# leocwlau

> Leading Edge Of Complex World Leadership Analytics Unlocked


## Posts

- [The Alert That Cries Wolf](https://leocwlau.com/2026/the-alert-that-cries-wolf/): An alerting system that cries wolf every Sunday trains its team to ignore the messenger. So when the real collapse fired, on time, the warning sat unread between two false alarms, and the crisis peaked before anyone knew. Alerts are a trust system, and trust is what false alarms spend.
- [Readiness Before Revenue](https://leocwlau.com/2026/readiness-before-revenue/): Every growth offer walks a line between service and betrayal. An upgrade pitched to a customer barely using the product trains them to read every message as a sales call. Four questions of adoption, readiness, value and timing decide which kind of offer is sent, because revenue waits for readiness.
- [Nobody Owns the Customer](https://leocwlau.com/2026/nobody-owns-the-customer/): Four systems hold four accurate fragments of the same customer, and together they describe a person who does not exist. The win-back email lands a day after the complaint. Here is why fragmentation is an ownership problem, and why no platform purchase can fix it.
- [The Tagging Plan Is Not the SDR](https://leocwlau.com/2026/the-tagging-plan-is-not-the-sdr/): Every analytics team maintains a meticulous SDR, and the reports still lie. The ledger records machinery, never purpose. A tagging plan starts with the user journey and justifies every data point before choosing a variable. Here is why that difference decides whether tracking is correct or commercially relevant.
- [The Oberon Debugger: From Workspace to API](https://leocwlau.com/2026/the-oberon-debugger-from-workspace-to-api/): Adobe Analytics' hidden Oberon debugger reveals the exact Reporting API 2.0 requests behind every Workspace visualisation. Instead of reverse-engineering JSON payloads, you can capture a working report and turn it into live dashboards, real-time alerts, and programmatic integrations that scheduled exports and bulk data feeds simply cannot deliver.
- [The Silent Battle for eVar7 in Adobe Analytics](https://leocwlau.com/2026/the-silent-battle-for-evar7-in-adobe-analytics/): A missing tagging dictionary is the real reason Adobe Analytics reports break across regions. The Core + Local allocation model reserves eVars, props, and events for global standards and local flexibility, so enterprise implementations survive the next acquisition, rebrand, or regional expansion.
- [AI Citation Visibility Is Finally Here](https://leocwlau.com/2026/ai-citation-visibility-is-finally-here/): Search visibility is no longer enough. New AI Performance reports in Bing Webmaster Tools show when and how your content is cited in AI-generated answers, across topics, intents, and citation share. It is the practical step publishers have needed to start optimising for the age of generative answers.
- [Hit, Visit, or Visitor: The Segmentation Scope That Changes Your Reports](https://leocwlau.com/2026/hit-visit-or-visitor-the-segmentation-scope-that-changes-your-reports/): Adobe Analytics segments live or die by one setting: scope. Hit, visit, and visitor containers each answer a different question, yet the same condition can return three completely different results. Choose the wrong scope, and your report silently lies. Learn how to pick the right one before adding any conditions.
- [When the Data Platform Decides to Own the Customer](https://leocwlau.com/2026/when-the-data-platform-decides-to-own-the-customer/): Databricks CustomerLake is not another CDP. It is a bet that the data platform that already stores everything now wants to decide who sees what, when, and why. Architecture, governance, and pricing are on its side. The real test is whether marketers will let an infrastructure vendor run the show.
- [Adobe vs Google Analytics: Which Fits Your Business Stage?](https://leocwlau.com/2026/adobe-vs-google-analytics-which-fits-your-business-stage/): The Adobe Analytics versus Google Analytics 4 debate is not about crowning a winner. It is about matching the platform to your organisation's maturity, scale, and ambition. Here is how to choose without falling into vendor tribalism.
- [Reimagining Marketing Leadership in the Age of AI](https://leocwlau.com/2026/reimagining-marketing-leadership-in-the-age-of-ai/): How must marketing leadership evolve when artificial intelligence becomes not merely a tool, but a force that transforms how we plan, execute, and measure our work? This article explores the shift from generative AI to agentic systems, the tension between democratisation and control, the challenge of maintaining brand differentiation, and what marketing leaders must remember as they architect human-machine collaboration in the age of infinite generation.
- [From Use Case to Organisational Capability](https://leocwlau.com/2026/from-use-case-to-organisational-capability/): In some recent discussions with people I met at conferences or during catch-ups, we came across a story that was repeated.
