conversion performance or analyze the effectiveness of a specific campaign or channel, Conversion Clustering in Maya has got you covered.

For instance, let’s say you need to group all of your transactions in a single Conversion Cluster, however, you use a different name to describe it on every platform:

  • On Google Analytics, it’s “Transaction”
  • On Google Ads, it’s “Purchase”
  • On Facebook, it’s “subscription_completed”
  • On Quora, it’s “acquired”

You may love switching between tabs and adding manually every single conversion, but with Maya maybe we have a better way to do it. You simply create a rule, where a conversion contains any of the above names, and set the value to “Main Conversion – Purchase” (#magic):

Create Conversions clusters based on their name or service

This makes your workflow faster and more efficient, leaving you with more time to focus on other important aspects of your marketing strategy.

Landing Pages Clustering

Continuing with the clustering extravaganza, let’s talk about Landing Pages Clustering. With this feature, you can group your landing pages based on a variety of criteria, including domain, URL slug, and even the content of a specific URL path level or total slash count. 

Let’s take an ecommerce site for example. The URL structure is like  You can use Maya’s landing page clustering to group all the landing pages that contain products of the same subcategory, creating a rule for the url_path_level_2 attribute:

Create landing pages cluster using url path and slash counts, or domain

By clustering your landing pages, you can get a better understanding of how different groups of pages are performing and identify areas for improvement. And the best part? You can use your landing page clusters in Maya’s default reports, such as the SEO, the Analytics, or the Speed Report. So not only are you improving your website’s user experience, but you’re also optimizing for speed and performance. Talk about killing two birds with one stone!

Keywords, Queries & Word Groups

In a world where people can’t spell, use strange synonyms, and write questions like Yoda, keeping track of search terms can be a nightmare. But with the right Clustering, your team can finally make sense of this madness. By grouping similar keywords, queries, and word variations into clusters, you can analyze your search data more efficiently and extract valuable insights that would otherwise be lost in the digital abyss.

Word Groups

But beware! Any value you add to your Word Groups is a “phrase match”, meaning that if you add the value “ads” to a group, Maya will group not just “ads” but also “pads”, “lads”, “dads” and so on. To avoid errors, use the “Excluded” functionality to specify words you don’t want to be included in a group. With Word Groups, you can ensure your data is accurately categorized and analyzed.

Create a word group with the typing variations of your brand name

Let’s start with Word Groups, the building blocks of Query and Keywords Clustering. With Word Groups, you can categorize different words or phrases, such as your competitors, or different spelling variations, into one group. These groups can then be used for Query & Keywords Clustering, allowing you to analyze search terms based on any criteria you need.

Imagine you need to group all the spelling variations of your company’s brand name for reporting purposes. Instead of adding a custom rule for each variation to all of your clusters, you can simply update your existing Word Group and all your reports will be updated in real-time. This saves you time and hassle, leaving you free to focus on analyzing your data.

Keywords and Queries Clustering

So, you’ve already formed your word groups? Excellent! Now it’s time to put them to use in your Keywords and Queries Clusters.

The process is pretty straightforward once again. You create custom rules and conditions for each cluster, but with the added magic of word group matching. This means you can easily add new variations of words or phrases to your word groups and they will be automatically included in your clusters. No more manually adding every query!

Create a query cluster by using your word groups

But what’s the real value of all this segmentation, you ask? Well, for starters, it saves you time and provides valuable insights into your users’ behavior. You can now track a whole group of similar search terms instead of manually accumulating every metric. This is a game-changer for performance teams looking to understand their audience better and create more targeted campaigns.

With Keywords & Queries Clustering, you can answer important questions like: How did our branded keywords perform last month? Are there any increases in impressions for ETL-related keywords? Should we start targeting and bidding on keywords that have the phrase “data warehouse” in them? These insights help you make informed, data-driven decisions that can significantly impact your business.


As a marketer, you know that data is king, but with so much of it, it can be tough to see the forest for the trees. That’s where data clustering comes in, allowing you to quickly identify the most valuable insights and make informed decisions based on your marketing data. Luckily, Maya makes data management dead simple and affordable, with a range of powerful features designed to help you get the most out of it.

So what are you waiting for? Sign up for Maya now and take your marketing data to the next level. And if you still have questions, our awesome team is always here to help with a personalized demo. Still not sure about how it works? Book a demo with our awesome team to show you around!


Data clustering is a data technique that helps you organize large amounts of data into smaller, more manageable groups.

Data clustering can help marketing teams to gain a deeper understanding and a more holistic view of their large amount of data, to make well-informed decisions based on metrics.

With your Maya account, you’ll be able to segment all of your available data into meaningful clusters, based on specific characteristics, performance, or context. Read more about how to work with Clusters here.

Table of Contents

Get Started with Maya today!