Dynamic Audience Segmentation

Segmentation is one of those concepts that sounds simple until you actually try to do it properly. Most teams start with broad categories like age, location, or gender, but the real insight comes when you start looking at how users act - how often they visit, how recently they engaged, how much value they bring, and which patterns naturally form across those dimensions. The goal of segmentation isn’t to label users, it’s to understand the structure of their behavior. That’s what data-driven segmentation methods allow us to do. K-Means, for example, helps you find natural patterns hidden in behavioral data. You decide how many groups you want to explore, and the algorithm does the heavy lifting, assigning each user to the cluster that best represents their behavior. It’s simple, efficient, and powerful for large datasets where you want to explore engagement trends without predefining who belongs where. When you need to see relationships instead of just results, hierarchical clustering becomes more useful. It builds a tree-like view showing which users are similar and where meaningful divisions exist. You don’t need to commit to a single number of segments. You can cut the tree at different points to explore how granular your understanding should be. It’s particularly helpful for moderate datasets where interpretability matters as much as precision. Then there’s DBSCAN, a method designed for reality - where user behavior is messy, irregular, and full of noise. Unlike K-Means, DBSCAN doesn’t assume clusters are neat or circular. It groups users by density, identifying natural clusters and automatically separating outliers. This makes it especially valuable for complex behavioral or clickstream data where some users behave in ways that don’t fit any conventional pattern. If you want something more business-focused and immediately actionable, RFM segmentation (Recency, Frequency, Monetary) remains a classic for a reason. By scoring how recently and how often users engage, and how much they contribute, you can pinpoint who’s loyal, who’s at risk, and who’s gone silent. It’s simple but effective for linking behavior to ROI and retention strategies. Finally, once you have meaningful segments, classification models can keep them alive. You can train a model to automatically assign new users to the right segment as data flows in, turning segmentation from a static exercise into a living system that adapts as behavior changes.

Hot take: If you're still segmenting customers solely by ARR and company size, you're leaving money on the table. After a painful realization, we completely overhauled our segmentation model: Our highest-paying enterprise customers weren't necessarily the most profitable or successful. Traditional segmentation missed these critical factors: Product usage patterns Growth potential (not just current spend) Support cost-to-revenue ratio Implementation complexity Use case maturity The result? We were over-serving some accounts and under-serving others based on flawed assumptions. Our new dynamic segmentation model includes: User adoption velocity Feature utilization depth Growth readiness score Technical maturity index Success potential metric The impact? 47% reduction in time-to-value 32% increase in expansion revenue More precise resource allocation Happier customers (and CS team!) A startup paying you $30K might have better product-market fit and growth potential than an enterprise paying $200K but struggling with adoption. Modern customer segmentation should be fluid, multi-dimensional, and focused on success potential, not just current value. What factors do you consider in your segmentation model? #CustomerSuccess #SaaS #GrowthStrategy #CustomerExperience

"What’s the ideal audience size for LinkedIn ads?"   ↑ An important question I get every few weeks, and it recently surfaced in the Fibbler community. Without the right audience, advertising is pointless. So what is the answer? For years, I defaulted to the classic rule of thumb: ~50,000 per audience segment, but 3 years ago, I stopped as it's misleading. I've been in and around over 1,000+ accounts now and have seen audiences from 1,000 people to 12m (y𝘦𝘴, 12 𝘮𝘪𝘭𝘭𝘪𝘰𝘯) achieve top 1% results. 𝐓𝐡𝐞 𝐨𝐧𝐥𝐲 𝐫𝐢𝐠𝐡𝐭 𝐚𝐧𝐬𝐰𝐞𝐫 - your audience size is your audience size, it's just tactic dependent. The question people 𝘴𝘩𝘰𝘶𝘭𝘥 be asking is "how do I know I've targeted the right audience?" The variables in targeting the right audience are: → Strategy (why this audience) → Segmentation (can you split it up) → Penetration (do you want new reach or to be frequent) → Tactics (brand tactics require looser audiences than activation) When thinking about 𝐬𝐭𝐫𝐚𝐭𝐞𝐠𝐲, the questions you need to ask are: → Who is this offer for? → Who is actually going to care? As LinkedIn is mainly B2B, I match the answer to these questions to targeting and I ALWAYS first start with the company. The 3 options: ↳ Company list (most accurate) ↳ Company size + industry (next best) ↳ Company size (for those with industry-agnostic solutions) Then I work on defining who the people we need to target are. Some variations we often use (there is no right answer here): Functions + Seniority + Skills + JT Exclusions Supertitles + JT Exclusions Functions + JT Exclusions Supertitles + Skills + Excl Function + Groups + Excl 𝐒𝐞𝐠𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 is the next thing to consider. The reasons you should segment: → Geography i.e. do you advertise to different time zones? → Internal structure i.e. having priority companies based on size? → Buying committee i.e. is MQL:SQL rate higher for certain functions? If the answer to any of these is yes, you should consider segmenting your audience pool by that variable. If you have a mass market product, then I'd suggest staying with as large an audience as possible. 𝐏𝐞𝐧𝐞𝐭𝐫𝐚𝐭𝐢𝐨𝐧 is achieved only by a very simple ratio Budget:Audience The higher the budget and narrower the audience, the higher the frequency. The lower the budget and wider the audience, the lower the frequency. You can control this by ↳ How many targeting variables you add ↳ How many AND layers you apply ↳ How many exclusions you appy ↳ How much budget you spend Finally, you need to consider 𝐭𝐡𝐞 𝐭𝐚𝐜𝐭𝐢𝐜 - in short, the most important point is how tight or loose you WANT to be with this targeting. Be looser with roles for brand awareness and tighter if you have say an incentivised offer. — Bottom line: Only segment for logic and understanding, not to satisfy a rule-of-thumb number. Your audience should be exactly as big (or small) for your company/goal - nothing more, nothing less.