- [The Agentic Web: Beyond the Browser and Into the Functional Layer](https://leocwlau.com/2026/the-agentic-web-beyond-the-browser-and-into-the-functional-layer/): The internet is shifting from "page views" to "action completed," marking the death of the traditional browser for task completion. In this new Hybrid-Agentic ecosystem, AI agents navigate a website’s Functional Layer to execute complex tasks with mathematical precision, bypassing manual clicks and scrolls. As the Machine Economy emerges, businesses must adapt to serve both human experience and agentic utility. Explore how the Agentic Shift is rebuilding the digital world for speed, trust, and autonomy.
- [Beyond the Prompt - From Chatbots to Self-Hosted Agentic AI with Openclaw](https://leocwlau.com/2026/beyond-the-prompt-from-chatbots-to-self-hosted-agentic-ai-with-openclaw/): Moving from standard chatbots to Agentic AI with Openclaw offers a more autonomous experience by shifting toward independent, task-oriented intelligence. By self-hosting the system within a Docker container using Ollama, I established a sustainable, private environment that treats the AI as a separate individual with its own dedicated accounts, effectively avoiding expensive API token usage. Official documentation is often insufficient, where some information is missing or glossed over. While users might expect AI chatbots to be helpful guides, they often prove unreliable for technical troubleshooting, providing "thoughtfully incorrect" suggestions that require verification. Achieving a successful setup ultimately necessitates manual research, combining official documents with web crawling and human intelligence to resolve hurdles. Because automated tools and standard guides fall short, users must be prepared to manually intervene and perform their own technical verification to bridge the gaps in the provided instructions.
- [Automated Adobe Launch Verification with n8n](https://leocwlau.com/2025/automated-adobe-launch-verification-with-n8n/): Learn how to build an automated Adobe Launch verification crawler using n8n and Puppeteer to proactively detect missing or misconfigured tags across your website, preventing data loss with cost-effective, scheduled monitoring and Google Sheets reporting.
- [Time Measurement and Reporting in Adobe Analytics](https://leocwlau.com/2025/time-measurement-and-reporting-in-adobe-analytics/): Master Adobe Analytics time spent: backward calculation secrets, why last hits get 0s, and how to accurately measure process completion time with plug-in.
- [Understanding Attribution in Adobe Analytics](https://leocwlau.com/2025/understanding-attribution-in-adobe-analytics/): In today’s data-driven marketing, understanding the customer journey is key to optimizing ROI. Adobe Analytics’ attribution assigns credit to touchpoints, revealing channel effectiveness, and contribution from various customer behaviours. There are three approaches for dynamic multi-touch analysis: Marketing Channels for acquisition, eVars for preferences, and Attribution Models in the Analysis Workspace.
- [MarTech Maturity in the Lens of Digital Customer Lifecycle](https://leocwlau.com/2025/martech-maturity-in-the-lens-of-digital-customer-lifecycle/): None of us would like to chase technology. Setting up a MarTech maturity framework with a digital customer lifecycle and regular assessments could give us a clear direction on where to go.
- [From Adobe Analytics Learner To Champion](https://leocwlau.com/2024/from-adobe-analytics-learner-to-champion/): A journey from Adobe Analytics learner to champion over a decade covering learn, excel, share, and champion. Observe, think, and try are the trio in learning. Demand, curiosity, empathy, and sharing knowledge internally and externally are important factors in going from learner to champion.
- [Combining A/B Testing and Personalization](https://leocwlau.com/2024/combining-a-b-testing-and-personalization/): In today's digital landscape, personalization is key to enhancing user experiences and driving conversions. But how can we truly measure its impact? There are some approaches to merge A/B testing and personalization, unlocking the potential to deliver tailored experiences while evaluating effectiveness.
- [How to pass the Adobe Analytics Architect Exam](https://leocwlau.com/2024/how-to-pass-the-adobe-analytics-architect-exam/): Sharing my journey of preparing for and successfully clearing this challenging certification in one go. It has a bit of a showoff 😁, some references for preparation, and experience from the examination. If you aim to become an Adobe Analytics Architect, I hope this can offer some help to secure a one-time pass.
- [Recalculate Adobe Target Result](https://leocwlau.com/2024/recalculate-adobe-target-result/): Go beyond the basics and unlock powerful insights by integrating Adobe Target with Adobe Analytics via A4T, to analyze testing and personalization activities using a wide range of metrics. Addressing the limitation of the A4T solution by leveraging Data Warehouse and Data Feed to perform offline calculations using Python to perform Welch's t-test and proportion test.
