Driving Impact with Segmentation & Personalization in Adobe Analytics

Transcript

Hi everyone. Thanks so much for joining today’s skills exchange session. I’m so excited to have you here. My name is Monica White. I’m a Senior Customer Success Manager here at Adobe. I’ve been here for over 10 years now and I’ve been part of the Customer Success org since 2020. I’m certified in both analytics and CJA and I’m passionate about helping customers get the most value out of their Adobe solutions, especially by highlighting new features and functionality as our products evolve. Now, before we dive into the content, I always like to share a little bit about myself outside of work because we’re all more than just our job titles, right? I am based in the Seattle area and when I’m not working with customers, I love being active. I’m really into long distance running. I play CoRec soccer with my husband and we try to get out hiking wherever we can, especially with all the amazing trails we have in here in the Pacific Northwest. You got to take advantage. We also love to travel, but when I’m relaxing at home, you usually find me hanging out with my husband Stu, our dog Kiko and our cat Hopper. So that’s a little bit about me. Today we’ll cover why segmentation and personalization matter, creating and managing segments, spotlight segment IQ, implementing personalization strategies and cover a real world example. Then we’ll close with a Q&A session. Feel free to add your questions in the chat bot or chat pod while we’re going though. So the benefits of segmentation and personalization. Let’s start with a high level overview of what segmentation and personalization are and why they matter.

Segmentation is a filter for your data. It allows marketers and analysts to isolate subsets of visitors based on behaviors, attributes and events, enabling more target analysis and action. It’s foundational to understanding your audience and tailoring experiences that resonate. Personalization builds on segmentation by delivering relevant content based on user behavior and profile data. Personalization improves engagement, conversion rates and overall ROI. In short, segmentation helps you understand your audience and personalization helps you act on that understanding. In analytics, segmentation is powered by two key tools which we will be looking at today. Segment builder, this is your canvas for creating segments and segment manager. Once segments are created, segment manager helps you organize and operationalize them. This creates a collaborative workflow that data scientists and marketing analysts can build on and refine segments for their specific needs or save them in a shared library.

How to build a segment. Creating and using segments in Adobe Analytics is straightforward and with the segment builder in Adobe Analytics workspace, there are three main ways of creating segments. From scratch and segment builder, from pre-built segments and key touch points and follow up reports. It is a powerful tool that allows you to define roles using dimensions, metrics and containers. Segments can be saved, shared and reused across reports and dashboards.

Additionally, real time segmentation is supported and segments can be applied on the fly. There are several ways to access segment builder but I will focus on showing you two for now. First, analytics in the top navigation, select components, go to segments and then you can start building your segment by starting with the title and the description from there.

Second, in analysis workspace, if you open a project, click new or existing and then select new segment and you can build it from there. Once started, you can create a segment of users, for example, who viewed a product page but didn’t convert or a segment that came in through a campaign and purchased something as a result. This segment can then be analyzed to understand why these users did or didn’t make a purchase and what can be done to encourage them to convert. One of the most powerful things about Adobe segment builder is how intuitive it is once you understand the architecture behind it. It’s built on a simple but flexible model that uses containers to define the scope of your segment logic.

Let’s dig a little further into understanding segment builder. In segment builder, everything is organized into three nested containers, visitor, visit and hit. Think of them like Russian nesting dolls. Each one fits inside the next and each defines a different level of granularity. I’ll break down how Adobe analytics organizes data using a simple but relatable analogy, a visit to the museum. The visitor is the outermost container. Think of the visitor container as the person visiting the museum. This is a unique individual who might come once or return multiple times throughout the year. In analytics, this represents a persistent user ID like a cookie or CRM ID. That lets us track behavior across multiple sessions and devices. A visit is the middle container. Now, each time that person walks into the museum, that’s a visit. Maybe they spend an hour exploring exhibits, stop at the cafe, browse at the gift shop. All of that activity is grouped into one session. In Adobe analytics, a visit container captures all hits that happen during a single session, usually ending after 30 minutes of an activity. Now, a hit is the most specific. Every time a visitor stops and looks at an exhibit, reads a plaque or scans a QR code, that’s a hit. It’s the most granular level of data. On a website, this could be a page view, a button click or a video start. Each of these interactions is tracked individually within that session.