The one thing that can ruin a brilliant marketing campaign. The wrong segmentation. Every great marketing campaign starts with the right audience. And the best B2B campaigns start with an audience agreement contract between marketing and sales. Yes, the teams must sign a contract. 😜 The classic, ‘let’s align’ dance. But in all seriousness, get the audience wrong, and you probably just wasted some time and money. The best thing to do is to have a source of truth for your audience segments. I recently saw a Common Room implementation that gave me an aha moment. Nurturing leads must be dynamic today. 👉 Old way: Nurture based on a single conversion events and a persona definition. 👉 New way: Nurture based on a set of always up-to-date signals, targeted at the right accounts and contacts with segments that update automatically as behavior changes, and plays that launch without rebuilding workflows from scratch." If you always have up-to-date audience segments that are learning in real-time, it’s going to be a lot easier to run simple brand and demand plays. One of my favorites is using LinkedIn engagement signals to retarget accounts and buyers. The key is to retarget them with ads not coming from a brand, but from people they trust, already know, and respect. Another good one is using first party conversion event data to enroll new or exciting contacts into a targeted email and ad campaign. Conversions like webinar attendance, downloading a research report, and checking out an interactive product demo. Have the segments ready to go, the trigger on stand by, and the play designed.  I like to think of this as a much better version of leader nurturing. The approach we’re taking at EasyLlama is thoughtfully and thoroughly working with our sales team members to pick the right accounts, the right contacts, and the right targeting. We’re using a combination of new accounts, existing accounts in our CRM, and prioritizing accounts based on signals of buying interest and intent. Tan Tran, Kenna Rooney, Caroline Shine, and Chris Edwards are working closely together. We're now meeting weekly and sharing ideas, questions, feedback to make sure we're engaging the right accounts, with the right messaging and offers, using the appropriate channels. Keeping accounts engaged using a targeted, specific, and multi-channel strategy is a must today. And when AI is powering segmentation and orchestration, it’s not about more automation, it’s about better judgment at scale. Educating people about their pain points, showing them examples of success, getting them to learn from customers, and using real, trusted experts is how you’re going to stay top of mind and create a differentiated buying experience that matters just as much as your product. 

🤯 The $95 Billion Question: How Do You Size the "X-Curious" Audience? Streaming content spend is set to top broadcast at $95 billion in 2025. With that massive investment comes a challenge: finding the next engine of growth. The focus has shifted from serving core fans to converting the adjacent "X-curious" audience. For example, platforms like Netflix and Crunchyroll are targeting the global "Anime-curious" segment, a strategy validated by the launch of dedicated 24/7 channels for this specific audience. The central strategic conflict: Is "X-curious" a real segment, or just trend-following (FOMO)? Answering this dictates budget allocation. You cannot size this kind of fuzzy market with simple Top-Down methods. My Approach to Sizing the Unknown-Unknown: **Identify Standalone Proxies**: Don't rely on generic industry data. Find specific, non-typical behavioral complements that signal genuine pre-existing affinity. This means looking at adjacent consumption that aligns with the content's theme Remember the Big Mac story when MacDonald's just got into big data? **Conduct Controlled Segmentation**: Before strategy is built, run small, self-contained audience prospecting experiments on ad platforms. This converts abstract affinity into hard, bottom-up metrics like a cost-per-acquisition or a conversion rate. **Triangulate & Segment by Motivation**: Use the real-world performance data to generate proprietary, customer-driven insights that identify the true drivers of the new prospective audiences. This is where you distinguish: 💡 The Genuinely Curious: They need education and onboarding (Growth via Content Strategy). 💡 The FOMO/Trend Driven: They need maintenance and hype to keep the category hot (Growth via Momentum and Hype). This process moves the segment from the "Unknown-Unknown" to a quantifiable "Known-Unknown." More importantly, it creates a mechanism to strategically deploy budgets for sustainable growth, rather than chasing a passing trend. What methods do others pull for sizing adjacent or "curious" markets? How do you distinguish between real market interest and a passing trend? #MarketSizing #GrowthStrategy #GoToMarket #AudienceSegmentation #Streaming #BusinessDevelopment

