<?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 Kanishka Agarwal on Medium]]></title>
        <description><![CDATA[Stories by Kanishka Agarwal on Medium]]></description>
        <link>https://medium.com/@kashishaggarwal987?source=rss-beddb795f10b------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Stories by Kanishka Agarwal on Medium</title>
            <link>https://medium.com/@kashishaggarwal987?source=rss-beddb795f10b------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 17:42:29 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kashishaggarwal987/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[Beyond Beautiful Dashboards: Turning Data into Decisions with Power BI]]></title>
            <link>https://medium.com/@kashishaggarwal987/beyond-beautiful-dashboards-turning-data-into-decisions-with-power-bi-abe7ac8da23d?source=rss-beddb795f10b------2</link>
            <guid isPermaLink="false">https://medium.com/p/abe7ac8da23d</guid>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[power-bi]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[business-intelligence]]></category>
            <dc:creator><![CDATA[Kanishka Agarwal]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 14:38:35 GMT</pubDate>
            <atom:updated>2026-09-14T14:38:35.664Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MG9kxIQPG8BrDXvvmB0b_g.jpeg" /></figure><p>A sales dashboard shows revenue rising by 20%. The charts look encouraging, and the report appears to tell a success story. Then someone asks a simple question: “If we sold more, why did our profit fall?”</p><p>That question reveals the difference between displaying data and understanding it.</p><p>An effective dashboard helps readers recognise a situation, investigate possible explanations, and decide what to do next. Its appearance matters because clear design makes information easier to understand. However, attractive visuals cannot compensate for incorrect calculations or missing business context.</p><p>Power BI can support this process by connecting data preparation, modelling, calculations, and visual exploration. The analyst’s responsibility is to make those elements serve a meaningful business question.</p><p>1. Begin with the Decision</p><p>Before selecting a visual, identify who will use the report and what they need to decide.</p><p>A sales manager might want to evaluate a promotion. An operations manager might need to adjust staffing. A purchasing manager might be reviewing supplier costs. Although they may work with related data, their decisions require different information.</p><p>For this article, the question is:</p><p>Did higher sales improve profitability enough to justify continuing a promotion?</p><p>Revenue alone cannot answer this. We need net sales, cost of goods sold, gross profit, and gross margin. Depending on the decision, advertising expenses, fulfilment costs, and customer behaviour may also matter.</p><p>Defining the question first helps keep the report focused.</p><p>2. Examine the Story Behind Sales Growth</p><p>Consider the following hypothetical retailer. These figures are illustrative and do not represent results from a completed Power BI project.</p><p>Metric April May<br>Net sales ₹10,00,000 ₹12,00,000<br>Cost of goods sold ₹7,00,000 ₹9,30,000<br>Gross profit ₹3,00,000 ₹2,70,000<br>Gross margin 30% 22.5%</p><p>Net sales increased by 20%, while gross profit decreased by 10%. Gross margin declined by 7.5 percentage points, from 30% to 22.5%.</p><p>These measures tell different parts of the same story. The business generated more revenue but retained less gross profit from those sales.</p><p>Gross profit also differs from net profit. The example excludes operating expenses such as rent and salaries, so it cannot establish the business’s final profitability.</p><p>The next step is investigation. Higher supplier costs, larger discounts, or a shift towards lower-margin products could explain the decline. The totals alone do not establish which explanation is correct.</p><p>3. Check the Data Before Trusting the Visuals</p><p>A convincing chart can still communicate an incorrect result.</p><p>Duplicate transaction lines may inflate sales. Missing costs may overstate profit. Incorrect dates may place transactions in the wrong period. Returns also need consistent treatment.</p><p>Understanding what each row represents is essential. If an order contains three product lines, counting rows would count three records rather than one order. An order-count measure must reflect the actual business definition.</p><p>Before designing the report, the analyst should check identifiers, data types, missing values, and reconciliation with source totals.</p><p>A clear model supports this work. Microsoft’s star-schema guidance describes fact tables containing events or observations and dimension tables describing entities such as products, stores, and dates. These structures support filtering, grouping, and summarisation. [1]</p><p>The relationships between tables also need checking. A well-organised model makes calculations easier to understand and maintain.