<?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 Fahrell Giovanny on Medium]]></title>
        <description><![CDATA[Stories by Fahrell Giovanny on Medium]]></description>
        <link>https://medium.com/@fahrellgiovanny?source=rss-967249932650------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*jbfa5lbjj4Gas4iPJkSNEQ.jpeg</url>
            <title>Stories by Fahrell Giovanny on Medium</title>
            <link>https://medium.com/@fahrellgiovanny?source=rss-967249932650------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 13:25:57 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@fahrellgiovanny/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[How IDEAL Makes Mortgage Product Selection More Transparent and Easier]]></title>
            <link>https://medium.com/@fahrellgiovanny/how-ideal-makes-mortgage-product-selection-more-transparent-and-easier-ddc865b80791?source=rss-967249932650------2</link>
            <guid isPermaLink="false">https://medium.com/p/ddc865b80791</guid>
            <category><![CDATA[homeownership-indonesia]]></category>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[mortgage]]></category>
            <category><![CDATA[kpr]]></category>
            <category><![CDATA[homeownership]]></category>
            <dc:creator><![CDATA[Fahrell Giovanny]]></dc:creator>
            <pubDate>Wed, 19 Feb 2025 05:16:20 GMT</pubDate>
            <atom:updated>2025-02-19T05:24:40.021Z</atom:updated>
            <content:encoded><![CDATA[<p><em>In Indonesia, the journey to homeownership can often feel </em><strong><em>overwhelming</em></strong><em>, especially when navigating the </em><strong><em>complexities</em></strong><em> of mortgage products. With a diverse range of options from multiple banks and property developers, potential homeowners are often left </em><strong><em>confused</em></strong><em> about which mortgage product best </em><strong><em>suits</em></strong><em> </em><strong><em>their</em></strong><em> </em><strong><em>needs</em></strong><em>.</em></p><blockquote><strong>IDEAL Indonesia (www.ideal.id)</strong>, a multi-platform application designed to <strong>simplify</strong> the mortgage selection process, making it more <strong>transparent</strong>, <strong>accessible</strong>, and <strong>user-friendly</strong>.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JNyqRdi86XqWJlpI" /></figure><h3>The Challenge: Simplifying a Complex Process</h3><p>Mortgage selection is inherently <strong>complicated</strong>. It involves understanding various loan types, interest rates, tenures, and eligibility criteria. In Indonesia, where a significant portion of the population is <strong>unbanked</strong> or has limited <strong>financial</strong> <strong>literacy</strong>, this complexity becomes even more pronounced. IDEAL Indonesia <strong>recognized</strong> this challenge and set out to create a solution that would not only simplify the process but also educate users along the way.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jfrl9lc6omosjx9jZtekRg.png" /><figcaption>Old Product Page (mobile)</figcaption></figure><h3>The Solution: A User-Centric Approach</h3><p>IDEAL Indonesia’s platform allows users to <strong>explore</strong> mortgage options from over 30 major banks, financial institutions, and reputable property developers. The process begins with users submitting their mortgage <strong>preferences</strong> and <strong>customer</strong> <strong>profiles</strong>. Based on this information, the platform automatically <strong>sorts</strong> and <strong>filters</strong> mortgage products, presenting users with the tailored and most <strong>suitable</strong> <strong>options</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CHr3v8LEa6haoUul3fzCUA.png" /><figcaption>New Product Page (mobile)</figcaption></figure><h3><strong>Key Features of the Platform:</strong></h3><p><strong>Product Page with Clearer and Detailed Information</strong><br>The product page on IDEAL Indonesia’s platform is designed to provide users with all the <strong>essential</strong> information needed to make an <strong>informed</strong> <strong>mortgage</strong> <strong>decision</strong>. At a glance, users can view <strong>key</strong> <strong>details</strong> such as interest rates, loan tenure, bank type (Conventional or Sharia), and initial installment amounts. This comprehensive presentation allows users to <strong>quickly</strong> <strong>compare</strong> products and <strong>understand</strong> the financial implications of each option. By offering clear, upfront information, IDEAL <strong>eliminates</strong> <strong>confusion</strong> and <strong>empowers</strong> users to <strong>confidently</strong> choose the mortgage product that best <strong>suits</strong> <strong>their</strong> <strong>needs</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CHr3v8LEa6haoUul3fzCUA.png" /></figure><p>IDEAL Indonesia ensures transparency and confidence in decision-making by providing detailed product information for each mortgage option. <strong>Beyond</strong> <strong>basic</strong> details like loan rates, tenure, and initial installment amounts, users can access a comprehensive breakdown, including the total <strong>loan lifetime value</strong>, a clear <strong>installment structure</strong>, and any <strong>additional</strong> fees or <strong>charges</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4SHr6vywfJY_l5cMaYH2pg.png" /><figcaption>New Cart Page (mobile</figcaption></figure><p>For <strong>added</strong> <strong>convenience</strong>, users can <strong>download</strong> a detailed loan <strong>calculation</strong> directly from the platform, which is sent to their email for further review or sharing with family and financial / mortgage <strong>advisors</strong>. This level of detail <strong>empowers</strong> users to make well-<strong>informed</strong> mortgage decisions with <strong>clarity</strong> and <strong>confidence</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-kWEWiW4OVTF1AIVuOT4qQ.png" /><figcaption>Download Calculation Detail</figcaption></figure><p><strong>Mortgage Cart for Comparison</strong><br>IDEAL Indonesia’s mortgage cart feature is designed to make the selection process even more user-<strong>friendly</strong> by allowing users to add <strong>up</strong> <strong>to</strong> <strong>three</strong> mortgage products from different banks for <strong>direct</strong> <strong>comparison</strong>. Once added to the cart, users can view these products <strong>side</strong> <strong>by</strong> <strong>side</strong>, enabling them to easily <strong>evaluate</strong> key factors such as interest rates, loan tenures, monthly installments, and total loan costs. This side-by-side comparison eliminates the need to <strong>switch</strong> between multiple pages or remember <strong>complex</strong> details, <em>streamlining the decision-making process.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j854OgswrrUKybLsIOZDYA.png" /><figcaption>Product Comparison</figcaption></figure><p>To further enhance user success, IDEAL requires users to add <strong>at least two products</strong> before applying. This not only encourages users to compare options and make informed decisions but also <strong>increases their chances of receiving an offer letter</strong> from at least one bank, making the mortgage application process more efficient and accessible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mCE0rYbDHjljw9JnVyp8zw.png" /><figcaption>Minimum Two Mortgage Product</figcaption></figure><p><strong>Educational Icons and Simplified Flow</strong><br>To help users better <strong>understand</strong> the products, IDEAL introduced <strong>icons</strong> that highlight the <strong>advantages</strong> of each mortgage option. Additionally, the platform was redesigned to remove <strong>unnecessary</strong> steps, such as a separate product detail page, and instead, all relevant information is now available directly on the cart page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sVmd0MszjpkHlidIhUXtLg.png" /><figcaption>Product Highlight</figcaption></figure><h3>The Impact: Improved User Experience and Higher Conversion Rates</h3><p>The results of IDEAL Indonesia’s efforts have been impressive. After implementing <strong>these</strong> <strong>changes (late 2023)</strong>, the platform saw a significant increase in user engagement and conversion rates:</p><ul><li><strong>Higher Conversion Rates</strong>: The percentage of registered users who completed a mortgage application increased from 6.71% to 14.93%.</li><li><strong>More Educated Users</strong>: 92.21% of applicants applied with two or more mortgage products, indicating that users were better informed and more confident in their choices.</li><li><strong>Reduced Drop-Off Rates</strong>: The drop-off rate on the product page decreased by 29.33%, showing that users found the new design more intuitive and easier to navigate.</li></ul><h3>Conclusion: A Step Forward in Financial Inclusion</h3><p>IDEAL Indonesia’s approach to <strong>simplifying</strong> <strong>mortgage</strong> selection is a testament to the power of user-centric design. By making the process more transparent and educational, the platform has not only improved the user experience but also increased the likelihood of users finding and securing the right mortgage product for their needs.