<?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 Abhishek Dinavahi on Medium]]></title>
        <description><![CDATA[Stories by Abhishek Dinavahi on Medium]]></description>
        <link>https://medium.com/@abhishekdinavahi?source=rss-c75a582a9160------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Stories by Abhishek Dinavahi on Medium</title>
            <link>https://medium.com/@abhishekdinavahi?source=rss-c75a582a9160------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 01:29:59 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@abhishekdinavahi/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[Market Basket Analysis and its implementation in retail stores:]]></title>
            <link>https://abhishekdinavahi.medium.com/market-basket-analysis-and-its-implementation-in-retail-stores-5298123f630d?source=rss-c75a582a9160------2</link>
            <guid isPermaLink="false">https://medium.com/p/5298123f630d</guid>
            <category><![CDATA[retail]]></category>
            <category><![CDATA[alteryx]]></category>
            <category><![CDATA[market-basket-analysis]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Abhishek Dinavahi]]></dc:creator>
            <pubDate>Tue, 28 Sep 2021 10:53:38 GMT</pubDate>
            <atom:updated>2021-09-28T10:53:38.282Z</atom:updated>
            <content:encoded><![CDATA[<p>Did you ever go to a retail chain and wonder why in majority of layouts, beverages and chips are placed together? In fact there is an urban legend of <strong>Beer and Diaper</strong> which states that middle aged men who purchase diaper also pickup beer on their way home. But, how would one come to the conclusion that a set of categories are to be placed together for a better cross sale? Well, the solution to that is <strong>Market Basket Analysis.</strong></p><p><strong>What is Market Basket Analysis?</strong></p><p>Market Basket Analysis or MBA in short is a technique that provides the association levels between different items an. It looks for combinations of items that occur together in transactions frequently and helps retailers identify trends or patterns.</p><p>For the tech geeks out there, it’s an “<strong>if statemen</strong>t”, where if purchase (milk) then purchase (bread). So MBA forms patterns between milk and biscuit and rates it on different attributes.</p><p>{Milk}-&gt;{Bread}</p><p>There are two main attributes/features in MBA.</p><p><strong>Antecedent</strong> — The <strong>If</strong> part<br><strong>Consequent</strong> — the <strong>Then</strong> part</p><p><strong>How would I perform this analysis if my dataset is really huge?</strong></p><p>Well, nothing to worry about. There are already models which will compute MBA for you without the need of putting in too much effort. The three main metrics on which MBA is calculated are</p><p><strong>Support: </strong>It is the popularity of an item. Well, to put it in simpler terms, it is the ratio of transactions involving an item to the entire transaction count</p><p>Ex: Let us say you have bought five packets of Lays Magic masala chips packet and have bought 30 different items. Support for the Lays packet would be 5/30 = 0.16</p><p><strong>Confidence: </strong>Likelihood that customer who bought both P1 and P2. Confidence divides the number of transactions involving both the products by the transactions involving only one product. This will help us determine which product is influencing the purchase of which product.</p><p>Ex: Confidence (Lays-&gt;Coca Cola) = (Transactions involving both Lays and Coke)/Transactions involving only Lays</p><p>In our case, let us assume<strong> </strong>that<strong> </strong>we have<strong> </strong>purchased<strong> 6 Lays and 2 Cokes.</strong></p><p>Confidence(Lays-&gt;Coca Cola) = (2/30)/(6/30) = 0.33</p><p><strong>Lift: </strong>Helps us in determining<strong> </strong>increase in sale of P1 when we sell to P2.<br>Lift(A =&gt; B) = Confidence(A, B) / Support(B)</p><p>Lift can be interpreted as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/497/1*cY3bLwtwFzrTudoZJr7Bdg.png" /><figcaption>Lift Values and Interpretation</figcaption></figure><p><strong>Do I have to code for getting my MBA?</strong></p><p>Well not really. You can perform your MBA in coding language of your choice, but incase you are a non-coding business analyst, you have a variety of tools that help you perform the same analysis.</p><p>Alteryx has dedicated tools for performing MBA</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/824/1*DAxNGky7yyT8KWnOBd6XwA.png" /><figcaption>Market Basket Analysis results</figcaption></figure><p>Alteryx provides us the flexibility of choosing different algorithms for MBA like Apriori and ECLAT with multiple configurations for parameter tuning for the most optimized results.</p><p>Any thoughts on how else MBA can be used? Please list it down in the comments.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5298123f630d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[RFM Analysis and how companies can use it for better efficiency:]]></title>
