<?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[Excella Labs / Technology - Medium]]></title>
        <description><![CDATA[New ideas, experiments, and experiences by Excella’s Software Engineering, UX, DevSecOps, AI, and Data Xperts. - Medium]]></description>
        <link>https://medium.com/excella-labs-engineering?source=rss----cdbe8ef7f205---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Excella Labs / Technology - Medium</title>
            <link>https://medium.com/excella-labs-engineering?source=rss----cdbe8ef7f205---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 14:02:38 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/excella-labs-engineering" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Five Things from DORA You Can Implement Now for Maximum Impact]]></title>
            <link>https://medium.com/excella-labs-engineering/five-things-from-dora-you-can-implement-now-for-maximum-impact-1f74cefc3b71?source=rss----cdbe8ef7f205---4</link>
            <guid isPermaLink="false">https://medium.com/p/1f74cefc3b71</guid>
            <category><![CDATA[security]]></category>
            <category><![CDATA[dora]]></category>
            <category><![CDATA[efficiency]]></category>
            <category><![CDATA[impact]]></category>
            <dc:creator><![CDATA[Cameron Hopkin]]></dc:creator>
            <pubDate>Tue, 25 Mar 2025 11:17:10 GMT</pubDate>
            <atom:updated>2025-03-25T11:17:10.519Z</atom:updated>
            <content:encoded><![CDATA[<p>In the fast-paced world of software development, we’re constantly searching for ways to deliver better products faster while keeping costs under control. Suppose you’ve been in tech for even a short time. In that case, you’ve likely heard of DORA — the DevOps Research and Assessment team whose groundbreaking research has changed how we think about high-performing technology organizations.<br>It’s important to note that implementing DORA’s key findings doesn’t necessitate a massive organizational overhaul. Some of the most impactful changes can be made with relatively modest investments. Let’s explore five DORA insights you can implement today. This will not only boost efficiency, strengthen your security posture, and deliver measurable cost savings but also empower you with the capability to make a significant impact.</p><figure><img alt="A woman working in an office with graphs on the screen. AI Generated Image" src="https://cdn-images-1.medium.com/max/1024/1*gyKcBrNcVpW3DogtDs-fow.jpeg" /></figure><h3>Why DORA Metrics Matter More Than Ever</h3><p>Before discussing specific actions, let’s address the elephant in the room: Why should you care about DORA metrics in today’s economic climate?</p><p>The answer is straightforward. When budgets tighten, the pressure to deliver value intensifies. Organizations can no longer afford the luxury of inefficient processes or wasteful practices. DORA’s research gives us a data-driven framework directly correlating certain practices with improved performance and reduced costs.</p><p>In 2023, organizations classified as ‘elite performers’ by DORA standards deployed code 973 times more frequently than low performers while maintaining superior stability. More importantly, these elite performers recovered from incidents 6,570 times faster. The financial implications become crystal clear when you consider the average cost of downtime (approximately $5,600 per minute, according to Gartner). This underscores the urgency and importance of implementing DORA’s insights.</p><h3>1. Implement Disciplined Trunk-Based Development for Faster Delivery and Fewer Conflicts</h3><p>One of DORA’s most consistent findings is the strong correlation between trunk-based development and high performance. Yet many teams use complex branching strategies that delay integration and create merge conflicts.</p><p><strong>What you can do today:</strong> Begin transitioning to a trunk-based development approach by:</p><ul><li>Limiting branch lifetimes to less than one day</li><li>Setting up comprehensive automated tests that run on every commit</li><li>Using feature toggles judiciously with a clear management strategy</li></ul><p>While feature toggles (or flags) can be powerful tools for implementing trunk-based development, teams must carefully manage these toggles to avoid creating technical debt. For each toggle you implement, make a clear plan for its eventual removal, use consistent naming conventions, and implement proper monitoring. Remember that unmanaged feature toggles can become liabilities over time.