- [Harnessing GenAI Chatbot to Generate A/B Testing Ideas](https://leocwlau.com/2024/harnessing-genai-chatbot-to-generate-a-b-testing-ideas/): A/B testing is vital for optimizing web page performance, but generating impactful ideas can be time-consuming. Using a GenAI Chatbot can streamline ideation, enable data-driven decisions, and boost web page conversions.
- [What Makes a Success: Delivery and Presentation](https://leocwlau.com/2024/what-makes-a-success-delivery-and-presentation/): It is an incredible honour to be invited to serve as a judge at the Digital Media Awards 2024, a privilege that I have a chance to look into some of the most innovative and impactful campaigns. It comes to me that there are two primary factors when I am reviewing entries and call them success or not, the capability to deliver and the capability to present. That can go further to personal and professional development as well.
- [Tab, Tab, Tab, And Website Traffic Analytics](https://leocwlau.com/2024/tab-tab-tab-and-website-traffic-analytics/): We are using tabs for web browsing a lot which can seriously impact the web analytics of the site: user behaviour across multiple tabs and windows, sharing cookies, isn't distinguished in data analysis. This behaviour affects accurate session tracking and attribution, complicating the understanding of user journeys. Solutions include improving website design and tracking, shifting towards user-centric analytics, and augmenting quantitative data with qualitative research.
- [Measuring Customer Journeys For Life Insurance In Asia Using Adobe Analytics](https://leocwlau.com/2024/measuring-customer-journeys-for-life-insurance-in-asia-using-adobe-analytics/): Wenzhou-Kean University collaborates with Hong Kong Baptist University, adopting Adobe Analytics to equip students with transborder skills in web analytics. Invited by Hong Kong Baptist University to share the caption with students from Wenzhou-Kean University attending an Adobe Analytics training conducted by HKBU to present Asia's insurance industry landscape, and how the Internet and digital tools enable better customer experiences, process efficiency, deeper engagement, and the measurement.
- [Can We Really Replace The Cookie?](https://leocwlau.com/2023/can-we-really-replace-the-cookie/): What are the concerns around HTTP cookies such as security and privacy issues, and the challenges of a cookieless environment and how it impacts brands' digital marketing efforts? We should be focusing on first-party data, enhancing the overall customer experience, using cross-channel analytics, marketing automation, and customer data platforms to adapt to the post-cookie era.
- [Tracking Adobe Experience Cloud ID With WebSDK](https://leocwlau.com/2023/tracking-adobe-experience-cloud-id-with-websdk/): The ECID is a unique identifier in Adobe Experience Cloud/Platform that helps recognize individuals and devices. While ECID is not available as a dimension in Adobe Analytics, capturing it as a custom dimension is desired for data linkage. The older AppMeasurement tracking allows assigning the ECID using the getMarketingCloudVisitorID() function. However, with the new WebSDK, the getIdentity command is used asynchronously, posing challenges in capturing and assigning the ECID. Workarounds involve retrieving the ECID separately and using custom actions with JavaScript Promises.
- [Converting event_list And post_event_list In Adobe Analytics Data Feed](https://leocwlau.com/2022/converting-event_list-and-post_event_list-in-adobe-analytics-data-feed/): The event_list and post_event_list columns are heavily encoded and impossible to use directly. Showing how to convert the encoded event_list and post_event_list columns in Adobe Analytics Data Feed into meaningful event names and split them into individual columns using Apache Spark to make the data more interpretable and easier to analyze.
- [User Retention By Days Of Access](https://leocwlau.com/2021/user-retention-by-days-of-access/): Instead of the traditional day since the last visit or cohort analysis, calculating user retention based on the number of days users access a portal or app over multiple visits provides a more comprehensive understanding of retention. Using PySpark with Adobe Analytics Data Feed to extract the required data, calculate the weekly and monthly access rates, apply time-decayed weighting for recent months, and visualize the results. The goal is to gain better insights into how often users engage with the platform over time.
- [Rebuilding Adobe Analytics Full Path Report With Spark](https://leocwlau.com/2021/rebuilding-adobe-analytics-full-path-report-with-spark/): The full path report is missing in the new Analysis Workspace in Adobe Analytics and rebuilding using Apache Spark. However, we can rebuild it using Apache Spark and data from the Adobe Analytics Data Feed, by reading the hit data, filtering valid page names, grouping by visit, ordering the page sequences, removing duplicates if needed, and counting the occurrences of each unique path.