This container model gives you a ton of flexibility. You can build segments that reflect a user’s full history, zoom in on specific sessions or isolate individual interactions. And because they’re nested, you can combine them to create really precise definitions. If you go into the components tab, click on segments. After clicking on a segment, you will land on the segment page. Click on the add option located in the left panel, and then you’ll see visit, visitor or hit. Select what makes most sense for your segment and move forward.

So now that we’ve covered how segments are structured using containers, let’s take it a step further and talk about how we can use those containers to understand the order of user behavior. That’s where sequential segments come in. Sequential segments are incredibly useful when you want to analyze how people move through your site, not just what they did, but the order in which they did it. They allow you to define a sequence of actions or interactions. Continuing with the museum visit example, you might want to find visitors who entered the website and viewed a specific exhibit page, then added tickets to their visit or added tickets to their cart to visit that museum exhibit in person. Or maybe you want to identify users who started a checkout and then didn’t complete it. In Segment Builder, you use the then operator to define this kind of logic. It tells Adobe Analytics to look for one action that happened after another. And just like within regular segments, you can apply this logic at the visitor, visit or hit level depending on how broad or narrow you want to go. So in short, sequential segments help you go beyond static attributes and start mapping out the journey. There’s a great way to bring more nuance and storytelling into your analysis. Segments from Fallout Analysis. Now that we’ve seen how to build segments from scratch in Segment Builder, let’s talk about a few other ways you can create in use segments in Adobe Analytics. One quick and powerful method is through Fallout reports and analysis workspace. If you’re analyzing a funnel and notice a drop off at a certain step, you can right click and instantly create a segment of users who dropped off or continued. It’s a great way to turn insights into action without leaving your analysis. The great thing is it opens up Segment Builder and pre populates the sequential segments exactly matching the touch point you need and can give all you need to do is give a title and description then save from there to report on later or share with your team. You can also start with pre built segments. Adobe provides a library of common ones like mobile users, new visitors or cart abandoners and the component tab on the right or on the left side. These options make segmentation more flexible and accessible no matter where you are in your analysis workflow. Segment Manager. Now let’s discuss Segment Manager and why it’s such an important part of building a scalable segmentation workflow. You access by going into analytics, click the components tab, then segments and the top navigation. From there, you’ll see it display on the top tab, tag, share, delete, rename, approve, copy or export to CSV. There are four key reasons to lean into segmentation manager. First organization and governance. It helps you keep your segments clean and structured. You can tag, approve and track where segments are being used, whether it’s in dashboards, alerts or shared projects. Second is for collaboration. Segment Manager supports a shared library so teams can reuse and refine each other’s segments. This avoids duplication and ensures consistency across reporting and personalization. Third is access control. You can manage who can view or edit each segment, which is especially helpful when working with sensitive data or cross functional teams. Four, last but not least, is scalability. As your segmentation strategy grows, Segment Manager becomes the central hub for managing hundreds of segments across teams and use cases. So while Segment Builder is where you define logic, Segment Manager is where you operationalize it, making sure your segments are usable, shareable and accessible. Now let’s get into a segmentation workflow. Here’s an example of a best practice workflow for creating and managing segments. Work with your teams to build out an established workflow if you don’t already have one. First is planning segments. Outline steps and best practices to follow when planning your segments. Build your segments. Build and edit your segments for use in all analytics capabilities. Third, tag your segments. Tag segments for ease of organization and sharing. Four, approve segments. Approve segments to make them canonical. Four, apply segments. You can apply segments directly from a report from the segments rail. Six, share segments. Share your segments with the intended audience and other analytics tools and Adobe Target and other tools in the Adobe Experience Cloud. Seven, filter segments. Filter by tags, owners and other filters like mine, shared with me, favorites and approved. Finally, mark segments as favorites. Marking segments as favorites is another way or as a favorite is another way to organize them for ease of use. Now that we have an understanding of how to build and manage segments, I want to quickly spotlight a feature that’s often overlooked. Segment IQ. I’ve noticed in many of my conversations with analytics owners and in reviewing analytics adoption scores that this feature is frequently underutilized. One of the main reasons there’s a common misconception that segment IQ is part of customer journey analytics, and it’s actually quite the opposite. Segment IQ is exclusive to analytics. So what does it do? Segment IQ uses machine learning to uncover statistically significant differences between two segments. It’s a fast, automated way to identify what makes one audience behave differently from another without having to manually dig through dozens of metrics. For example, let’s say you have a segment of users who frequently visit your product page but never convert. Segment IQ can compare them to converters and surface the key behavioral and demographic differences. That insight can help you tailor your messaging offers or even site experience to better engage that group. It is an incredibly powerful tool for identifying high value audiences and optimizing for ROI. And it’s already built into the platform. So if you haven’t explored it yet, it’s definitely worth a look. One of the key features within segment IQ is the segment comparison panel, and it’s a powerful one. This tool lets you compare an unlimited number of segments and uses machine learning to surface the most statistically significant differences between them. It automatically scans all dimensions and metrics you have access to. So you don’t have to guess where to look. What makes it especially useful is that it helps you uncover the traits that are driving performance. For example, it can show you what your highest converting audiences have in common or where different segments overlap in ways you might not expect. This is incredibly valuable when you’re trying to refine and personalize experiences or prioritize which audiences to invest in. So here’s a demo. Let’s say you’re trying to understand why one group of users converts at a much significantly higher rate than another. You create two segments. One, a June campaign conversion, and segment B, everyone else. Use the segment comparison panel in analytics. It runs a machine learning analysis across all available dimensions and metrics like traffic sources, device type, time on site, campaign ID, and more. It might even reveal that segment A, your converters are significantly more likely to use desktop devices and spend more than three minutes on a product page. While segment B, non-converters might skew towards mobile users coming in from social media with the shorter time on page. These insights help you pinpoint what’s working and what’s not. Maybe your mobile experience needs improvement or your social campaigns are attracting less qualified traffic. Instead of manually digging through dozens of reports, the segment comparison panel surfaces differences for you so that you can act faster and smarter. So we’ve just looked at the segment comparison panel which helps you uncover key differences between audiences. Now let’s look at another powerful feature, comparing segments in fallout. This feature lets you overlay multiple segments directly in a fallout visualization to compare different audiences and how they move through a funnel. It’s a great way to spot where drop-offs happen and how behaviors vary between groups without needing to build separate reports.

