Applied Data Analytics for Impact

Introduction CRM data is rarely stored in a single, analysis ready table. Customer details, interactions, transactions, and campaigns are usually split across multiple datasets, often with inconsistent keys and repeated records. As a result, analysts frequently encounter inflated metrics, broken joins, and confusing totals. These issues are not caused by SQL itself, but by how joins, aggregations, and deduplication are applied. Getting these fundamentals right is essential for trustworthy CRM analytics. Most CRM metrics depend on combining tables correctly: counting unique customers attributing interactions to campaigns summarising behaviour over time If joins are misaligned or duplicates are not handled deliberately, metrics quietly drift. Dashboards may look correct but tell the wrong story. Intermediate SQL skills allow analysts to express clear analytical intent , not just retrieve data. Thinking before writing SQL Before writing queries, I focus on three questi...

Introduction Many dashboards fail quietly because the transformation logic is scattered, manual, or undocumented. Analysts often clean data “just enough” to make visuals work, without designing transformations that are structured, repeatable, and auditable. In BI environments, this leads to fragile reports. A small schema change breaks refreshes. A new column introduces inconsistencies. Over time, confidence in the numbers erodes. Power Query sits exactly at this fault line between raw data and analytics. Power Query is not just a prep tool. It is an ETL layer embedded inside BI workflows . When transformations are well designed: data refreshes become predictable logic is transparent and reviewable models remain stable as data evolves downstream DAX stays simple When they are not, analysts compensate with complex measures and manual fixes, increasing technical debt. Structured transformations reduce that debt. Intermediate technical explanation: how to think abou...

 Introduction CRM datasets are rarely analysis ready. They often contain duplicated records, inconsistent text fields, missing values, and dates stored in multiple formats. While tools like Power BI and Excel can handle some cleaning, analysts frequently face a point where repeatable, scalable data preparation is required. This is where Python becomes essential. The challenge isn’t just cleaning data once. It’s designing a process that works reliably as new CRM data arrives. Poor data quality directly impacts: customer counts segmentation accuracy campaign performance metrics downstream modelling and forecasting If cleaning logic lives only in ad hoc steps or manual fixes, errors reappear quietly over time. A Python based approach allows analysts to formalise assumptions, document decisions, and reproduce results consistently. In CRM analytics, this reliability is foundational. Intermediate technical explanation: how to think about CRM data cleaning Before...

 Introduction CRM data is rarely clean by default. Records are entered by different teams, updated at different times, and stored across multiple tables that were never designed for analytics. As a result, analysts often struggle with inconsistent metrics, confusing filters, and dashboards that break as soon as requirements change. In most cases, the root issue isn’t the visuals or the calculations. It’s the data model underneath . Why this problem matters A poorly designed data model leads to: Double counted customers KPIs that change unexpectedly when filters are applied Complex DAX written just to “fix” modelling issues Dashboards that are hard to maintain or scale In CRM analytics, where insights often drive engagement strategy, segmentation, and forecasting, unreliable numbers quickly erode trust. A clean data model acts as the foundation that keeps analytics consistent, explainable, and reusable. Modelling concept Fact vs Dimension thinking A reliable...

  Introduction Despite the rise of modern analytics tools, Excel remains one of the most widely used tools in data analysis. What separates an average Excel user from a highly effective data analyst is not the number of functions they know, but how they structure data, solve problems, and support decision-making. In this post, I share a practical approach to building advanced Excel skills that reflect real-world analytics work rather than exam-style knowledge. Why Excel Still Matters for Data Analysts Excel is often the first tool used to: explore unfamiliar datasets validate assumptions perform quick analyses communicate insights to non-technical stakeholders Understanding how to use Excel well improves analytical thinking, regardless of the tools used later. Thinking in Tables, Not Worksheets Experienced analysts treat Excel as a structured data tool, not a canvas. Key habits include: using Excel Tables consistently keeping raw data separate from analy...

 Introduction Many analytics issues are not caused by complex models or incorrect logic. They come from quiet data quality failures that go unnoticed until results are questioned. Missing values, duplicate records, unexpected spikes, or invalid dates can all distort insights. When these checks rely on manual review, they are inconsistent and easy to forget. This is why analysts need automated data quality checks , even for simple datasets. When data quality checks are informal or ad hoc: dashboards lose credibility analysts spend time firefighting instead of analysing errors propagate into forecasts and models trust in analytics declines Automation turns data quality from a reactive task into a governed process . It ensures that datasets meet basic standards before they are used for reporting or decision making. In CRM and similar analytical domains, this consistency is critical. Intermediate technical explanation: what data quality really means At an analy...

  Introduction Dashboards often receive the most attention in analytics projects, but the quality of insights depends far more on the underlying data model than on visual design. Poor modelling choices can lead to misleading KPIs, inconsistent metrics, and a lack of trust in reporting. In this post, I explore how raw CRM data can be transformed into meaningful KPIs through thoughtful data modelling decisions. The focus is on structure, clarity, and alignment with real decision making needs rather than tool specific features. Why KPIs Fail Before They Reach Dashboards Many KPI issues originate long before reporting begins. Common causes include: Ambiguous metric definitions Inconsistent grain across tables Mixing transactional and aggregated data Unclear relationships between entities KPIs derived from poorly structured fields When these issues exist, even well designed dashboards struggle to provide reliable insight. Understanding Raw CRM Data Structure CRM...