</p><p>4. Define Consistent Measures with DAX</p><p>Suppose a transaction-line table called Sales contains two numeric columns:</p><p>* NetSalesAmount: revenue after discounts and returns.<br>* CostAmount: the corresponding total cost of goods for that line.</p><p>The following definitions would be created as separate measures:</p><p>Net Sales =<br>SUM(Sales[NetSalesAmount])<br>Cost of Goods Sold =<br>SUM(Sales[CostAmount])<br>Gross Profit =<br>[Net Sales] — [Cost of Goods Sold]<br>Gross Margin % =<br>DIVIDE([Gross Profit], [Net Sales])</p><p>These are example measures for the stated table structure. They have not been executed against an accompanying Power BI file.</p><p>Format the final measure as a percentage. For May, ₹2,70,000 divided by ₹12,00,000 gives 22.5%.</p><p>Microsoft documents that DIVIDE returns a blank for division by zero unless an alternative result is supplied. [3] This avoids presenting an undefined margin as a meaningful zero.</p><p>The calculation also uses a ratio of totals. Simply averaging product margin percentages can misrepresent overall performance because products contribute different amounts of revenue.</p><p>5. Give Every Visual a Clear Purpose</p><p>The report layout should follow the reader’s questions.</p><p>KPI cards could display net sales, gross profit, and gross margin, each with its comparison period. Bar charts could show April and May side by side. With more historical data, separate trend charts could help investigate whether the change is temporary or persistent.</p><p>Product-level comparisons could show where profit declined most. However, such comparisons require product-level data; they cannot be inferred from the monthly totals.</p><p>Microsoft recommends considering the audience, highlighting important information, and avoiding clutter. [2]</p><p>Titles should communicate meaning. “Sales increased while gross margin declined” tells readers more than “Monthly Analysis.”</p><p>Use readable labels, visible units, and consistent colours. Avoid relying on colour alone to indicate improvement or deterioration. When charts use different scales, make those differences clear.</p><p>6. Make Exploration Easy to Understand</p><p>In the Power BI service, a dashboard provides a single-page overview, while underlying reports support more detailed analysis. [2]</p><p>A report can help readers investigate particular periods, products, or stores. The selected context must remain visible so that a local result is not mistaken for company-wide performance.</p><p>Comparisons also need care. April and May have different numbers of calendar days, and trading days or holidays may affect sales. Revenue per trading day could provide additional context where appropriate.</p><p>A reporting-period label and refresh timestamp help readers understand what the report covers. Testing common filter combinations helps identify misleading totals or comparisons before the report is shared.</p><p>7. Turn the Finding into a Testable Action</p><p>Suppose further analysis shows that the profit decline is concentrated in heavily discounted products.</p><p>A practical recommendation could be to test a smaller discount on selected products. Before starting, the team should identify the owner, duration, and measures used to evaluate the trial.</p><p>Gross profit, order volume, and customer response could be monitored together. A discount change that improves margin but causes a substantial loss of sales may not deliver the intended business result.</p><p>Where practical, compare the trial with similar products or stores using the existing offer. Account for differences in demand and seasonality.</p><p>An observed relationship between discounts and lower margins does not prove that discounts caused the entire decline. The recommendation should reflect that uncertainty.</p><p>8. Review What Happened Next</p><p>Analysis should continue after a recommendation is made.</p><p>Record the proposed action, its owner, the review date, and the outcome. If performance improves, investigate whether the change explains it. If results disappoint, revisit the assumptions.</p><p>This creates a repeatable process:</p><p>Define the question → Prepare the data → Calculate measures → Explain the findings → Test an action → Review the outcome.</p><p>The value of the dashboard becomes easier to judge when it supports a decision that can be evaluated.</p><p>9. Learning and Motivation</p><p>A business education and technical analytics training can complement each other. Business knowledge helps frame the question, while analytical skills help investigate it.</p><p>For a learner studying at Imarticus Learning, a useful reflection would connect a specific lesson or project experience with this process. Examples might include discovering an incorrect calculation, choosing a more suitable chart, or learning to explain a result clearly.</p><p>The most meaningful reflection describes what actually happened, what changed in the learner’s understanding, and how that lesson influenced the article.</p><p>Conclusion</p><p>A useful dashboard combines trustworthy data, meaningful measures, clear comparisons, and practical interpretation.</p><p>The retail example shows why these elements matter. Higher sales initially appeared encouraging, but examining gross profit and margin changed the discussion.