            <link>https://abhishekdinavahi.medium.com/rfm-analysis-and-how-companies-can-use-it-for-better-efficiency-fc16040ddd8?source=rss-c75a582a9160------2</link>
            <guid isPermaLink="false">https://medium.com/p/fc16040ddd8</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[alteryx]]></category>
            <category><![CDATA[rfm]]></category>
            <category><![CDATA[rfm-analysis]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Abhishek Dinavahi]]></dc:creator>
            <pubDate>Mon, 20 Sep 2021 11:28:23 GMT</pubDate>
            <atom:updated>2021-09-20T11:28:23.913Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>What is RFM Analysis?</strong></p><p>RFM Analysis is a marketing methodology/technique to group customers and rank them based on the metrics of Recency, Frequency and Monetary from their past transactions to segment customers and perform targeted marketing campaigns for a better throughput.</p><p>RFM Analysis quantitatively ranks customers on the basis of</p><p><strong>Recency (R): </strong>When was the last transaction made? Companies usually try to measure recency in days. But this may vary depending on the product or the industry. If it’s an FMCG product like milk, then recency would be on a daily basis, but if it was a manufacturing equipment, the recency might have to calculate in years.</p><p><strong>Frequency (F): </strong>Frequency of the purchase the customer has made over a given period of time. First time buyers might be good target for targeted follow-up advertising to convert them into more frequent customers.</p><p><strong>Monetary (M): </strong>Amount spent by a customer over a period of time. Customer who have a high transaction value have a better probability of spending more and could potentially add more value to the business.</p><p><strong>How to compute RFM analysis:</strong></p><p>RFM analysis scores customers on each of the three main factors. In general, each customer is given a score from 1 to 5 is given, with 1 being the least and 5 being the highest. However, depending on the industry, various implementations of an RFM analysis system may use slightly different values.</p><p>Organizations average these values together, then sort customers from highest to lowest to find the most valuable customers, while some businesses weigh the values differently while calculating the final scores.</p><p><strong>Analytics for enabling RFM Analysis:</strong></p><p>The tool I would be using today is <strong>Alteryx</strong>, a self-service analytics tool that enables citizen data scientists and business analysts achieve complex tasks without the burden of writing code. An analyst can create an automated workflow in Alteryx with different rules such that, Alteryx can directly source data from the database (SAP, Oracle etc), apply complex logics, and output the results to the corresponding stake holders without any manual intervention required enabling the analyst to focus more on deriving insights.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/934/1*kShZC_jWqFhTT1pWz67kng.png" /></figure><p><strong>How to analyze these results?</strong></p><p>Customers with the top scores across all the segments are our best customers with high spend, frequent visits. These are our most loyal customers who should be safeguarded.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/581/1*JsKVmzlnlYFA0P6nDDuGVw.png" /></figure><p>Customers with good frequent visits but low monetary score can be target with some discounts to increase their purchasing power to transition them into customers with high spending.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/593/1*nIJ8UTJLyVHATTu5woHvOA.png" /></figure><p>Multiple such combinations can be put in place to target customers for the best revenue generation by a company.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fc16040ddd8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Impact of Alteryx on mitigating the risk of Duplicate Payments]]></title>
            <link>https://abhishekdinavahi.medium.com/impact-of-alteryx-on-mitigating-the-risk-of-duplicate-payments-78ab268b99a9?source=rss-c75a582a9160------2</link>
