<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Aman Aggarwal on Medium]]></title>
        <description><![CDATA[Stories by Aman Aggarwal on Medium]]></description>
        <link>https://medium.com/@aman.aggarwal83?source=rss-468957d144da------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*rdmGVE1Dy0phzdF8IUv0fg.jpeg</url>
            <title>Stories by Aman Aggarwal on Medium</title>
            <link>https://medium.com/@aman.aggarwal83?source=rss-468957d144da------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 02:05:36 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@aman.aggarwal83/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Mastering SQL Window Functions: A Practical Guide]]></title>
            <link>https://medium.com/@aman.aggarwal83/mastering-sql-window-functions-a-practical-guide-38c32d69cbd4?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/38c32d69cbd4</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[business]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Fri, 30 Jan 2026 16:50:40 GMT</pubDate>
            <atom:updated>2026-01-30T16:50:40.985Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oDM5ki_fdBficmOiDWN6nA.png" /></figure><p>If you’ve been working with SQL for a while, you’ve probably encountered scenarios where you need to perform calculations across rows while still maintaining the individual row details. Maybe you wanted to calculate running totals, rank items within groups, or compare each row to an aggregate value. This is where window functions shine.</p><p>Window functions are one of SQL’s most powerful features, yet they often remain underutilised because they can seem intimidating at first. In this guide, I’ll demystify window functions and show you how they can simplify complex queries and unlock new analytical possibilities.</p><h3>What Are Window Functions?</h3><p>A window function performs a calculation across a set of rows (a “window”) that are related to the current row. Unlike aggregate functions with GROUP BY, window functions don’t collapse rows — they preserve them while adding calculated values.</p><p>Think of it this way: regular aggregate functions smash your data together into summary rows, while window functions let you peek at neighboring rows or the whole dataset while keeping every row intact.</p><p>The basic syntax looks like this:</p><pre>function_name() OVER (<br>    [PARTITION BY partition_expression]<br>    [ORDER BY sort_expression]<br>    [ROWS or RANGE frame_specification]<br>)</pre><h3>The Main Types of Window Functions</h3><h4>1. Ranking Functions</h4><p>These assign ranks to rows based on specific criteria.</p><p><strong>ROW_NUMBER()</strong> — Assigns a unique sequential number to each row:</p><pre>SELECT <br>    employee_name,<br>    department,<br>    salary,<br>    ROW_NUMBER() OVER (PARTITION BY department ORDER BY salary DESC) as row_num<br>FROM employees;</pre><p>This gives each employee a number within their department, ordered by salary.</p><p><strong>RANK()</strong> — Assigns ranks with gaps for ties:</p><pre>SELECT <br>    student_name,<br>    score,<br>    RANK() OVER (ORDER BY score DESC) as rank<br>FROM exam_results;</pre><p>If two students tie for 1st place, the next student gets rank 3 (not 2).</p><p><strong>DENSE_RANK()</strong> — Like RANK(), but without gaps:</p><pre>SELECT <br>    product_name,<br>    category,<br>    revenue,<br>    DENSE_RANK() OVER (PARTITION BY category ORDER BY revenue DESC) as rank<br>FROM products;</pre><p>With DENSE_RANK(), if two products tie for 1st, the next one gets rank 2.</p><h4>2. Aggregate Window Functions</h4><p>You can use familiar aggregate functions (SUM, AVG, COUNT, MIN, MAX) as window functions:</p><p><strong>Running Total:</strong></p><pre>SELECT <br>    transaction_date,<br>    amount,<br>    SUM(amount) OVER (ORDER BY transaction_date) as running_total<br>FROM transactions;</pre><p><strong>Moving Average:</strong></p><pre>SELECT <br>    date,<br>    sales,<br>    AVG(sales) OVER (<br>        ORDER BY date <br>        ROWS BETWEEN 6 PRECEDING AND CURRENT ROW<br>    ) as seven_day_avg<br>FROM daily_sales;</pre><p><strong>Percentage of Total:</strong></p><pre>SELECT <br>    region,<br>    sales,<br>    sales / SUM(sales) OVER () * 100 as pct_of_total<br>FROM regional_sales;</pre><h4>3. Value Functions</h4><p>These let you access values from other rows:</p><p><strong>LAG()</strong> — Access previous row’s value:</p><pre>SELECT <br>    date,<br>    price,<br>    LAG(price, 1) OVER (ORDER BY date) as previous_price,<br>    price - LAG(price, 1) OVER (ORDER BY date) as price_change<br>FROM stock_prices;</pre><p><strong>LEAD()</strong> — Access next row’s value:</p><pre>SELECT <br>    date,<br>    temperature,<br>    LEAD(temperature, 1) OVER (ORDER BY date) as next_day_temp<br>FROM weather_data;</pre><p><strong>FIRST_VALUE() and LAST_VALUE()</strong> — Get first or last value in the window:</p><pre>SELECT <br>    employee_name,<br>    department,<br>    salary,<br>    FIRST_VALUE(salary) OVER (<br>        PARTITION BY department <br>        ORDER BY salary DESC<br>    ) as highest_salary_in_dept<br>FROM employees;</pre><h3>Understanding the PARTITION BY Clause</h3><p>PARTITION BY divides your result set into groups (partitions), and the window function is applied separately to each partition. It’s like GROUP BY, but it doesn’t collapse rows.