</p><p>Organizations that adopt disciplined trunk-based development with well-managed feature toggles can significantly reduce integration issues while increasing deployment frequency. The security benefit is equally compelling: shorter-lived branches mean vulnerabilities get detected and fixed faster, reducing your window of exposure to potential attacks.</p><h3>2. Automate Your Change Approval Process</h3><p>DORA’s research consistently shows that external change approval processes (like traditional Change Advisory Boards) negatively correlate with performance. Yet despite the clear evidence, many organizations still rely on these manual gates.</p><p><strong>What you can do today:</strong></p><ul><li>Replace manual approvals with automated quality gates</li><li>Implement automated policy checks in your deployment pipeline</li><li>Use security scanning tools that run automatically on code changes</li></ul><p>The potential cost savings here are significant. First, you eliminate the hours spent in change approval meetings (on average, 4–8 hours per week for senior engineers). Second, you accelerate your delivery timeline, bringing value to market faster. One government client we worked with saved an estimated $340,000 annually by replacing their CAB process with automated controls, while simultaneously strengthening their security posture through consistent enforcement of policies. This should inspire and motivate you to implement these changes.</p><h3>3. Embrace Observable Systems as Your Security Foundation</h3><p>Elite performers spend 50% less time remediating security issues according to DORA’s research. The key difference? They build observability into their systems from the ground up. Observable systems are those that provide clear, real-time insights into their operation, making it easier to detect and respond to security issues.</p><p><strong>What you can do today:</strong></p><ul><li>Implement structured logging across all applications</li><li>Deploy distributed tracing for critical user journeys</li><li>Create dashboards that highlight security anomalies</li></ul><p>Observability isn’t just about performance monitoring — it’s about establishing an omnipresent security foundation that detects unusual patterns before they become breaches. One retail client implemented these changes and identified an attempted data exfiltration that their traditional security tools missed. Beyond security, their improved observability reduced mean time to resolution by 72%, translating to approximately $180,000 in annual savings from reduced downtime.</p><h3>4. Make Deployment Pain Visible</h3><p>DORA found that teams who measure deployment pain, which refers to the difficulties and challenges experienced during the deployment process, and actively work to reduce it show dramatically improved performance across all metrics. Yet most organizations don’t systematically track this critical indicator.</p><p><strong>What you can do today:</strong></p><ul><li>Survey team members after each deployment about pain points</li><li>Create a “deployment pain” dashboard visible to leadership</li><li>Allocate 20% of each sprint to addressing the highest pain areas</li></ul><p>Making deployment friction visible creates organizational motivation to address these issues. A healthcare client implemented this approach and discovered that nearly 40% of their deployment time was spent on manual database change scripts. By automating this process, they reduced deployment times by 65% and eliminated a significant source of production incidents, saving approximately $290,000 annually in combined costs from faster deployments and reduced outages.</p><h3>5. Implement Team-Level Value Stream Mapping</h3><p>DORA’s research highlights the importance of understanding and optimizing the entire value stream, which is the sequence of activities that deliver a product or service to a customer, yet many organizations remain blind to where time and resources are actually being spent.</p><p><strong>What you can do today:</strong></p><ul><li>Conduct a simplified value stream mapping exercise with each team</li><li>Identify the top three bottlenecks in your delivery process</li><li>Create visible metrics around these bottlenecks and set improvement targets</li></ul><p>Stream mapping sounds simple, but the impact can be profound. A manufacturing client discovered through this exercise that their QA process consumed 68% of their total delivery time but only found issues in 3% of releases. By reimagining their testing approach with automated testing first, they reduced their overall delivery cycle by 58% while improving quality, resulting in approximately $420,000 in annual development cost savings.