</p><p>For a country like Indonesia, where financial literacy and access to banking services are still growing, IDEAL’s platform is a <strong>significant</strong> <strong>step</strong> <strong>forward</strong> in making <strong><em>homeownership</em></strong><em> </em><strong><em>more</em></strong><em> </em><strong><em>accessible</em></strong><em> </em><strong><em>to</em></strong><em> </em><strong><em>everyone</em></strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ddc865b80791" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Most Important Key Metric: Time to Value (TTV)]]></title>
            <link>https://medium.com/@fahrellgiovanny/most-important-key-metric-time-to-value-ttv-9b6bbec28275?source=rss-967249932650------2</link>
            <guid isPermaLink="false">https://medium.com/p/9b6bbec28275</guid>
            <category><![CDATA[timetovalue]]></category>
            <category><![CDATA[north-star]]></category>
            <category><![CDATA[product-development]]></category>
            <category><![CDATA[key-metrics]]></category>
            <category><![CDATA[product-manager]]></category>
            <dc:creator><![CDATA[Fahrell Giovanny]]></dc:creator>
            <pubDate>Mon, 24 Apr 2023 14:52:40 GMT</pubDate>
            <atom:updated>2023-04-24T14:52:40.109Z</atom:updated>
            <content:encoded><![CDATA[<h4>As PMs, our ultimate goal is to create products that provide value to our customers. But how do we measure that value?</h4><p>One metric that has become increasingly popular in recent years is <em>Time to Value (TTV)</em>. TTV measures the time it takes for a customer to receive the core value of your product after they begin using it. Let’s explore why TTV is critical, how to measure it, how to use it to guide your product development journey, and finally how to reduce it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bV8hEJtWGB1SDx6tm25fVA.png" /><figcaption>TTV</figcaption></figure><h3>Why Time to Value Matters</h3><p>Time to Value is an essential metric for any PMs to track because it directly correlates to customer satisfaction and retention. Customers who quickly see and experience the value of the product are more likely to continue using it and refer it to others. Contrarily, customers who don’t know the value immediately will likely stop using your product and eventually leave negative reviews.</p><p>In today’s competitive marketplace, it has become increasingly important for us to deliver results from specific products and services immediately. As such, companies that can deliver value quickly have a competitive advantage and are more likely to succeed in the long term.</p><p>In addition, tracking TTV can help you identify areas where your product may be falling short. For example, if customers take a long time to see the value, it may be because your onboarding process needs to be simplified or it does not self-explanatory.</p><h4>How to Measure Time to Value</h4><p>Measuring TTV requires a deep understanding of your product and your customers. Here are the steps to follow:</p><ol><li><strong><em>Define what “value” means for your entire product journey.</em><br></strong>Before measuring TTV, you need to define what are the things that constitute value for your product. This could be anything from achieving a specific outcome to completing a task.</li><li><strong><em>Identify the key actions that lead to the value.</em><br></strong>Once you’ve defined the core value, identify the key steps customers need to take to achieve it. For example, if the value is completing a product purchase, the key activities might be adding a product, completing a payment, and finally receiving the product.</li><li><strong><em>Measure the time it takes users to complete these actions.</em><br></strong>Once you’ve identified the key steps, measure the time customers take to achieve them. This will give you your baseline TTV.</li></ol><h4><strong>How to Use TTV</strong></h4><p>Now that you know how to measure TTV, the next step is to use it to guide your product development journey.</p><ol><li><strong><em>Identify areas for improvement.</em></strong><br>Use your data to identify areas where your product could be improved. For example, if customers are taking a long time to complete a necessary action, it may be because it has a bad user experience. You can then work to streamline the process and reduce TTV.