</p><p>Without PARTITION BY, the window function operates on the entire result set:</p><pre>-- Running total across all rows<br>SUM(amount) OVER (ORDER BY date)<br><br>-- Running total per customer<br>SUM(amount) OVER (PARTITION BY customer_id ORDER BY date)</pre><h3>Understanding Frame Clauses</h3><p>Frame clauses (ROWS or RANGE) define which rows within the partition are included in the window:</p><pre>-- Current row and 2 preceding rows<br>ROWS BETWEEN 2 PRECEDING AND CURRENT ROW<br><br>-- All rows from start to current<br>ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW<br><br>-- Current row and all following rows<br>ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING<br><br>-- Centered window (2 before, current, 2 after)<br>ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING</pre><p><strong>ROWS</strong> operates on physical row positions, while <strong>RANGE</strong> operates on logical ranges of values — ROWS is more commonly used and usually what you want.</p><h3>Wrapping Up</h3><p>Window functions unlock a new level of analytical power in SQL. They let you perform complex calculations that would otherwise require self-joins, subqueries, or application-level processing. Once you get comfortable with the syntax, you’ll find yourself reaching for them constantly.</p><p>Start simple — try adding ROW_NUMBER or a running total to one of your existing queries. As you get comfortable, explore more advanced patterns like moving averages and year-over-year comparisons. Before long, window functions will become an essential part of your SQL toolkit.</p><p>What complex queries have you simplified with window functions? Share your experiences in the comments below!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=38c32d69cbd4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why VLOOKUP is Old News: Meet XLOOKUP and Copilot in Excel]]></title>
            <link>https://medium.com/@aman.aggarwal83/why-vlookup-is-old-news-meet-xlookup-and-copilot-in-excel-26c9168af6f5?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/26c9168af6f5</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[excel]]></category>
            <category><![CDATA[vlookup]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[business]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Mon, 01 Sep 2025 20:34:33 GMT</pubDate>
            <atom:updated>2025-09-01T20:46:13.936Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sZrgEnPGRm1zC9333NcSfA.png" /></figure><p>For more than three decades, <strong>Excel</strong> has been the backbone of data analysis, reporting, and business decision-making across industries. From financial models and sales reports to marketing dashboards and quick ad-hoc analyses, Excel remains one of the most widely used tools in the world.</p><p>But even with its dominance, one function in particular — <strong>VLOOKUP</strong> — has carried both fame and frustration. While it helped millions of professionals link and retrieve data, it also became infamous for its quirks and limitations.</p><p>Today, Microsoft has not only given us a far better alternative in <strong>XLOOKUP</strong>, but it’s also ushering in the next era of productivity with <strong>Copilot in Excel</strong>.</p><h3>Why VLOOKUP is Outdated</h3><p>If you’ve ever worked with VLOOKUP, you know the struggle:</p><ul><li><strong>One-directional search</strong> → It only works when your lookup column is on the left. Need to look to the left? VLOOKUP can’t help you.</li><li><strong>Column index fragility</strong> → If your table changes, formulas break because VLOOKUP relies on fixed column numbers.</li><li><strong>Match mode confusion</strong> → The default approximate match (TRUE) often produced incorrect results when users forgot to add FALSE.</li><li><strong>Performance bottlenecks</strong> → On large datasets, heavy use of VLOOKUP slowed everything down.</li><li><strong>Poor readability</strong> → Nested or complex VLOOKUP formulas became difficult to debug or maintain.</li></ul><p>It worked, but let’s be honest: it was clunky, error-prone, and not built for modern data challenges.</p><h3>XLOOKUP vs VLOOKUP: What’s the Difference?</h3><p>Microsoft created <strong>XLOOKUP</strong> as a direct replacement for VLOOKUP (and even HLOOKUP). It’s easier, cleaner, and far more powerful.</p><p>Here’s why it’s better:</p><ul><li><strong>Flexible lookups</strong> → Search left, right, up, or down.</li><li><strong>Dynamic referencing</strong> → No need to count columns; you reference ranges directly.</li><li><strong>Exact match by default</strong> → Reduces accidental errors.</li><li><strong>Built-in error handling</strong> → Use IFNOTFOUND instead of wrapping with IFERROR.</li><li><strong>Cleaner formulas</strong> → Easier to read, easier to maintain.</li></ul><h3>A Quick Example</h3><p>Imagine you have a simple dataset of products and prices:</p><pre>+-----------+--------------------+---------------+<br>|  Product  |       Price        |      ID       |<br>+-----------+--------------------+---------------+<br>| Laptop    |        800         |      1        |<br>| Monitor   |        200         |      2        |<br>| Keyboard  |        100         |      3        |<br>+-----------+--------------------+---------------+</pre><p><strong>Goal:</strong> Find the price of a <em>Monitor</em>.</p><ol><li><strong>Using VLOOKUP:</strong></li></ol><pre>=VLOOKUP(&quot;Monitor&quot;, A1:B4, 2, FALSE)</pre><ul><li>You must specify column index <strong>2</strong> (Price column).</li><li>If another column is added before “Price,” this formula breaks.</li></ul><p><strong>2. Using XLOOKUP:</strong></p><pre>=XLOOKUP(&quot;Monitor&quot;, A1:A4, B1:B4, &quot;Not Found&quot;)</pre><ul><li>No column counting needed.</li><li>Cleaner, safer, and easier to maintain.</li></ul><h3>Copilot in Excel: The AI Leap Forward</h3><p>Just when users were getting comfortable with XLOOKUP, Microsoft introduced <strong>Copilot in Excel</strong>, powered by generative AI.</p><p>With Copilot, Excel is evolving from a <strong>tool you program</strong> into a <strong>tool you collaborate with</strong>. Imagine being able to:</p><ul><li><strong>Generate formulas</strong> by describing your problem in plain English.</li><li><strong>Summarise large datasets</strong> without manually writing pivot tables.