You can visually compare conversion paths across segments side by side and it helps you identify where specific audiences are struggling or succeeding. It’s especially helpful for testing hypotheses like whether mobile users drop off earlier than desktop users or if campaign traffic behaves differently than organic traffic. Here’s how it works. Open up any fallout visualization in Analysis Workspace, drag in your funnel steps like product view, add to cart or checkout, and then drag in the segments you want to compare. Maybe new versus returning visitors or mobile versus desktop. Adobe will overlay those segments in the same fallout chart so that you can instantly see how each group flows through that funnel. You’ll be able to spot where one group is dropping off more than another and that insight can guide everything from UX improvements to campaign targeting. Now that we’ve explored how to build and analyze segments, let’s talk about how segmentation actually powers personalization and how we can bring that to life across channels. Segmentation is the foundation of personalization. Once we’ve identified meaningful audience groups, the next step is activating those segments to deliver tailored experiences.

As we’ve just learned, segments can be created in analytics using Segment Builder, pre-built segments, or from visualizations like cohort, flow, or Venn diagrams. And then these help you define your audiences based on behavior, attributes, or journey patterns. Once created, segments are activated, meaning they’re made available to downstream tools for targeting and personalization. There are several key destinations for activation. One is Adobe Target for web and app personalization, A-B testing, and recommendations. RT-CDP for real-time profile enrichment and cross-channel activation. And other channels like email platforms, ad networks, and more. For the delivery, these segments are then used to deliver personalized experiences, whether it’s a homepage tailored to a user’s browsing behavior or an email triggered based on recent activity. Let’s say a user frequently visits pages related to your specific product category. You can use Adobe Target to highlight those products on their homepage. Or with RT-CDP, you can trigger a personalized email within three days of a key action, like using a feature or banning a cart. Personalization strategies can be rule-based, behavior-driven, or even predictive. And with tools like Target and RT-CDP, you can test, optimize, and scale those experiences across channels. Real-world examples and best practices are crucial for understanding the practical applications of segmentation and personalization. I’m sure you all have heard that Adobe likes to call ourselves customer zero. So this example is of how Adobe leveraged segmentation to increase engagement on our experience league platform. Adobe’s internal RT-CDP proof of concept used segmentation to trigger skill level-based nurture emails, resulting in higher engagement. The new personalized homepage on experience league increased click-through rates by nearly three times. Some best practices to follow when working on that personalization start with creating clear business questions, testing and iterating, and democratizing those insights via shared dashboards. For example, RT-CDP email nurture showed significant improvement in engagement compared to the control group. Sharing curated dashboards can help reduce data overwhelm and ensure that everyone in your organization has access to valuable insights. Before we wrap up, I want to leave you with a few key takeaways from today’s session. First, a basic understanding of what segments are and how to build them. Second, leveraging segment IQ for deeper insights. Third, applying segments to drive personalization.