</p><h3>Understanding the ROI of DORA Implementation</h3><p>The DevOps Research and Assessment (DORA) metrics offer an objective framework for measuring return on investment and business impact of initiatives like the ones we’ve discussed. While specific numbers vary by organization, research consistently shows that high-performing organizations implementing DORA principles outperform their peers in delivery speed and stability.</p><p>Organizations that embrace these practices reduce costs through efficiency gains and increase their ability to respond to market changes and deliver value faster. As Jeff Gallimore, Excella’s Chief Technology and Innovation Officer, has discussed in various forums, including the Federal Tech Podcast, these improvements translate directly to better organizational performance and competitive advantage.</p><p>The financial benefits come from multiple sources: reduced development time, fewer production incidents, faster failure recovery, and the ability to deliver features to market more quickly. The exact ROI will depend on your organization’s starting point and implementation approach, but the data is precise, and these investments pay off substantially.</p><h3>Beyond the Numbers: The Cultural Shift</h3><p>While the financial benefits are compelling, the most profound impact of implementing these DORA recommendations is often cultural. Teams that embrace these practices report higher job satisfaction, reduced burnout, and improved collaboration between development and operations.</p><p>In an industry where talent acquisition and retention costs continue to rise, creating an environment where technical professionals can do their best work becomes a competitive advantage.</p><h3>Taking the First Step</h3><p>The journey to elite performance doesn’t happen overnight but starts with a single step. Choose one of these five areas that resonates most with your current challenges and begin there. Measure your baseline, implement the changes, and track the improvement.</p><p>The data is precise: organizations that embrace DORA’s findings outperform their peers in delivery speed, operational stability, security posture, and cost efficiency. The question isn’t whether you can afford to implement these changes — it’s whether you can afford not to.</p><p><em>What DORA-inspired changes have made the most significant impact in your organization? I’d love to hear your experiences in the comments.</em></p><h3>References</h3><ul><li>Google Cloud. (2023). <em>Accelerate State of DevOps Report 2023</em>. Retrieved from <a href="https://dora.dev/research/2023/dora-report/">https://dora.dev/research/2023/dora-report/</a></li><li>Google Cloud Blog. (2023, October 5). <em>Announcing the 2023 State of DevOps Report</em>. Retrieved from <a href="https://cloud.google.com/blog/products/devops-sre/announcing-the-2023-state-of-devops-report">https://cloud.google.com/blog/products/devops-sre/announcing-the-2023-state-of-devops-report</a></li><li>DevOps.com. (2021, September 21). <em>Google DORA Report Shows Modest DevOps Gains</em>. Retrieved from <a href="https://devops.com/google-dora-report-shows-modest-devops-gains/">https://devops.com/google-dora-report-shows-modest-devops-gains/</a></li><li>GitLab. (2021, October 26). <em>How to make your DevOps team elite performers</em>. Retrieved from <a href="https://about.gitlab.com/blog/2021/10/26/how-to-make-your-devops-team-elite-performers/">https://about.gitlab.com/blog/2021/10/26/how-to-make-your-devops-team-elite-performers/</a></li><li>Atlassian. (n.d.). <em>Calculating the cost of downtime</em>. Retrieved from <a href="https://www.atlassian.com/incident-management/kpis/cost-of-downtime">https://www.atlassian.com/incident-management/kpis/cost-of-downtime</a></li><li>Atlassian. (n.d.). <em>Calculating the cost of downtime</em>. Retrieved from <a href="https://www.atlassian.com/incident-management/kpis/cost-of-downtime">https://www.atlassian.com/incident-management/kpis/cost-of-downtime</a></li><li>Federal Tech Podcast. (n.d.). <em>Jeff Gallimore on DORA metrics for federal IT systems</em>. Retrieved from <a href="https://podcasts.apple.com/us/podcast/ep-221-measuring-what-matters-evaluating-success-in/id1612819978?i=1000698697807">https://www.excella.com/</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1f74cefc3b71" width="1" height="1" alt=""><hr><p><a href="https://medium.com/excella-labs-engineering/five-things-from-dora-you-can-implement-now-for-maximum-impact-1f74cefc3b71">Five Things from DORA You Can Implement Now for Maximum Impact</a> was originally published in <a href="https://medium.com/excella-labs-engineering">Excella Labs / Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Protecting Your Digital Life]]></title>