</li><li><strong>Create charts and dashboards</strong> by asking for insights.</li><li><strong>Explain formulas step by step</strong> so you can actually learn as you work.</li></ul><p>This changes the game. Instead of spending hours tweaking formulas or debugging lookup errors, you can focus on insights and decisions.</p><h3>Why This Matters</h3><p>Excel is not going away anytime soon. Despite the rise of specialised analytics platforms and BI tools, Excel remains the <strong>common language of data</strong> in businesses worldwide.</p><ul><li>Finance teams still build models in Excel.</li><li>Marketing teams still prepare campaign performance reports.</li><li>Operations teams reconcile numbers.</li><li>Analysts still rely on Excel for quick data exploration.</li></ul><p>But the way we use Excel is evolving:</p><ul><li><strong>From manual functions (VLOOKUP)</strong> → to <strong>smarter functions (XLOOKUP)</strong>.</li><li><strong>From formula-heavy workflows</strong> → to <strong>AI-assisted analysis (Copilot)</strong>.</li></ul><h3>Final Thoughts</h3><ul><li><strong>VLOOKUP</strong>: A classic, but outdated and fragile.</li><li><strong>XLOOKUP</strong>: Modern, intuitive, and built for today’s Excel users.</li><li><strong>Copilot</strong>: The future of Excel, turning data analysis into a conversation.</li></ul><p>If you’re still relying on <strong>VLOOKUP</strong>, now is the time to switch. Embrace <strong>XLOOKUP</strong> today — and get ready for an <strong>AI-powered Excel era</strong> where Copilot helps you work smarter, not harder.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=26c9168af6f5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SQL for Data Cleaning: Simple Queries for Better Insights]]></title>
            <link>https://medium.com/@aman.aggarwal83/sql-for-data-cleaning-simple-queries-for-better-insights-05f1ef822f81?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/05f1ef822f81</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-cleaning]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Mon, 25 Aug 2025 20:14:00 GMT</pubDate>
            <atom:updated>2025-08-25T20:14:00.545Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dD4f7_odq-z8Aqx7uzhciQ.jpeg" /></figure><p>If you’ve ever worked with raw data, you know it’s <em>never perfect</em>. Messy data sneaks in through missing values, inconsistent formats, duplicates, or strange outliers. Before you can analyse it, you need to <strong>clean it</strong>.</p><p>While Python and R often get the spotlight for data wrangling, <strong>SQL is still the most powerful first step</strong> for cleaning data at the source. In this blog, we’ll walk through essential <strong>SQL techniques for data cleaning, </strong>with examples you can apply directly to your projects.</p><h3>1. Remove Duplicates</h3><p>One of the most common and essential data cleaning steps is <strong>removing duplicate records</strong>. Duplicates not only inflate results but can also lead to misleading insights if left unchecked.</p><p>Instead of bluntly deleting everything, you can use the ROW_NUMBER() window function to precisely identify and remove duplicates while keeping the <em>first valid record</em>.</p><pre>WITH ranked AS (<br>    SELECT <br>        id,<br>        email,<br>        ROW_NUMBER() OVER (<br>            PARTITION BY email   -- define what makes a row &quot;duplicate&quot;<br>            ORDER BY id          -- keep the lowest id as the original<br>        ) AS row_num<br>    FROM customers<br>)<br>DELETE FROM customers<br>WHERE id IN (<br>    SELECT id FROM ranked WHERE row_num &gt; 1<br>);</pre><p>This gives you full control over which record stays (e.g., the earliest, the latest, or the one with the highest value in a column).</p><h3>2. Handle Missing Values</h3><p>Missing values (NULL) are unavoidable in real-world datasets. If not handled properly, they can break calculations or cause incomplete analysis. In SQL, functions like <strong>COALESCE()</strong> or <strong>CASE WHEN</strong> help you replace nulls with default or meaningful values.</p><pre>-- Replace NULL phone numbers with &#39;Not Provided&#39;<br>SELECT <br>    id,<br>    COALESCE(phone, &#39;Not Provided&#39;) AS phone<br>FROM customers;</pre><p>The goal is to decide whether to <strong>fill, replace, or ignore</strong> missing values based on your business logic.</p><h3>3. Standardise Text Data</h3><p>Inconsistent text formatting (extra spaces, random casing) makes joins unreliable and summaries messy. Standardisation ensures uniformity across your dataset.</p><pre>-- Trim spaces and standardize casing<br>SELECT <br>    TRIM(name) AS clean_name,<br>    UPPER(city) AS city_upper,<br>    LOWER(email) AS email_lower<br>FROM customers;</pre><p>Functions like TRIM(), UPPER(), and LOWER() help make text data consistent and analysis-ready.</p><h3>4. Fix Data Types</h3><p>Data stored in the wrong format like numbers as text or dates as strings can block calculations and aggregations. Always cast columns to the right type.</p><pre>-- Convert string date to DATE type<br>SELECT <br>    CAST(order_date AS DATE) AS clean_date<br>FROM orders;<br><br>-- Convert string numbers<br>SELECT <br>    CAST(price AS DECIMAL(10,2)) AS clean_price<br>FROM products;</pre><p>Proper data types make queries faster, safer, and easier to work with.</p><h3>5. Validate &amp; Enforce Constraints</h3><p>Rather than constantly cleaning, you can <strong>prevent bad data from entering</strong> in the first place using database constraints.</p><pre>-- Example: emails must be unique<br>ALTER TABLE customers <br>ADD CONSTRAINT unique_email UNIQUE (email);<br><br>-- Example: price must be positive<br>ALTER TABLE products <br>ADD CONSTRAINT chk_price CHECK (price &gt; 0);</pre><p>Constraints act like data quality guards at the database level.