            <guid isPermaLink="false">https://medium.com/p/78ab268b99a9</guid>
            <category><![CDATA[accounts-payable-software]]></category>
            <category><![CDATA[vendors]]></category>
            <category><![CDATA[duplicate-payments]]></category>
            <category><![CDATA[accounts-payable]]></category>
            <category><![CDATA[alteryx]]></category>
            <dc:creator><![CDATA[Abhishek Dinavahi]]></dc:creator>
            <pubDate>Mon, 14 Jun 2021 02:48:22 GMT</pubDate>
            <atom:updated>2021-06-14T02:48:22.184Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/376/1*jkKyr1N7qy22jpetH5rolA.jpeg" /></figure><p><strong>What are Duplicate Payments?</strong></p><p>Duplicate Payments are additional payments made to a vendor/supplier that have already been paid. These anomalies usually occur owing to flaws in an organization’s accounts payable (AP) process that do not detect the presence of prior payments.</p><p>For Example, if a “<strong>Company ABC”</strong> procured a “<strong>Material A”</strong> worth 2 crores from “<strong>Vendor X”</strong>, a duplicate payment will mean that the AP department of “<strong>Company ABC”</strong> pays “<strong>Vendor X”</strong> twice for a single transaction losing out on 2 crores worth of money.</p><p>Duplicate Payments can add up to substantial losses if left undetected. In addition, companies spend significant amount of time trying to identify such payments which could have been used in more productive things otherwise.</p><p><strong>Causes of Duplicate Payments:</strong></p><p>There might be several reasons for duplicate payments ranging from manual errors to complex internal business processes. Some of the causes include</p><p>· Discrepancies in Vendor Master File</p><p>· Data Entry Errors</p><p>· Pricing Errors (In case of International Vendors)</p><p>· Payments sent via multiple channels.</p><p>· Internal changes made to software.</p><p>It can be time consuming and difficult for a business to detect a duplicate payment before the payment is processed, as duplicates aren’t by definition exact replicas of one another. For example, duplicate invoices might have different users posting them, different dates or even different invoice numbers. This might occur when vendors send both an electronic and paper invoices, or if they re-send the same invoice after not having been paid by the agreed-upon date.</p><p>A thorough audit of Accounts Payable (AP) department is required to mitigate such requests. This can be done internally but could prove expensive and inefficient.</p><p><strong>How can Alteryx Help?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/491/1*Czo3tJeSLmmngldFLozUdw.png" /></figure><p>An analyst can create an automated workflow in Alteryx with different rules such that, Alteryx can directly source data from the database (SAP, Oracle etc), apply complex logics, and output the results to the corresponding stake holders without any manual intervention required from the AP department’s end, enabling them to focus on solving more complex problem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/602/1*H03uP1ONEhn-jDjRAeGkHA.png" /></figure><p><strong>Execution Methodology:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/602/1*RIPHMRA6EAUX96Zz2lPsNg.png" /></figure><p><strong>Sourcing Data</strong>: If we consider our ERP to be SAP, analysts normally spend time downloading data from tables like ACDOCA or FBL1N. Due to the high transaction size of these tables, it will prove to be a herculean task to download the data.</p><p>Alteryx can help you directly connect to SAP HANA, so that you don’t have to manually download the data at any point. Alteryx can source data on different filter criteria from these tables.</p><p><strong>Rule Creation: </strong>Post<strong> </strong>getting<strong> </strong>our data in Alteryx, we can create multiple rules owing to the wide variety of functionalities Alteryx can offer. The rules/combinations could be as follows</p><p>· <strong>Rule 1:</strong> Same (User, Vendor, Posting Date, Document Type, Amount)</p><p>· <strong>Rule 2:</strong> Same (User, Document Type, Vendor, Date, Amount) Different Document Numbers</p><p>· <strong>Rule 3:</strong> Same (Vendor, Amount, Document Type, Account Number)</p><p>This level of detail will help us cover a wider ground helping us better mitigate the risk of Duplicate Payments.</p><p><strong>Output the Results: </strong>We can set the results in such a way that, once all the complex analysis is done, Alteryx will automatically mail these documents to stakeholders.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78ab268b99a9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Free Will-Is it an Illusion?]]></title>