            <link>https://medium.com/excella-labs-engineering/protecting-your-digital-life-4e157e244319?source=rss----cdbe8ef7f205---4</link>
            <guid isPermaLink="false">https://medium.com/p/4e157e244319</guid>
            <category><![CDATA[security]]></category>
            <category><![CDATA[privacy]]></category>
            <category><![CDATA[digital-life]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[self-protection]]></category>
            <dc:creator><![CDATA[Cameron Hopkin]]></dc:creator>
            <pubDate>Mon, 30 Dec 2024 18:20:23 GMT</pubDate>
            <atom:updated>2024-12-30T18:20:23.690Z</atom:updated>
            <content:encoded><![CDATA[<h3>Protecting Your Digital Life: A Guide to Privacy and Security for Everyone</h3><p><em>A straightforward guide to online safety for everyone</em></p><figure><img alt="3D image of a digital lock symbolizing strong consumer privacy and security. It highlights the idea of impenetrability against external attempts to breach it." src="https://cdn-images-1.medium.com/max/1024/0*cKYkoAHCRf54lKsS" /><figcaption><em>Photo generated with ChatGPT 4o</em></figcaption></figure><p>In today’s world, our lives are increasingly connected to the internet. While this brings tremendous convenience, it also opens doors to potential risks. Whether you’re tech-savvy or just learning to use a smartphone, protecting your privacy and security is essential. The good news? You don’t need to be a technology expert to keep yourself safe online.</p><h3>Why Should You Care About Digital Privacy?</h3><p>Think of your digital privacy like your home security. You must protect your digital information just as you lock your doors and windows. Scammers and cybercriminals are constantly finding new ways to access personal information, from bank accounts to family photos. In 2023 alone, Americans lost over $10 billion to online scams, with older adults often targeted most aggressively.</p><h3>Essential Security Steps Anyone Can Take</h3><h3>1. Protecting Your Email — Your Digital Front Door</h3><p>Your email account is like the front door to your digital life. If someone gains access to it, they can access your other accounts by requesting password resets. Here’s how to secure it:</p><h3>Setting Up Strong Authentication</h3><p>Two-factor authentication (2FA) adds an extra layer of security beyond your password. Think of it as having a key and an alarm code for your house. Here’s how to set it up safely:</p><p><strong>Best Option: Authentication Apps</strong></p><p>Think of an authentication app as a unique key that changes every 30 seconds. It’s much safer than getting codes through text messages, and it’s easier than it sounds:</p><ol><li>Download an authentication app like Google Authenticator or Authy on your smartphone.</li><li>When logging in, you’ll enter your password plus a code from the app.</li><li>Family members can help set this up during their next visit</li></ol><p><strong>If You Don’t Have a Smartphone:</strong></p><p>Don’t worry — there’s a great alternative:</p><ol><li>Consider using a physical security key (like a YubiKey)</li><li>These are small devices that plug into your computer.</li><li>They’re highly secure and simple to use</li><li>Many cost around $25–50, but their security is worth the investment.</li></ol><h3>Email Privacy Services</h3><p>Services like SimpleLogin or AnonAddy let you create “proxy” email addresses that forward to your actual email. Think of it as having a personal mail forwarding service that protects your actual address. Here’s why it’s helpful:</p><ol><li>You can create different email addresses for various services.</li><li>If one service is compromised, your actual email stays private.</li><li>You can quickly delete compromised addresses without affecting your primary email.</li><li>It’s like having a spam filter and privacy shield in one</li></ol><h3>2. Protecting Your Credit — A Powerful Free Tool</h3><p>One of the strongest defenses against financial fraud is entirely free: credit freezes. Here’s what you need to know:</p><h3>What is a Credit Freeze?</h3><p>A credit freeze is like putting a strong lock on your credit file. When it’s frozen:</p><ol><li>No one can open new accounts in your name.</li><li>It’s completely free by law.</li><li>You can temporarily lift it when you need to apply for credit.</li><li>It doesn’t affect your credit score or existing accounts</li></ol><h3>How to Freeze Your Credit:</h3><p>The process is straightforward and can be done in an afternoon:</p><ol><li>Contact Equifax: Visit Equifax.com or call 1–800–349–9960</li><li>Contact Experian: Visit Experian.com or call 1–888–397–3742</li><li>Contact TransUnion: Visit TransUnion.com or call 1–888–909–8872</li></ol><p>Keep your PINs or passwords safe — you’ll need them to unfreeze your credit later.