</p><h3>6. Normalise Date &amp; Time Fields</h3><p>Dates often arrive in multiple formats, making time-based analysis painful. Normalizing them into structured components makes reporting easier.</p><pre>-- Extract year, month, and weekday<br>SELECT <br>    EXTRACT(YEAR FROM order_date) AS year,<br>    EXTRACT(MONTH FROM order_date) AS month,<br>    TO_CHAR(order_date, &#39;Day&#39;) AS weekday<br>FROM orders;</pre><p>This allows consistent grouping and filtering across reports and dashboards.</p><p>Data cleaning may not be the flashiest part of analytics, but it’s the foundation of trustworthy insights. By mastering these SQL techniques like removing duplicates, handling missing values, standardising formats, fixing data types, and validating with constraints, you can transform raw, messy tables into reliable datasets. The cleaner your data, the more confident you can be in every report, dashboard, or machine learning model that follows. Think of SQL as your <strong>first line of defence</strong> against bad data, it saves time, reduces complexity, and ensures you’re always working with the truth.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=05f1ef822f81" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Time Metrics in Web Analytics: A Complete Guide]]></title>
            <link>https://medium.com/@aman.aggarwal83/understanding-time-metrics-in-web-analytics-a-complete-guide-8cd9211764f0?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/8cd9211764f0</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[time]]></category>
            <category><![CDATA[web-analytics]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Mon, 11 Aug 2025 06:23:49 GMT</pubDate>
            <atom:updated>2025-08-11T06:23:49.777Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XMxFW4worAhNNv72l8zSHw.jpeg" /></figure><p>Accurately measuring how users spend time on your website is essential for optimising content, improving user experience, and driving business goals. However, time-based metrics in web analytics — such as <strong>Average Engagement Time</strong>, <strong>Average Session Duration</strong>, <strong>Average Time on Page</strong>, <strong>Time Spent per Visit</strong>, and <strong>Total Time Spent</strong> — can be confusing. This guide explains each metric, how it is calculated, and when to use it.</p><p>In this blog, we’ll break down what each of these metrics really means, how they’re calculated, when they matter, and how they differ across analytics platforms like Google Analytics and Adobe Analytics.</p><h3>1. Average Engagement Time</h3><h4>Definition :</h4><p>Average Engagement Time captures the active time users interact with your site, including scrolling, clicking, typing, or keeping the browser tab in focus.</p><h4><strong>Calculation :</strong></h4><pre>Average Engagement Time = Total Engaged Time ÷ Total Sessions (or Users)</pre><h4><strong>Why it matters :</strong></h4><p>This metric distinguishes between passive page loads and actual user interactions, revealing genuine engagement levels.</p><h3>2. Average Session Duration</h3><h4>Definition :</h4><p>Average Session Duration measures the average length of a session from entry to exit or after a period of inactivity (default is 30 minutes in GA).</p><h4>Calculation :</h4><pre>Average Session Duration = Total Session Duration ÷ Total Sessions</pre><h4>Why it matters :</h4><p>To understand how much time visitors typically spend on your site across multiple pages in a single visit.</p><h3>3. Average Time on Page</h3><h4>Definition :</h4><p>Average Time on Page measures how long users spend on a specific page, excluding sessions where only one page is viewed.</p><h4>Calculation :</h4><pre>Average Time on Page = Total Time on Page ÷ (Pageviews − Exits/Bounces)</pre><h4>Why it matters :</h4><p>Ideal for evaluating the performance of individual blog posts, landing pages, or articles.</p><h3>4. Average Time on Site</h3><h4>Definition :</h4><p>Average Time on Site is similar to Average Session Duration but typically excludes bounce sessions.</p><h4>Calculation :</h4><pre>Average Time on Site = Total Time on Site ÷ Total No. of Non-Bounce Session</pre><h4>Why it matters :</h4><p>Provides a cleaner view of engaged visits by removing sessions without interaction beyond the initial page.</p><h3>5. Total Time Spent</h3><h4>Definition :</h4><p>Total Time Spent is the cumulative time users spend on your site (or a section of it) across all sessions.</p><h4>Calculation :</h4><pre>Total Time Spent = Sum of Time on Site for All Sessions</pre><h4>Why it matters :</h4><p>Useful for high-level summaries, such as total hours spent on a content hub in a given month.</p><h3>Conclusion</h3><p>Time-based metrics offer a window into how visitors interact with your website, but understanding their nuances is critical for making informed decisions. Each metric serves a different analytical purpose:</p><ul><li>Use <strong>Average Engagement Time</strong> to evaluate how actively users are interacting.</li><li>Use <strong>Average Session Duration</strong> and <strong>Average Time on Site</strong> to understand visit length.</li><li>Use <strong>Average Time on Page</strong> to assess individual content performance.</li><li>Use <strong>Total Time Spent </strong>for broader performance evaluation across campaigns or sections.</li></ul><p>Rather than relying on a single metric, the most valuable insights come from using these metrics together in context. Segment them by traffic source, user type, or content category to gain deeper understanding and optimise accordingly.</p><p>Ultimately, success in digital analytics comes not from tracking every metric available, but from choosing the right metrics that align with your business goals, user experience strategy, and content performance objectives.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cd9211764f0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SQL Joins Explained: A Complete Technical Deep Dive With Examples]]></title>