            <link>https://abhishekdinavahi.medium.com/free-will-is-it-an-illusion-24d0e8dc5098?source=rss-c75a582a9160------2</link>
            <guid isPermaLink="false">https://medium.com/p/24d0e8dc5098</guid>
            <category><![CDATA[sam-harris]]></category>
            <category><![CDATA[action]]></category>
            <category><![CDATA[illusion]]></category>
            <category><![CDATA[free-will]]></category>
            <category><![CDATA[choices]]></category>
            <dc:creator><![CDATA[Abhishek Dinavahi]]></dc:creator>
            <pubDate>Wed, 27 Jan 2021 16:44:41 GMT</pubDate>
            <atom:updated>2021-01-27T16:44:41.614Z</atom:updated>
            <content:encoded><![CDATA[<p>I was once having a discussion with my uncle and was introduced to ideologies of Sam Harris who, along with many scientists believe that free will is an illusion. Every single decision/action we take under the assumption of freewill is influenced by our past actions and the consequences of that actions. Nothing we do is original. Similar to a Machine Learning algorithm, we are just utilizing the results of our past actions to make better informed decisions either to improve the results or to avoid them.</p><p>Let us take an example. Suppose you go to a hotel, there is a high probability that you order the same item you always have. This is because, the items you order have already been tried and tested by you multiple times and you like the taste of it. Incase you order something new, you might have read about that dish somewhere in a magazine or your friend might have suggested it to you. So as you see here, your motive of choosing an item is being influenced by either your own taste, or something you might have read or heard sometime in the past. Let us take an extreme scenario, wherein you (after reading this article), trying to prove me wrong close your eyes and randomly scroll your finger across the menu and select an item. Does that mean you have done something out of free will ? <br>Well no! Your motive of closing your eyes and selecting an item that way has only been driven by your urge to prove an article wrong. So your action was motivated through this article which you have read in the past (before selecting an item the next time you go to a restaurant).</p><p>This same logic can be applied to every scenario in life. What do you feel? Does the thought of having no free will disappoint you? Let me know in the comments.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=24d0e8dc5098" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[All hail the Multi-Row Formula!! Alteryx Designer]]></title>
            <link>https://abhishekdinavahi.medium.com/all-hail-the-multi-row-formula-alteryx-designer-3f1944333d31?source=rss-c75a582a9160------2</link>
            <guid isPermaLink="false">https://medium.com/p/3f1944333d31</guid>
            <dc:creator><![CDATA[Abhishek Dinavahi]]></dc:creator>
            <pubDate>Fri, 01 Jan 2021 09:56:13 GMT</pubDate>
            <atom:updated>2021-01-01T14:13:03.478Z</atom:updated>
            <content:encoded><![CDATA[<p>We often come across cases where we are required to perform calculations between different rows of either the same column or different columns to derive insights. Examples include calculating the percentage difference or any other complex conditions. Excel offers an unparalleled experience in performing such calculations owing to its flexible nature.</p><p>However, many times, we notice excel hanging when we try to perform such operations on huge datasets, resulting in delay of our deliverable. It is one of the reasons, corporations across the world invest in alternative ETL tools to perform such complex operations to save both time and automate a process.</p><p>We will talk about how to leverage this functionality of performing row level interactions in Alteryx. Alteryx is a self-service ETL analytics tool with an intuitive user interface which can be used to Extract, Transform and Load Data.</p><p>There is an extremely useful tool called “<strong>Formula”</strong> in Alteryx which can perform complex operations like conversions, spatial, numerical, text operations, conditional functions etc. However, the formula tool only works on a column basis. It interacts with your data on a column basis, where in it fills or populates our rows in a column based on our input. It does not make rows inside a column interact with each other. So, what is the solution? Enter the <strong>Multi-Row Formula.