</li><li><strong><em>Set goals.</em></strong><br>Set TTV goals for your product and track your progress towards them. This will help you continuously improve and deliver the best value to your customers.</li><li><strong><em>Use TTV as a competitive advantage.</em></strong><br>If your product has a faster TTV than your competitors, use it as a competitive advantage in your marketing and sales efforts. However, continuous improvement is needed to ensure your TTV is faster.</li></ol><h4>How to reduce time to value:</h4><ol><li><strong><em>Streamline processes:</em></strong><em> </em><br>Identify any bottlenecks in your processes and streamline them to reduce delays and increase efficiency. This can include eliminating unnecessary steps, automating user tasks, giving the right recommendations, and improving user communications.</li><li><strong><em>Optimize features.</em></strong><br>Focus on the features of your product or service that deliver the most value to your customers and optimize them to make them even more effective. This may involve adding new features or improving existing ones.</li><li><strong><em>Enhance the user experience.</em></strong><br>Ensure your product or service is easy to use and understand. Consider conducting user testing to identify pain points and areas for improvement and then implement changes to improve the overall user experience.</li><li><strong><em>Prioritize customer feedback.</em></strong><br>Listen to your customers and incorporate their feedback into your product development process. This can help you identify areas where you can improve your product or service to meet their needs better and deliver value more quickly.</li><li><strong><em>Use data analytics.</em></strong><br>Utilize data analytics to gain insights into customer behavior and preferences. This can help you identify trends and patterns that can inform your product development and marketing strategies, ultimately helping you reduce time to value.</li><li><strong><em>Be creative.</em></strong><br>Encourage the teams to think outside the box and develop new ideas to improve your product or service. This can help you stay ahead of the competition and deliver value quickly to your end users.</li></ol><h4>Conclusion</h4><p>Time to Value is a critical metric for product managers to track. It measures the time it takes for customers to see the value of your product after they begin using it. By tracking TTV, you can identify areas for improvement, set goals, and use them as a competitive advantage. Reducing time to value is often achieved by streamlining processes, optimizing features, and enhancing user experience.</p><p>By doing so, businesses can deliver value to their customers more quickly and effectively. Focusing on TTV will help you create products that provide value to your customers and drive business growth.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9b6bbec28275" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The LinkedIn Swindler]]></title>
            <link>https://medium.com/@fahrellgiovanny/the-linkedin-swindler-82f35186e3a8?source=rss-967249932650------2</link>
            <guid isPermaLink="false">https://medium.com/p/82f35186e3a8</guid>
            <category><![CDATA[linkedin-scammer]]></category>
            <category><![CDATA[indonesia]]></category>
            <category><![CDATA[linkedin-swindler]]></category>
            <category><![CDATA[hiring-tips]]></category>
            <category><![CDATA[linkedin]]></category>
            <dc:creator><![CDATA[Fahrell Giovanny]]></dc:creator>
            <pubDate>Sun, 31 Jul 2022 15:59:56 GMT</pubDate>
            <atom:updated>2022-08-07T12:25:25.146Z</atom:updated>
            <content:encoded><![CDATA[<h4>It is happening in Indonesia, and we just didn&#39;t care enough.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xXS9hNod3-8IJtXac-Z_Zw.jpeg" /></figure><p><strong>The</strong> new and highly rated Netflix documentary &quot;The Tinder Swindler&quot; was released in early 2022. The film was about a conman who posed on Tinder&#39;s dating app as a wealthy jet setter and diamond trader as a status symbol. Using his Tinder profile, he accumulated wealth and scammed many women to help him in a made-up crisis.</p><p>Wealth and high status are attractive. That is why not only can we find the swindler for Tinder but also <em>&quot;The LinkedIn Swindler.&quot;</em></p><p><em>What are your typical LinkedIn Swindler red flags?