            <link>https://medium.com/@aman.aggarwal83/sql-joins-explained-a-complete-technical-deep-dive-with-examples-9572eda4664d?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/9572eda4664d</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[sql-joins]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Sun, 27 Jul 2025 20:18:05 GMT</pubDate>
            <atom:updated>2025-07-28T07:52:24.869Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/584/1*n2aiqxcZGx2mab2k0uZHGQ.png" /></figure><p>Whether you’re writing your first SELECT query or building a complex data pipeline, understanding <strong>SQL joins</strong> is crucial. Joins are the glue that holds relational databases together. They let you combine data across tables in powerful ways.</p><p>In this blog, we’ll break down:</p><ul><li>What SQL joins are and why they matter</li><li>Different types of joins</li><li>Real-life examples with visuals</li><li>Tips for writing efficient join queries</li></ul><h3>What is a SQL Join?</h3><p>Imagine a company with two tables:</p><ul><li>employees — contains employee details</li><li>departments — contains department info</li></ul><p>To answer questions like <em>“Which department does each employee belong to?”</em>, you need to combine both tables. That’s where SQL joins come in.</p><blockquote>A <strong>JOIN</strong> clause lets you retrieve related data from multiple tables based on a common field, usually a <strong>foreign key</strong>.</blockquote><h3>Types of SQL Joins</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AFbAWYW0arbodqTefijolQ.png" /></figure><ol><li><em>INNER JOIN</em> — Returns only matching rows from both tables</li><li><em>LEFT JOIN</em> — All rows from the left table, and matched ones from the right</li><li><em>RIGHT JOIN</em> — All rows from the right table, and matched ones from the left</li><li><em>FULL OUTER JOIN</em> — Returns all rows when there is a match in either table</li><li><em>CROSS JOIN</em> — Cartesian product — every combination of rows</li></ol><p>Let’s break each one down with examples. Assume we’re using the following two tables :</p><p>Table: employees</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*NDQvxM5Q9UEzK9rffLVxqw.png" /></figure><p>Table: departments</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/664/1*obDSBUjDlrr9fy8csqWO0w.png" /></figure><ol><li><strong>INNER JOIN</strong><br>Returns only the rows where there’s a match in <strong>both</strong> tables.</li></ol><pre>SELECT e.name, d.dept_name<br>FROM employees e<br>INNER JOIN departments d ON e.dept_id = d.dept_id;</pre><p>Result :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/1*DRMSj1xoJjRqlPFhcLrwvw.png" /></figure><p><strong>2.</strong> <strong>LEFT JOIN</strong><br>Returns all rows from the <strong>left</strong> table and matches from the right. If there’s no match, you’ll get NULLs.</p><pre>SELECT e.name, d.dept_name<br>FROM employees e<br>LEFT JOIN departments d ON e.dept_id = d.dept_id;</pre><p>Result :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/667/1*j8hep3GRsqqHLOPBnmNobw.png" /></figure><p><strong>3. RIGHT JOIN</strong><br>Returns all rows from the <strong>right</strong> table, matched with the left table.</p><pre>SELECT e.name, d.dept_name<br>FROM employees e<br>RIGHT JOIN departments d ON e.dept_id = d.dept_id;</pre><p>Result :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/663/1*DZ4PK-TJNVNwpql4VgT3SA.png" /></figure><p><strong>4. FULL OUTER JOIN</strong><br>Returns all rows from <strong>both</strong> tables. If there’s no match, it fills in NULLs.</p><pre>SELECT e.name, d.dept_name<br>FROM employees e<br>FULL OUTER JOIN departments d ON e.dept_id = d.dept_id;</pre><p>Result :</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/658/1*RSoDNnzcb7yimOg31UiGLQ.png" /></figure><p><strong>5. CROSS JOIN</strong><br>Returns every possible combination of rows between both tables.</p><pre>SELECT e.name, d.dept_name<br>FROM employees e<br>CROSS JOIN departments d;</pre><p>With 4 employees and 4 departments, you’ll get <strong>16 rows</strong> (4×4). Use this carefully!</p><h3>Best Practices for SQL Joins</h3><ol><li><strong>Use aliases</strong> for readability: e, d, etc.</li><li><strong>Filter after joining</strong>, not before — especially in LEFT/RIGHT joins.</li><li><strong>Index join keys</strong> for faster performance.</li><li><strong>Watch out for duplicates</strong> — especially in one-to-many joins.</li><li>**Avoid SELECT *** in joins — be explicit with columns to reduce clutter.</li></ol><p>SQL joins may seem intimidating at first, but once you understand them, you unlock a world of powerful insights. They allow you to model relationships, analyze patterns, and answer complex questions across tables with ease.</p><p>Whether you’re a data analyst, engineer, or backend developer — mastering joins is an investment that pays off every day.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9572eda4664d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Medallion Architecture]]></title>
            <link>https://medium.com/@aman.aggarwal83/medallion-architecture-93cfc95f3b90?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/93cfc95f3b90</guid>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[medallion-architecture]]></category>
            <category><![CDATA[data-engineering]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Tue, 15 Jul 2025 09:06:27 GMT</pubDate>