</p><p>Power BI can make those relationships easier to explore. The analyst must still explain what the evidence supports, acknowledge what remains uncertain, and propose a sensible next step.</p><p>Beautiful dashboards attract attention. Carefully reasoned analysis helps turn that attention into a better business decision.</p><p>References</p><p>1. <a href="https://learn.microsoft.com/en-us/power-bi/guidance/star-schema">Microsoft Learn: Understand star schema and the importance for Power BI</a><br>2. <a href="https://learn.microsoft.com/en-us/power-bi/create-reports/service-dashboards-design-tips">Microsoft Learn: Tips for designing a great Power BI dashboard</a><br>3. <a href="https://learn.microsoft.com/en-us/dax/divide-function-dax">Microsoft Learn: DIVIDE function in DAX</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=abe7ac8da23d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[SQL to Power BI: Turning Raw Sales Data into Actionable Business Insights]]></title>
            <link>https://medium.com/@kashishaggarwal987/when-i-first-started-learning-data-analytics-i-thought-it-was-all-about-writing-sql-queries-and-13d0c73d2199?source=rss-beddb795f10b------2</link>
            <guid isPermaLink="false">https://medium.com/p/13d0c73d2199</guid>
            <dc:creator><![CDATA[Kanishka Agarwal]]></dc:creator>
            <pubDate>Fri, 03 Jul 2026 16:03:59 GMT</pubDate>
            <atom:updated>2026-07-03T16:58:38.391Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AD4ONquIsc4azRCNqPe9tw.png" /><figcaption>COFFE SHOP SALES — DASHBOARD</figcaption></figure><p>When I started working on this project, my main objective was not simply to create charts but to understand how data visualization can simplify complex business information. I chose the Coffee Shop Sales dataset because it closely represents a real retail business where hundreds of transactions take place every day. Although the dataset initially appeared to be just rows of numbers, I discovered that every transaction contained valuable insights about customer behavior and business performance. After cleaning and analyzing the data in SQL, I imported it into Power BI to build an interactive dashboard that could tell a meaningful story.</p><p>The dashboard begins with three key performance indicators — <strong>Total Sales ($157K)</strong>, <strong>Total Orders (33,527)</strong>, and <strong>Total Quantity Sold (48,233)</strong> — which provide an instant snapshot of the business. Instead of reading through lengthy reports, users can quickly understand the overall performance by looking at these KPIs. To analyze sales trends over time, I created a daily sales trend visualization that highlights fluctuations throughout the month. This helped me understand whether sales remained consistent or experienced noticeable increases and decreases during different periods.</p><p>One of the most interesting parts of the dashboard is the comparison between <strong>weekday and weekend sales</strong>. Visualizing this information made it easier to understand how customer purchasing patterns change depending on the day of the week. Such insights can help businesses make informed decisions about staffing, inventory planning, and promotional campaigns. I also included a <strong>store location analysis</strong>, which compares the performance of different branches. Seeing the revenue generated by each location helped me identify which stores performed consistently well and which locations might require additional attention.</p><p>To better understand customer preferences, I created visualizations for <strong>product categories</strong> and <strong>product types</strong>. These charts clearly show which beverages and food items contribute the most to overall sales. Rather than making assumptions, the dashboard provides data-driven evidence about customer demand. A business owner can use this information to optimize inventory, introduce targeted promotions, or focus on high-performing products.</p><p>My favorite visualization in the dashboard is the <strong>Sales by Days and Hours heatmap</strong>. It offers a clear picture of customer activity throughout the week and different times of the day. By identifying peak business hours, coffee shop managers can plan employee schedules more efficiently and ensure sufficient stock is available during busy periods. I found this visualization particularly valuable because it transforms thousands of individual transactions into a simple pattern that anyone can understand within seconds.</p><p>Throughout this project, I realized that the true power of Power BI lies not in creating attractive dashboards but in communicating insights that support better decision-making. Every chart, KPI, and filter was designed with a purpose — to answer a business question rather than simply display data. This experience taught me that effective data visualization combines technical skills with storytelling, enabling raw data to become meaningful information that drives smarter business decisions.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=13d0c73d2199" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>