</em></p><blockquote><strong>They scam people into believing that they can do something that they are not capable of to help them secure a job</strong></blockquote><p>It is pretty self-explanatory. The swindlers have actual employment. However, they use glorified and exaggerated job descriptions on LinkedIn to lure the recruiters. Followed by excellent communication skills to convince the recruiter further that they could do what they wrote in the job description.</p><p>Interviews biased positively towards candidates who were articulate and could bullshit well. While that&#39;s a good skill, it&#39;s not 100% correlated to success <a href="https://www.xendit.co/en/blog/the-argument-for-a-trial-day/">(Xendit, 2020)</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j1fQEwchRkqGPgBqFtLyQQ.png" /><figcaption>Exhibit 1</figcaption></figure><p>The job positions and descriptions might initially look normal. But after a thorough read, you might find that there are lots of inconsistencies. For example, in the swindler&#39;s internship position, he works at Icon+. However, the description says that it was from Accenture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V2K-BxBieFPRbzN0PMmoxA.png" /><figcaption>Exhibit 2</figcaption></figure><p>As you can see, the swindler also did another internship for Sierra Solutions Group for three months. However, the description says that he has worked for three years. Additionally, the keywords the swindler mentioned are not the tech stack and tools used by the company. Therefore all of it is gibberish and jargon words just to beautify the resume.</p><blockquote><strong>They work for multiple companies at once</strong></blockquote><p>Because of the lack of skills, the swindlers know that working for one company and working for a long term is not their modus operandi. The idea is not to work until the company notices and terminates their employment. So, it would make sense for them to have multiple employers.</p><p>The implication of that is a lot. The swindlers cannot deliver to any company expectation (in which they do not have the skills to do well), thus, leaving the company hiring them empty-handed. Usually, the swindlers will come up with many excuses not to work, such as family issues, health problems, real-life conflicts, etc. Ultimately, they will 100% refuse to work in the office as this will expose their operations.</p><blockquote><strong>Fake positions especially when they are able to be employed on legit companies</strong></blockquote><p>It is easier to find a job when you have excellent working experience. This can also create a good impression when the interview happens. So, it will make the swindlers generate more reach and interest from the recruiter.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pAUh8pP-NkEinnBvHpin-g.png" /><figcaption>Exhibit 3</figcaption></figure><p>After only working since July 2018, he has already become an Engineering Manager on Traveloka. Not <em>suspicious</em> at all, right?</p><p>My ex-employer recently hired 2 LinkedIn swindlers as well (Both of them are QA Engineers). At first, my company realized its need for an employee in a particular field. Unfortunately, when the supposedly needed employees arrived, the swindlers &quot;hired&quot; were just there without any real contribution to the company.</p><p>So, what can your company do to <strong>prevent</strong> and <strong>minimize</strong> the risk of this phenomenon?</p><ol><li><strong><em>Do background checks</em></strong><em>, always (Grab use </em><a href="https://grab.careers/privacy-policy/"><em>3rd Party Background Check</em></a><em>).</em></li><li><strong>Reference Check </strong><em>(Tokopedia requires reference checks)</em>.</li><li><strong>Technical Test </strong>to reflect the potential candidate&#39;s skills.</li><li><strong>Work Trial </strong><em>(Xendit did a </em><a href="https://www.xendit.co/en/blog/the-argument-for-a-trial-day/"><em>three-day working trial</em></a><em>)</em>.</li><li><strong>Use GetContact</strong> (it works wonder)</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3rUOCa_bR2KCk3Q65RTiyQ.png" /><figcaption>Exhibit 4</figcaption></figure><p>I am not against working in multiple places at once. On the contrary, I respect people who can do so. I appreciate those who can multitask while still possessing a good focus on what they do and can deliver extraordinary results in the end.</p><p>I fully condemn the ones who abuse the systems, the irresponsible ones whose goals are only to gain profit when the company that hires them suffers from a significant loss.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gVRzF4-LmfrY0Wc2BzVOVg.png" /><figcaption>Exhibit 5 (si paling sukses)</figcaption></figure><p><em>Click the </em>👏🏻 <em>below, so other people will see this on Medium.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=82f35186e3a8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Tokopedia Build Their First Graph Database]]></title>