            <atom:updated>2025-07-15T09:06:27.787Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/820/1*Zx3B2besAXy5AbBwGnTD-w.png" /></figure><p>Imagine your data as uncut gemstones: rough, scattered, and full of potential. <strong>Medallion Architecture</strong> is the polishing process that transforms these raw stones into gleaming jewels, guiding information through three distinct stages — Bronze, Silver, and Gold — until it shines with clarity and purpose.</p><p>In this blog, we’ll unpack each layer, reveal why this approach delivers trustworthy insights, and show you how it plays out in a real-world setting.</p><h3><strong>What is Medallion Architecture?</strong></h3><blockquote>Medallion Architecture is a modern data design framework that organizes information flow into three structured layers — Bronze, Silver, and Gold — each representing a step toward cleaner, more valuable data.</blockquote><ol><li><strong>Bronze Layer — <em>Raw Ingested Data<br></em></strong><em>a. Raw Data Ingestion :</em> Captures data directly from source systems (APIs, logs, databases) in its original, unprocessed format.<br><em>b. Immutable Storage :</em> Stores data as-is — often append-only — to preserve the source of truth for future reprocessing or auditing.<br><em>c. Foundation for Processing :</em> Acts as the input layer for all downstream transformations in Silver and Gold.</li><li><strong>Silver Layer — <em>Cleaned &amp; Enriched Data<br></em></strong><em>a. Purpose :</em> Transforms messy raw data into structured, reliable, and query-ready tables — serving as the foundation for the Gold Layer.<br><em>b. Cleans and Standardizes :</em> Removes duplicates, fixes nulls, formats fields (like timestamps), and ensures consistent schema across datasets.<br><em>c. Enriches and Joins:</em> Merges raw data with reference tables (e.g., user or product info), adds derived columns, and applies business logic.</li><li><strong>Gold Layer — <em>Business-Ready Data</em></strong><br><em>a. Business-Ready Data :</em> Contains aggregated, curated, and presentation-friendly data modeled for specific business use cases.<br><em>b. Supports Reporting &amp; Dashboards :</em> Optimized for tools like Power BI, Tableau, or Looker — enabling self-service analytics and decision-making.<br><em>c. High Trust &amp; Quality :</em> Data in this layer is validated, consistent, and serves as the single source of truth for KPIs and executive reporting.</li></ol><h3>Why Use Medallion Architecture?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*70TNE97aSLywhrBYcQeeow.png" /></figure><blockquote>When managing data at scale, things can get messy fast — disconnected pipelines, unclear data sources, and painful debugging. That’s where Medallion Architecture steps in — not just as a framework, but as a philosophy for building resilient, flexible, and trustworthy data systems.</blockquote><p>Here’s why it’s gaining traction:</p><ol><li><em>Full Data Lineage</em><br>Track each insight back to the exact raw record. When questions arise, you can point to the original Bronze data, audit every transformation in Silver, and validate Gold-level metrics in an instant.</li><li><em>Incremental Development</em><br>Rather than tackling a monolithic pipeline, you advance layer by layer. Ingest Bronze data first, then add Silver cleanses, and finally polish into Gold outputs — so you can deliver value continuously, not just at “go‐live.”</li><li><em>Modular Processing</em><br>By isolating transformations within each layer, you avoid redundant work. Clean once in Silver, then reuse that data for multiple Gold‐layer reports. Compute costs drop, and you gain the flexibility to spin up new use cases in minutes.</li><li><em>Robust Error Handling</em><br>If a downstream job breaks, you don’t lose sleep over partial loads. Simply reprocess from the last known-good Bronze snapshot. Your pipeline stays consistent, auditable, and ready to recover at any moment.</li><li><em>Scalability</em><br>With Medallion Architecture, you get a repeatable framework that grows with your needs. New data sources plug into Bronze; advanced analytics plug into Gold — without disrupting existing workflows.</li></ol><p>Medallion Architecture is more than just a design pattern — it’s a practical and scalable approach to managing the complexity of modern data systems. By organizing data into clearly defined Bronze, Silver, and Gold layers, it promotes better data governance, transparency, and efficiency across the entire pipeline. Start with a single data source and layer, and evolve your architecture as your data needs mature. With this structured foundation in place, you’re not just processing data — you’re enabling reliable, enterprise-grade analytics that can grow with your organization.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=93cfc95f3b90" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Neeraj Chopra — India’s Golden Boy]]></title>
            <link>https://medium.com/@aman.aggarwal83/neeraj-chopra-indias-golden-boy-79ba46c82a27?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/79ba46c82a27</guid>
            <category><![CDATA[javelin-throw]]></category>
            <category><![CDATA[neeraj-chopra]]></category>
            <category><![CDATA[india]]></category>
            <category><![CDATA[golden]]></category>
            <category><![CDATA[olympics]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Sun, 08 Jun 2025 19:12:37 GMT</pubDate>
            <atom:updated>2025-06-08T19:12:37.285Z</atom:updated>