📊 𝗗𝘆𝗻𝗮𝗺𝗶𝗰 𝗔𝗕𝗖 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗦𝗲𝗴𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗶𝗻 𝗣𝗼𝘄𝗲𝗿 𝗕𝗜 An ABC analysis helps identify which 𝗰𝘂𝘀𝘁𝗼𝗺𝗲𝗿𝘀, 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀, 𝗼𝗿 𝗼𝘁𝗵𝗲𝗿 𝗱𝗶𝗺𝗲𝗻𝘀𝗶𝗼𝗻𝘀 drive most of your revenue or margin and which contribute the least, providing insight into their value and relative importance. This dynamic version shows one way to make the analysis interactive and flexible, but it can easily be adapted to different business contexts and needs. In this setup, users can: ✅ Choose the classification period (𝗟𝗮𝘀𝘁 𝟲, 𝟭𝟮, 𝗼𝗿 𝟮𝟰 𝗺𝗼𝗻𝘁𝗵𝘀) calculated backward from the selected month ✅ Adjust thresholds for 𝗖𝗮𝘁𝗲𝗴𝗼𝗿𝘆 𝗔 and 𝗖𝗮𝘁𝗲𝗴𝗼𝗿𝘆 𝗕 using sliders (you can also choose 80/20, which aligns with the Pareto principle) ✅ Choose whether to classify by 𝗥𝗲𝘃𝗲𝗻𝘂𝗲 or 𝗚𝗿𝗼𝘀𝘀 𝗣𝗿𝗼𝗳𝗶𝘁 ✅ Filter by 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗨𝗻𝗶𝘁 💡 Quickly answer questions like: – “Is a small group of Category A customers responsible for the majority of total revenue (the 20/80 principle)?” – “How dependent is our revenue on our largest (Category A) customers?” – “What share of all customers falls into Category C, the ones contributing the least to total revenue?” ⚙️ Setup highlights ▪️Disconnected table for 𝗔𝗕𝗖 𝗰𝗮𝘁𝗲𝗴𝗼𝗿𝗶𝗲𝘀 ▪️Parameters for 𝗰𝗮𝘁𝗲𝗴𝗼𝗿𝘆 𝘁𝗵𝗿𝗲𝘀𝗵𝗼𝗹𝗱𝘀 (𝗔 %, 𝗕 %) ▪️Disconnected table for 𝗽𝗿𝗲𝘀𝗲𝘁 𝗽𝗲𝗿𝗶𝗼𝗱𝘀 ▪️𝗦𝘂𝗽𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝗹 𝗰𝗮𝗹𝗲𝗻𝗱𝗮𝗿 used for Revenue, Gross Profit, and GM% calculations based on preset periods ▪️𝗗𝗔𝗫 𝗹𝗼𝗴𝗶𝗰 for ABC classification using cumulative revenue or gross profit % (and closest thresholds), combined with a derived visual-level filter This version was inspired by techniques from the 𝗗𝗮𝘁𝗮 𝗩𝗶𝘇 𝗙𝗼𝗿𝗴𝗲 𝗰𝗼𝗺𝗺𝘂𝗻𝗶𝘁𝘆 and discussions with Achmad Farizky and Gustaw Dudek 🙌 💬 How do you segment your customers? What approach works best?

🚀 What’s New for Data Cloud & Marketing Cloud? Dynamic Segments + Broadcast Flows unlock real-time, high-volume messaging. In the October 2025 Data Cloud update, two major features were introduced: Dynamic Segments and Broadcast Flows. In particular, "Dynamic Segments" suddenly appeared in the new segment creation screen, leaving many users wondering: --- "How do I use this?" 🤔 --- "What's different from a traditional segment?" 🤯 My latest article explains, as simply as possible, how these new features work, when to use them, and how to configure them step by step. Blog: https://lnkd.in/g2hSq9uU ============================== This combo unlocks powerful use cases for operational alerts, service notifications, and instant audience targeting. 🔍 Dynamic Segments — Key Points Only available inside Broadcast Flows (for now) Mix static + dynamic conditions - - - Static = fixed filter (e.g., Gender = Female) - - - Dynamic = filter changes at execution time based on API parameters No “Publish” step — always uses the latest saved version Supports up to 100,000 records per execution Count/Preview works, but you must temporarily replace dynamic parameters with static values 📡 Broadcast Flows — Key Points Segment-based version of On-Demand Flow Triggered via REST API or SubFlow Perfect for real-time bulk sends ⚠️ Limitations: - No Schedule  - No Debug - No Wait Element 🧩 Why This Matters By combining both features, organizations can: ✔ Send real-time operational alerts ✔ Notify customers instantly based on dynamic inputs ✔ Improve CX and reduce support cases ✔ Execute large-volume sends without prepublishing segments 📘 Example Use Cases 🛫 Airline gate changes sent instantly to affected passengers 🎢 Theme park weather/emergency alerts 🏠 Proactive service outage notifications 🔌 Utility maintenance alerts by postal code 🏁 Conclusion If your business needs fast, large-scale, dynamic messaging, Dynamic Segments + Broadcast Flows are true game changers. #Salesforce #MarketingCloud #DataCloud #MomentMarketer #MarketingChampion #MarketingChampions 🔗 Connect with me on LinkedIn: https://lnkd.in/gkeKyShH