- [Optimise Custom Dimension Usage In Adobe Analytics](https://leocwlau.com/2021/optimise-custom-dimension-usage-in-adobe-analytics/): Here is the second half of my sharing in the Adobe Experience Maker The Skill Exchange event about optimising custom dimension usage.
- [Optimising Click Tracking With Adobe Launch](https://leocwlau.com/2021/optimising-click-tracking-with-adobe-launch/): It was my honour to be invited to join an Adobe online event to share my experience using Adobe Analytics, the recording is available at Experience Makers The Skill Exchange.
- [Setting Up My Spark Lab](https://leocwlau.com/2021/setting-up-my-spark-lab/): When learning Spark and testing with small datasets, I can simply run a local Spark instance with the following command which essentially creates a local Spark instance using all cores.
- [Visitor ID, The Must-Have Custom Dimension](https://leocwlau.com/2015/visitor-id-the-must-have-custom-dimension/): Understanding your visitors' behavior is essential for optimizing your website and tailoring your marketing strategies effectively. With this must-have custom dimension, you'll have the power to analyze visitor patterns, track user engagement, and make data-driven decisions to enhance your online presence
- [The Analytics Cycle](https://leocwlau.com/2015/the-analytics-cycle/): Strategic thinking and careful planning are required when it comes to implementing analytics solutions. The analytics cycle follows a plan-do-check-act framework. It all starts with defining or refining your conversion goals, and aligning them with your business objectives. The analytics cycle is a continuous journey that requires adaptation and improvement.
- [Optimizing Paid Search Experience](https://leocwlau.com/2014/optimizing-paid-search-experience/): A seamless journey for visitors who click on your ads and land on your website is the key to conversion and should be meticulously designed to cater to customers' needs and intentions. Optimizing the experience to guide searchers towards your products, differentiate yourself from competitors, and ultimately drive conversions.
- [Testing Behavioral Differences Between Two Sets Of Visitor](https://leocwlau.com/2014/testing-behavioral-differences-between-two-sets-of-visitor/): In the realm of web analytics, built-in reports provide valuable insights, but they often lack statistical measures to determine the significance of the reported data. By comparing the means of two different sets of data, we can uncover valuable insights. Understanding the behavioural differences between visitor segments can be immensely valuable for optimizing marketing strategies, website design, and user experience. By leveraging statistical analysis, we can make data-driven decisions and unlock new growth opportunities.
- [Sharing Of HKU MSc (ECom&IComp) Program](https://leocwlau.com/2014/sharing-of-hku-msc-ecomicomp-program/): Are you ready to take your career to new heights and gain a competitive edge in the ever-evolving digital landscape? Whether you're looking for part-time study options or aiming to align your studies with your current role in digital marketing, the MSc (ECom&IComp) program at HKU might be the perfect fit for you.
- [Big Data Is Only About Variety](https://leocwlau.com/2014/big-data-is-only-about-variety/): When talking about big data, we usually talk about the three V's: volume, velocity, and variety. While volume and velocity are indeed important, variety plays a pivotal role in differentiating big data from traditional data management and analysis.
- [Connecting Analytics In Physical And Electronic Worlds](https://leocwlau.com/2014/connecting-analytics-in-physical-and-electronic-worlds/): By merging analytics from physical spaces and digital platforms, from retail stores to smart devices, data-driven insights can be harnessed to optimize customer experiences and drive business growth. With the advent of advanced technologies such as the Internet of Things (IoT) and location-based analytics, organizations now have the power to bridge the gap between the physical and digital worlds. By leveraging data collected from various touchpoints, businesses can gain a comprehensive understanding of customer behaviour, preferences, and trends.
- [Mobile App Vs Mobile Web](https://leocwlau.com/2014/mobile-app-vs-mobile-web/): Mobile app or not? While mobile apps have their advantages, opting for a mobile app may not always be the best approach, especially if it lacks regular updates or if a web counterpart already exists. Some key factors to consider when making this decision, such as background execution, hardware dependencies, offline usage, and overall readiness.
- [(The Emerging) User Experience](https://leocwlau.com/2014/the-emerging-user-experience/): User experience has been steadily gaining interest over time. It's an essential aspect of customer-centric approaches, including customer experience, digital experience, and more. The maturing of the internet has played a significant role in driving this shift. As the awareness of user experience rises, organizations are actively seeking ways to enhance the user experience across various touchpoints and throughout the entire user journey.