            <content:encoded><![CDATA[<h3>Neeraj Chopra — India’s Golden Boy</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PFeJLiqOCTL3WrhBVVjr4Q.jpeg" /></figure><p>In a country where cricket reigns supreme, legends like Sachin Tendulkar, Virat Kohli, and MS Dhoni have long held the power to keep millions of Indians glued to their screens, even late into the night. Their mastery on the cricket field has made them household names, capable of captivating a nation that eats, sleeps, and breathes cricket.</p><p>But recently, something remarkable happened. An Indian soldier from Haryana, with no bat in hand and no cricket pitch in sight, managed to unite the country in a way that few could have predicted. This time, it wasn’t a cricket match keeping us awake — it was the sight of a single man standing tall on the global stage, javelin in hand, ready to make history.</p><p>We’re talking about <em>Neeraj Chopra</em>, India’s golden boy. With every throw, he has redefined what it means to be a hero in India, showing that dreams can soar as high as the javelin he launches into the sky. He’s given us more than just a reason to stay up late; he’s given us a reason to believe, to hope, and to celebrate the strength of the human spirit.</p><p>Neeraj’s journey from the small village of Khandra in Haryana to Olympic glory is not just a story of athletic prowess. It’s a story of grit, determination, and the unwavering belief that no dream is too big. In a nation where cricket has always been the heartbeat of sports, Neeraj Chopra has made us sit up and take notice of a different kind of champion — one who has shown that with enough heart, even a track and field event can bring a nation to its feet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/442/1*F0LwZdOURXU9-bNteF7heg.png" /><figcaption>Chopra’s Medal Tally</figcaption></figure><p>His impact goes beyond the medals and accolades. Neeraj has inspired an entire generation to look beyond the boundaries of traditional sports, to embrace disciplines that once seemed out of reach. He has ignited a passion for athletics in a country where such sports often struggle for recognition. Young athletes now see in him a role model, someone who has proven that with dedication and belief, even the most unlikely dreams can be realized.</p><p>Neeraj’s influence extends far beyond the field. He’s become a beacon of hope and pride, embodying the resilience and strength of the Indian spirit. His success has sparked conversations about investing in diverse sports, about nurturing talent from all corners of the country, and about the power of perseverance in the face of adversity.</p><p>And then there’s the impact of Neeraj Chopra on social media and television. His every move, every throw, and every triumph is followed with a fervor that rivals the most celebrated cricket matches. Social media platforms are awash with posts and updates celebrating his achievements, as fans from all walks of life come together to share in his success. His inspiring journey has become a trending topic, with hashtags and memes capturing the imagination of millions.</p><p>Television channels have devoted extensive coverage to his events, showcasing his throws and capturing the emotions of his supporters. Special segments and documentaries highlight his training, his background, and the remarkable story behind his success. Neeraj has become a regular feature on talk shows and interviews, further cementing his place as a national hero.</p><p>Through these media channels, Neeraj Chopra’s influence has reached far and wide, bringing the excitement of track and field to living rooms across the country. He has not only broadened the scope of sports coverage in India but also demonstrated the profound impact that a single athlete can have in shaping public interest and inspiring change.</p><p>In every way, Neeraj Chopra’s rise to fame has illuminated the power of sports to unite and inspire. His presence on social media and television has amplified his message, making it clear that heroes come in many forms and that greatness can be found in the most unexpected places.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=79ba46c82a27" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Calculate Web Metrics from Adobe Analytics Data Feed]]></title>
            <link>https://medium.com/@aman.aggarwal83/how-to-calculate-web-metrics-from-adobe-analytics-data-feed-8c22a87cd3d3?source=rss-468957d144da------2</link>
            <guid isPermaLink="false">https://medium.com/p/8c22a87cd3d3</guid>
            <category><![CDATA[adobe]]></category>
            <category><![CDATA[data-feed]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[adobe-analytics]]></category>
            <category><![CDATA[sql]]></category>
            <dc:creator><![CDATA[Aman Aggarwal]]></dc:creator>
            <pubDate>Mon, 08 Apr 2024 15:22:36 GMT</pubDate>
            <atom:updated>2024-04-09T18:20:46.457Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/580/1*GebUwmhiQbhFzRxfxYbPEg.png" /></figure><p>In this detailed discussion, we will explore how to use Adobe Analytics data feeds to calculate metrics accurately. The official documentation can sometimes lack clarity, but I will provide clear guidance, step-by-step instructions, and practical examples to help you effectively compute metrics within Adobe Analytics. By the end of this exploration, you will have a solid understanding of how to leverage these data feeds for precise metric analysis. Let’s dive in!</p><p><strong>1.Page Views :</strong><br>Page views is a metric in Adobe Analytics that measures the number of times a webpage has been viewed or loaded by a unique visitor during a specific time period.<br><em>Definition</em> : Count the number of rows where a value is in post_pagename or post_page_url.</p><pre>SELECT COUNT(CASE WHEN post_pagename != &#39;&#39; OR post_page_url != &#39;&#39; THEN 1 ELSE NULL END) AS &quot;Page Views&quot;<br>FROM adobe_data_feed</pre><p><strong>2. Unique Visitors :</strong><br>A unique visitor is a term used in marketing analytics which refers to a person who has visited the website at least once and is counted only once in the reporting time period.<br><em>Definition</em> : Unique count of combination of post_visid_high and post_visid_low.</p><pre>SELECT COUNT(DISTINCT CONCAT(post_visid_high,post_visid_low) AS &quot;Unique Visitors&quot;<br>FROM adobe_data_feed</pre><p><strong>3. New Visitors :</strong><br>New visitors are those accessing your site for the first time on a specific device. They offer opportunities for engagement and insights into audience diversity.