I hope this session gave you a clearer picture on how segmentation fits into your analytics and personalization strategy and how to start using these tools more confidently. And with that, I’d love to open it up for any questions you might have. So much value there, Monica. Thanks so much for sharing your perspective and for joining us today. Awesome. Yeah, thanks for having me. Excited to be here. Great. Well, now it’s time to open up for the Q&A. If you’ve got questions about segments, audiences, or personalization strategy, please go ahead and ask. So our first question today is how does building segments in analytics translate into segments created in AEP for activation into the CDP, AJO, or CJA? Absolutely. The beauty of these segments between the tools is that they can be shared between any of these solutions to ensure that the campaigns and the details being transferred between the campaigns are aligned. So yes, absolutely. Okay. Our second question, can segments be passed between analytics and Marketo engage? Yes, they can. So right now there is not a direct or not a way to directly share though. There are connectors that can be activated. Between us though, with Marketo, it is being expanded all the time. So it could become easier, but yes, there is a way to do it. It is not quite as seamless as between Target and CDP though at this time. Great question. Here’s the one that I know as a lot of people that I talked to share this one as well. Is it possible to share segments between analytics and Adobe Target? Yes, absolutely. Analytics and Adobe Target are best friends and segments are shared between friends. We ensure that those campaigns, like I said before, they’re nice and seamless and they’re very well integrated.

Yeah, I agree. The A4T connector does a lot of great work between each other. So I highly suggest that everyone, if you’re not leveraging that, please do. Absolutely. Yes. Okay. Here’s a question. Can you build an abandoned cart program, especially if you do not have a direct connection into your e-commerce backend? Yes, you absolutely can create an abandoned cart program and have that create a connection, even if it’s not directly on the e-commerce backend. So yes, you can with certain API connectors, I believe. Adam, do you have additional insights there that you would like to share? Yeah. So there’s different ways you can do that. So first of all, there is through an API. If you want to leverage your API, that’s okay. If you want more real-time data or you don’t have access to an API, you can manually program that inside of analytics to grab that information from the cart. It won’t be, again, a native connector, but you can do that. Just take development work to make it work. Absolutely. And definitely connect with your Adobe team. They can help support and provide additional details. Experience League also has some great resources and communities that you can reach out to. There’s for different companies and people that have done similar things. I completely agree with that. Yeah. Our next question. In sequential segment, is it possible to get the traffic within the sequence only and do not capture it before or after? Absolutely. Yes. You can isolate or expand as much as you want with these segments. That’s really the beauty of Segment Builder. Like I said, you can get super granular to traffic within an hour in a very specific section of your site. So, yes, absolutely.

Okay. So, this next one, I love this question. It’s something that I wish I had way back when I was a young analyst working on things. So, can you use segment IQ with dimensions and or metrics? Absolutely. So, the beauty of segment IQ is it covers all dimensions and metrics. And instead of needing to manually compare, it now uses machine learning to identify anomalies between all of the dimensions and metrics. And it can, yeah, identify statistically significant anomalies or things that you might not even thought of in your manual review. So, absolutely, use segment IQ and it’ll capture all of those dimensions and metrics. And, yeah, like I said, I love that feature. I want everyone to be using it and leveraging it. Yes. Hands down. It’s seriously a game changer in a lot of people’s business practice. So, again, I definitely agree. Put that spotlight even more on it. Yeah. Yes. Exactly. Yes. So, here is a great question and it’s a real world example I think would be good to discuss. We create a unique ID for users in Marketo that have done an action on a website. Can we use that to help share the segments between Marketo and analytics? Yes. Especially since you have that unique ID, that is perfect to be able to share between Marketo, engage in analytics. Leveraging that single ID, but that’s exactly what you need to make that connection between analytics and Marketo. So, you have the perfect situation to track those users and those whatever actions that you’re interested in. And you can even expand what their current actions that you’re tracking now. So, yes, absolutely. Great question.