</strong></p><p>The Multi-Row Formula is available under the “Preparation” set. It allows us to utilize row data as part of the formula creation. It lets us decide how two or more rows must interact based on a condition. The Multi-Row Formula tool can only update one field per tool instance. If you would like to update multiple fields, a Multi-Row Formula tool must be added to your workflow for each field to be updated.</p><p>Let us go though an example by comparing the same scenario in both Excel and Alteryx.</p><p>You have found out that a particular material is being procured from the same vendor, in the same location and plant at different prices, and you are now tasked with finding out the percentage difference in price to see what the incurred revenue leakage is.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/648/1*XPE-s0APtutbTJSugrSEqA.png" /></figure><p>I can create a new column in F by the name Percentage Difference, and can enter the formula <br> ((E3 — E2)/E3)*100 ( (Upper Value — Lower Value)/(Upper Value)*100) to calculate my percentage difference and copy paste the same formula for all the cells.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/865/1*3-RTqYCFLmTppXITJ88k_A.png" /></figure><p>However, you will notice that the values populated in F3, F5 do not make any business sense. Ideally, from a business perspective F3 should not be calculated as the Material, Vendor, Location and Plant combination is different for row 3 and row 4. These are two different combinations and cannot be used together to calculate a percentage difference. This problem can be addressed with an if condition. Using an if condition, we can instruct excel to perform this operation if and only if the Material, Vendor, Location and Plant column’s values are same for the two rows that are being compared or that are being used to make the calculation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/927/1*_stiS32-oPIEo0beZxuw5Q.png" /></figure><p>Now, you can notice that the percentage difference is only calculated at instances where our condition is satisfied.</p><p>Let us take a look at how the same can be performed in Alteryx.</p><p>The configuration of the Multi-Row formula can be found below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/856/1*SHO0WR8GWdFJMxc850EfEA.png" /></figure><p>Similar to the formula tool, the Multi-Row formula tool offers you the feasibility of either updating an existing column or creating a new column altogether where in we can decide the Name and the datatype of the column.</p><p>Let us inspect each functionality of the tool. The variables tell us about the row levels. i.e., whether it is an active row, the preceding row, or the succeeding row. The configuration is as below.</p><p>Row — 1: The row before the active row.</p><p>Row + 0: This is the Active Row. The active row is where the result goes into.</p><p>Row + 1: The row after the active row.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/848/1*Ity9GZs0hapTaZi2AWZ80Q.png" /></figure><p>As you can notice, when we select a variable, we can see all the columns that are mapped to that variable out of which we can make our selections. The condition or the formula is entered in the Expression tool. If we want to compare this to excel, [Row + 1: Price] is E5 and [Price] is E4. Once the condition is entered, the Multi-Row formula will populate these values to the entire column similar to the copy paste functionality we have done for excel.</p><p>The Num Rows field can be used to set the row variables that display as in the expression editor. When set to 1, these variables display Row + 1 and Row — 1. When set to 2, we can see Row + 1, Row — 1, Row + 2, Row — 2, Row 0. To put it in simple terms, if you want to use <em>n</em> rows for your calculations, you populate the value <em>n</em> in Num Rows.</p><p>You might be wondering, what about our condition we have put in excel, where we wanted to calculate the percentage difference if and only if my Material, Vendor, Location and Plant are same for the two rows that are being used. Well, Alteryx has got you covered there. The Multirow formula has a GroupBy option which provides the exact same functionality. When you select the columns in the Groupby field, Alteryx will perform your condition entered in the Expression Window only if the combination of the column selected in the GroupBy field match between the rows.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FXhusU21yY7_URKBmqSAbw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3f1944333d31" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>