- [Digital Property Needs A Conversion Goal](https://leocwlau.com/2014/digital-property-needs-a-conversion-goal/): Simply focusing on traffic volume or downloads isn't enough, it is important to align your digital property's objectives with your organization's goals. By setting a conversion goal, you'll have a clear direction for optimizing your efforts and improving user engagement. By monitoring key performance indicators (KPIs) related to your conversion goal. This data-driven approach allows you to make informed decisions and enhance the productivity of your digital properties.
- [Analytics Outliner Detection](https://leocwlau.com/2014/analytics-outliner-detection/): Explores the challenges of monitoring multiple digital properties and offers a unique approach using standard scores for early problem detection. Discover how to improve tracking stability and react promptly to issues.

## Pages

- [Ollameter - The fuel gauge for your Ollama cloud allowance](https://leocwlau.com/tools/ollameter/): I am using Ollama for OpenClaw, n8n, and Codex, with a Pro subscription.
- [Services](https://leocwlau.com/services/): Practical data, MarTech and digital help for businesses that don’t have a big internal team.
- [Publications](https://leocwlau.com/publications/): The CFO asks why the paid media budget rose 30% while cost per acquisition barely moved. Sales complains that leads arrive with no context.
- [Adobe Analytics - A Champion’s Handbook](https://leocwlau.com/knowledge-hub/adobe-analytics-a-champions-handbook/): Every click, scroll, abandoned cart, and conversion holds intelligence that can transform how organisations compete, serve customers, and scale.
- [TagScanner - Find Out What Tracking Tools Are on Your Website](https://leocwlau.com/tools/tagscanner/): Many websites utilise tracking tools such as Google Analytics, Adobe Analytics, Facebook Pixel, and other marketing tags to collect visitor data.
- [Tools](https://leocwlau.com/tools/)
- [MarTech Maturity](https://leocwlau.com/knowledge-hub/martech-maturity/)
- [Knowledge Hub](https://leocwlau.com/knowledge-hub/): Welcome to the Knowledge Hub — a curated collection of frameworks, guides, and best practices designed to help enterprise marketing teams build, optimise, and scale their MarTech capabilities.
- [Self-Assessment Framework: Where Do You Stand?](https://leocwlau.com/knowledge-hub/martech-maturity/self-assessment-framework-where-do-you-stand/): Time required: 30-45 minutes for an individual assessment; 2-3 hours for a team workshop.
- [The Bottom Line: Connecting MarTech to Business Value](https://leocwlau.com/knowledge-hub/martech-maturity/the-bottom-line-connecting-martech-to-business-value/): Every CMO, CFO, and CIO has asked it: "We're investing millions in MarTech. What are we getting back?"
- [Enterprise Pillars: Governance & Scalable Architecture](https://leocwlau.com/knowledge-hub/martech-maturity/enterprise-pillars-governance-scalable-architecture/): For multinational organisations or any enterprise managing dozens of brands, the MarTech challenge is exponentially more complex than for a single-market business.
- [The 4 Stages of Growth: From Foundations to Personalization](https://leocwlau.com/knowledge-hub/martech-maturity/the-4-stages-of-growth-from-foundations-to-personalization/): Not every organisation starts at the same point, and that's by design. The MarTech Maturity Model is built on the understanding that growth is incremental.
- [The Concept of MarTech Maturity](https://leocwlau.com/knowledge-hub/martech-maturity/the-concept-of-martech-maturity/): In the current digital landscape, we are witnessing a strange paradox. Organisations have access to more sophisticated marketing technology than ever before.
- [Toolkit - Adobe Analytics](https://leocwlau.com/toolkit-adobe-analytics/): A simple web application to provide useful and consolidated information for Adobe Analytics administrators with lots of report suites to manage.
- [Contact](https://leocwlau.com/contact/): If you have questions or comments, feel free to drop me a message.
- [About Me](https://leocwlau.com/about-me/): Welcome to my website!
- [Privacy Policy](https://leocwlau.com/privacy-policy/): We may collect personal information from you when you voluntarily provide it to us while interacting with our blog site.

## Optional

- [Agent (MCP protocol)](websites-agents.hostinger.com/leocwlau.com/mcp)

[comment]: # (Generated by Hostinger Tools Plugin)