            <link>https://medium.com/tokopedia-product-team/how-tokopedia-build-their-first-graph-database-a3f4957eb70c?source=rss-967249932650------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3f4957eb70c</guid>
            <category><![CDATA[tokopedia]]></category>
            <category><![CDATA[product-deep-dive]]></category>
            <category><![CDATA[graph-database]]></category>
            <category><![CDATA[fraud]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Fahrell Giovanny]]></dc:creator>
            <pubDate>Fri, 21 Feb 2020 07:23:15 GMT</pubDate>
            <atom:updated>2024-11-20T17:36:20.783Z</atom:updated>
            <content:encoded><![CDATA[<h4>Learning points in developing Graph Technology for better fraud detection</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j-qR4BA1St2xO-8DUyI0Sg.jpeg" /></figure><p>Managing vast data points as a technology company can be complicated. Whether it’s making faster, more informed business decisions, improving flexibility to react to uncertainties, or increasing awareness of risks, having structured and connected data is crucial. A graph database excels in these areas by enabling organizations to create robust preventive measures, especially for fraud detection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f6y9LuJPlQcHMT6jsxqPlw.png" /><figcaption><a href="https://www.tokopedia.com/"><strong>Graph Visualization Example of Buyer and Seller Transaction in Tokopedia</strong></a></figcaption></figure><blockquote>What is a graph database?</blockquote><p>A graph database is a structure that manages relationships between data as equally important as the data itself. One common method is the property graph model, which consists of:</p><p>Nodes: Represent entities like users, products, or devices.<br>Relationships: Connections between nodes (e.g., “BOUGHT,” “REFERRED”).<br>Properties: Metadata describing nodes and relationships (e.g., timestamps, quantities).</p><p><strong>Tokopedia’s Journey with Graph Databases</strong><br><em>From Experimentation to Adoption</em></p><p>In December 2018, Tokopedia Risk Management began its graph database journey, initially using an experimental project known as The Graph. Unfortunately, it was riddled with bugs, making us the “test engineers” for an unstable tool. By May 2019, we decided to abandon The Graph.</p><p>By September 2019, we explored alternative graph database solutions, settling on a more popular and reliable option. As its usage grew, in 2020, we took a big leap of faith, dedicating a squad to expand graph database applications within Tokopedia Risk Management.</p><blockquote>Here are some <strong>key takeaways </strong>over the course developing Graph Database in Tokopedia.</blockquote><p><strong>We Have to Create a Specific Relation Name</strong></p><p>Initially, we used a generic relationship name, “<strong>HAS</strong>,” for all connections. This limited our ability to analyze specific relationships. Migrating to meaningful relationship names like “<strong>BOUGHT</strong>” or “<strong>REFERRED</strong>” unlocked deeper insights.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/987/1*bNdMbjewPZec5JgROkpNKg.png" /><figcaption><strong>Old Relation in Tokopedia Graph Database (Using Has)</strong></figcaption></figure><p><strong>Deadlock Might Happen</strong></p><p>Graph databases can face deadlocks when inserting large volumes of data simultaneously. Plan for this by managing transaction queues effectively.</p><p><strong>Unstructured Data, Different Way to Query</strong></p><p>Graph databases differ from SQL; they lack fixed schemas and avoid joins. Learning NoSQL querying concepts (e.g., Cypher for Neo4j) is essential.</p><p><strong>Resource Intensive</strong></p><p>Graph databases struggle with database-wide queries and heavy workloads, requiring additional server and RAM resources to maintain performance.</p><p><strong>Community vs. Enterprise</strong></p><p>While community editions are free, enterprise editions are more scalable and better suited for commercial needs. Evaluate based on your requirements.</p><p><strong>Data Can Be Dangerous</strong></p><p>With great data comes great responsibility. Log all activities involving your graph database to maintain transparency and accountability.