And here’s another one for SegmentIQ. Is it possible to use SegmentIQ to help identify spikes in visits to determine why, where, what is causing the anomaly? Yes, absolutely. That is one of the main awesome things about SegmentIQ. It identifies those spikes, anomalies, what could be causing it, and yeah, it could dig in and find something that you might not have been aware of at all. So, yes, definitely. Great. Well, these are awesome questions. Let’s keep going. If you have more, please share. The next one, if we use the segmentation slash segment IQ to figure out what groups are engaged with the site, would segment IQ show any groups that don’t exist in related to that segment? It might not pull in something completely unrelated because it is comparing the segments that you’ve identified. But if that makes sense, it could capture things that might be related to that one. Does that make sense? Yeah, it does. It might be good also to refer to the experience league. There’s a lot more information that you can really glean from your specific use case. So, whoever asked that, highly suggest you go there and read up more about that. You should be able to find a lot more rich content for SegmentIQ for your question. Yes. Yeah. Okay. Here’s a great one. Can I create a segment and send to CDP to create an audience? Absolutely. That is another one of the key beauties between these, Segment Builder and Analytics and CDP. You can build your segments however you want them in Analytics and then activate them in real time in CDP. So, yeah, they’re also a match made in heaven. Similar to real-time CDP. Yes, they are. I love how real-time CDP does share with everything within Analytics, CGA. It’s a very powerful tool. It really is. I’m glad that is there. Yes. Definitely leverage those awesome integrations and the way that they seamlessly work together and talk to each other to activate in real-time for your personalization strategies. Yes. Okay. Our next question. When you share a segment from Analytics to Target, is it shared in real-time or is there a lag? So, between Analytics and Target, I do believe there might be a lag. I don’t think it is incredibly significant. I think the sharing between Analytics and RT CDP is faster. I do believe there is a minor lag between Analytics and Target though. Again, Experience League is a fantastic resource for granular details like that and reach out to your Adobe team too. They can help support. And Adam, do you happen to have any additional insight to share there? Yeah. Yeah. It’s not immediate. So, as soon as it happens, it’s not going to show up immediately into Target or vice versa, but it does go quickly. Yeah. I was always impressed by how fast it does, but it really depends on what your business case is. So, if you have a specific use case, go into Experience League, see if that actually beats your requirements. And then, you know, if there is something else that needs to be done, they should have directions on how to do that. Yeah, absolutely. Thank you. Okay. Here is a great one. How do I decide whether to build a segment at the hit, visit or visitor level? That is an awesome question and I love that one. So, it depends on what your analysis goal is for the hit level, that’s for the really granular actions, for the visit level, that’s all behaviors within a specific session. And the visitor level, that is the long-term behavior across sessions. And that could be whatever timeframe you’re looking into. If you’re interested in visits within a week, then you can set it in that timeframe. If you’re interested in visits within a month or three months. So, again, it depends on the analysis goals that you have, but those nested containers make it amazing to dig in and get exactly the information that you want. Yeah, I agree. It is to use one or the other. So, I highly agree with that. Yeah, exactly. And even going back to the museum example, remember a hit is every single QR code scanned or painting visited, the visit is every action within that person’s day at the museum. And then visitor level, picture that as one person who really loves the museum and visits multiple times a month, and it captures all of their visits to the museum. Yeah, I love that analogy. So, just remember, sometimes if you are capturing a lot of data, it gets hard to navigate and weed through to get actual information that you can actually make decisions on. So, sometimes the hit level will be too much. It’s great, valuable information. But again, you really need to look at what your use case is. Yeah, you might not need to know every single painting someone looked at. You might just want to know what gallery they were interested in. You might not need to know every snack they bought at the cafe, just that they visited the cafe. Yeah. Well, I think we have time for one or two more questions. Awesome. What is the best way to validate that a segment is working as intended? That is a great question. So, leveraging the pre-test feature in Segment Builder will help you preview how many hits, visits, and visitors match whatever segment that you’re testing. So, you can see it ahead of time to make sure it’s hitting your criteria before you apply it to whatever visualization that you’re looking into building. Okay. Well, that’s all great. So, that’s unfortunately all the time we have. Thank you so much for joining us, Monica. It’s been a pleasure. Thanks for having me.