<br><em>Definition</em> : Count of users where visit_num = 1.</p><pre>SELECT COUNT(DISTINCT CASE WHEN visit_num = 1 THEN CONCAT(post_visid_high,post_visid_low) ELSE NULL END) AS &quot;New Visitors&quot;<br>FROM adobe_data_feed</pre><p><strong>4. Returning Visitors :</strong><br>Returning visitors are defined as individuals who have visited your website at least once before and have made a subsequent visit. They signify a segment of your audience that engages with your content on a recurring basis, providing crucial data on your website’s effectiveness.<br><em>Definition</em> : Count of users where visit_num &gt; 1.</p><pre>SELECT COUNT(DISTINCT CASE WHEN visit_num &gt; 1 THEN CONCAT(post_visid_high,post_visid_low) ELSE NULL END) AS &quot;Returning Visitors&quot;<br>FROM adobe_data_feed</pre><p><strong>5. Visits :</strong><br>The Visits metric indicates the total number of sessions across all visitors on your website. It provides a measure of overall traffic and engagement, reflecting the volume of interactions occurring within a specific timeframe.<br><em>Definition</em> : Unique count of combination of post_visid_high, post_visid_low, visit_num, and visit_start_time_gmt.</p><pre>SELECT COUNT(DISTINCT CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt)) AS &quot;Total Visits&quot;<br>FROM adobe_data_feed</pre><p><strong>6. Pages / Visit :</strong><br>It represents the average number of pages viewed by a visitor during a single session on your website. This metric provides insights into visitor engagement and the depth of interaction with your content.<br><em>Definition</em> : Ratio of Total page views and total visits.</p><pre>SELECT COUNT(CASE WHEN post_pagename != &#39;&#39; OR post_page_url != &#39;&#39; THEN 1 ELSE NULL END)/COUNT(DISTINCT CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt)) AS &quot;Pages/Visit&quot;<br>FROM adobe_data_feed</pre><p><strong>7. Bounces :<br></strong>It refers to the number of visits where the users enter a website but don’t click on any other pages before exiting. It’s a metric indicating the percentage of single-page visits.<br><em>Definition</em> : Count of hit where the visit_page_num = 1 for each visit.</p><pre>WITH cte AS(<br>SELECT SUM(visit_page_num) OVER(PARTITION BY CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt) ORDER BY visit_page_num DESC) AS bounce<br>FROM adobe_data_feed<br>) <br>SELECT COUNT(CASE WHEN bounce = 1 THEN 1 ELSE NULL END) AS &quot;Bounces&quot;<br>FROM cte</pre><p><strong>8. Bounce Rate :<br></strong>Percentage of visitors who land on a webpage and then leave without interacting further with the site or viewing other pages within the same site.<br><em>Definition</em> : Ratio of number of bounces and the total visits.</p><pre>WITH cte AS(<br>SELECT post_visid_high, post_visid_low, visit_num, visit_start_time_gmt,<br>SUM(visit_page_num) OVER(PARTITION BY CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt) ORDER BY visit_page_num DESC) AS bounce<br>FROM adobe_data_feed<br>) <br>SELECT COUNT(CASE WHEN bounce = 1 THEN 1 ELSE NULL END)/COUNT(DISTINCT CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt)) AS &quot;Bounce Rate&quot;<br>FROM cte<br><br>*Note : Bounce is the flag which describes whether a visit was single page visit (bounce = 1).</pre><p><strong>9. Total Time Spent on Site :<br></strong>Time on site, also known as session duration, is the total amount of time that someone spends navigating through your website.<br><em>Definition</em> : Sum of Difference of post_cust_hit_time_gmt and visit_start_time_gmt of the last hit of a visit for the all the visits.</p><pre>WITH cte AS(<br>SELECT visit_start_time_gmt, post_cust_hit_time_gmt,<br>ROW_NUMBER() OVER(PARTITION BY CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt) ORDER BY visit_page_num DESC) AS hit_number<br>FROM adobe_data_feed<br>)<br>SELECT SUM(CASE WHEN hit_number = 1 THEN (post_cust_hit_time_gmt - visit_start_time_gmt) ELSE NULL END) AS &quot;Total Time Spent&quot;<br>FROM cte<br><br>*Note : hit_number gives the order of hits for a visit, hit_number = 1 denoting the last hit for that visit.</pre><p><strong>10. Average Session Duration :<br></strong>It is a metric used in web analytics to measure the average length of time a user spends on a website during a single session.<br><em>Definition</em> : Ratio of total time spent on site and total engaged sessions.</p><pre>WITH cte AS(<br>SELECT visit_start_time_gmt, post_cust_hit_time_gmt, post_visid_high, post_visid_low, visit_num,<br>SUM(visit_page_num) OVER(PARTITION BY CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt) ORDER BY visit_page_num DESC) AS bounce,<br>ROW_NUMBER() OVER(PARTITION BY CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt) ORDER BY visit_page_num DESC) AS hit_number<br>FROM adobe_data_feed<br>)<br>SELECT SUM(CASE WHEN hit_number = 1 THEN (post_cust_hit_time_gmt - visit_start_time_gmt) ELSE NULL END)/(COUNT(DISTINCT CONCAT(post_visid_high, post_visid_low, visit_num, visit_start_time_gmt)) - COUNT(CASE WHEN bounce = 1 THEN 1 ELSE NULL END))  AS &quot;Average Session Duration&quot;<br>FROM cte<br><br>*Note : Formula for Average Session Duration = Total Time Spent / (Total Visits - Bounces), bounce and hit_number are utilized the same way as used above.</pre><p>I hope that the insights provided in this blog have been useful for understanding how to precisely compute different metrics using Adobe Analytics data feed. Each metric represents a meaningful aspect of user interaction on your digital platform. By mastering these calculations, you gain deeper insights into your audience’s engagement story that drive informed decision-making.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8c22a87cd3d3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>