</p><p><strong>Limitless Use Case</strong></p><p>Graph databases are not just for fraud detection. They’re used globally for real-time recommendation engines, identity management, social networks, artificial intelligence, and knowledge graphs. A famous example is the <strong>Panama Papers</strong>, where graph databases unveiled hidden relationships between offshore bank accounts, exposing assets of politicians, celebrities, and even royalty.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KzSqRD9sWUUoNrSmmmfEqg.png" /><figcaption><a href="https://medium.com/neo4j/graph-visualization-of-panama-papers-data-in-neo4j-9c08ca17039c"><strong><em>Panama Papers Graph Visualization</em></strong></a></figcaption></figure><p>At Tokopedia, we leveraged graph databases primarily for fraud detection but have extended its use to detect nuanced behaviors like <strong>Cross Buying</strong>, <strong>Cluster Cross Buying</strong>, and <strong>Concentrated Purchases</strong>.</p><h3>Detecting Fraud and Purchase Behaviors with Graph Databases</h3><h4>1. Cross Buying</h4><p>Cross buying occurs when a user purchases across multiple product categories, such as electronics, groceries, and fashion. While often legitimate, abnormal patterns can signal coordinated abuse.</p><ul><li><strong>How Graph Databases Help</strong>:<br>Graph databases map user-product relationships, linking users to purchased categories. Queries can detect unusual diversity in purchases during short timeframes.</li></ul><h4>2. Cluster Cross Buying</h4><p>Cluster cross buying happens when a group of users (a fraud ring) shows coordinated cross-category purchases to exploit rewards.</p><ul><li><strong>How Graph Databases Help</strong>:<br>By connecting users through shared attributes like IP addresses or devices, graph databases can detect dense clusters exhibiting similar buying behaviors.</li></ul><h4>3. Concentrated Purchases</h4><p>Concentrated purchases involve bulk buying of specific products, often exploiting volume discounts or limited-time promotions.</p><ul><li><strong>How Graph Databases Help</strong>:<br>Graph databases link users to purchased products and quantify transaction volumes, flagging unusually high purchases.</li><li><strong>Cross-Cluster Insight</strong>:<br>Overlay transaction patterns with shared attributes like devices or payment methods to uncover fraud rings coordinating bulk purchases.</li></ul><h3>Real-World Fraud Detection Scenarios</h3><ol><li><strong>Coordinated Referral Abuse</strong>:<br>Fraud rings create multiple accounts using shared devices or IPs, refer each other, and trigger rewards. Graph databases can expose referral loops and dense clusters.</li><li><strong>Promotional Exploitation</strong>:<br>By detecting shared attributes across accounts exploiting cross-category promotions, businesses can flag abuse early.</li><li><strong>Reward Claim Patterns</strong>:<br>Analyze claim nodes to identify users or clusters repeatedly claiming high-value rewards during short periods.</li></ol><p>No matter your business stage or challenges, there’s no need to hesitate when exploring the potential of graph databases. Their unique ability to analyze relational data can empower your company to learn and scale more efficiently than ever before.</p><p>At Tokopedia, graph databases transformed our fraud detection approach, uncovering hidden relational patterns that distinguished legitimate users from fraudulent ones. They allowed us to identify sophisticated behaviors like Cross Buying and refine our fraud prevention strategies for greater impact.</p><p>Whether you’re combating fraud, developing recommendation systems, or managing complex data relationships, graph databases offer a powerful, scalable solution to address these challenges effectively.</p><p><em>I would love to hear about what experience and idea you have encountered while developing graph databases.</em><br><em>Feel free to contact me at </em><strong><em>fahrell.giovanny@tokopedia.com</em></strong></p><p><em>Special shoutout to:<br> Alluka Team — Yuga, Rian, and Josua<br>‘Technical Architect’ — Zaki and ‘Principal Engineer’ — Tjandrayana</em></p><p><em>Click the </em>👏🏻 <em>below so other people will see this here on Medium.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3f4957eb70c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/tokopedia-product-team/how-tokopedia-build-their-first-graph-database-a3f4957eb70c">How Tokopedia Build Their First Graph Database</a> was originally published in <a href="https://medium.com/tokopedia-product-team">Tokopedia Product</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>