</p><h3>3. Daily Security Habits That Make a Difference</h3><h3>Creating Strong Passwords</h3><p>Think of passwords like keys to your digital doors. Here’s how to make them strong:</p><ol><li>Use a password manager like Bitwarden (it’s free) to generate and store complex passwords.</li><li>Never reuse passwords across different accounts.</li><li>Consider writing down your master password and storing it in a secure location.</li><li>Use long phrases that are meaningful to you but hard for others to guess</li></ol><h3>Spotting Scams</h3><p>Protecting yourself from scams is like developing a sixth sense. Here are the warning signs:</p><ol><li>Be wary of urgent messages demanding immediate action.</li><li>Never give personal information over the phone unless you initiate the call.</li><li>If something seems too good to be true, it probably is</li><li>When in doubt, ask a family member or friend for a second opinion</li></ol><h3>Recovering from a Scam: You’re Not Alone</h3><p>First and foremost, please know that falling for a scam doesn’t mean you’re foolish. Today’s scammers are sophisticated professionals who use advanced psychological tactics and technology. They’ve fooled doctors, lawyers, tech professionals, and even cybersecurity experts. Feeling embarrassed is natural, but remember: the fault lies with the criminals, not you.</p><p>Many people keep quiet about being scammed because they feel ashamed, but speaking up is powerful. It helps warn others and enables you to recover. Sharing your experience might save someone else from the same scam. Remember, scammers succeed not because their victims are naive but because they’re often kind, trusting, and willing to help others — all admirable qualities that scammers exploit.</p><h3>Taking Action After a Scam</h3><p>Even with the best precautions, problems can occur. Here’s your step-by-step recovery plan:</p><ol><li>Don’t panic — there are people that can help.</li><li>Contact your bank or credit card company immediately if you suspect fraud.</li><li>Report scams to the FTC at ftc.gov/complaint</li><li>Change passwords for any affected accounts.</li><li>Consider reaching out to family members or trusted friends for support</li></ol><h3>Getting Help and Support</h3><p>Remember, you’re not alone in this. There’s both practical and emotional support available:</p><ol><li>Victim Support Services: Many communities have free counseling services for fraud victims</li><li>Local libraries often offer free tech security classes.</li><li>Senior centers frequently host cybersecurity workshops.</li><li>Adult children or grandchildren can often help set up security measures.</li><li>Many banks offer free security consultations for customers.</li><li>AARP Fraud Watch Network provides free support services, including emotional support</li><li>Local police departments often have cybercrime units that can offer guidance.</li></ol><h3>Building Confidence After a Scam</h3><p>Recovery isn’t just about securing your accounts — it’s also about rebuilding your confidence. Here are some empowering steps you can take:</p><ol><li>Become a Fraud Fighter: Consider joining local consumer protection groups or online forums where you can help warn others about scams.</li><li>Share Your Story: When you’re ready, sharing your experience with trusted friends or family can be healing and help protect others.</li><li>Take It Step by Step: To rebuild your online confidence, start by taking small actions, such as updating your passwords or learning new security features.</li><li>Remember Your Strengths: The qualities that scammers exploit — kindness, trust, helpfulness — are valuable human traits. Don’t let scammers make you lose these parts of yourself.</li></ol><h3>Supporting Others</h3><p>If someone you know has been scammed:</p><ol><li>Local libraries often offer free tech security classes.</li><li>Senior centers frequently host cybersecurity workshops.</li><li>Adult children or grandchildren can often help set up security measures.</li><li>Many banks offer free security consultations for customers</li></ol><h3>Looking Ahead: Staying Secure in a Changing World</h3><p>Technology will continue to evolve, but the basic principles of security remain the same:</p><ol><li>Be cautious about sharing personal information.</li><li>Use strong authentication methods.</li><li>Keep software and devices updated.</li><li>Have a plan for when things go wrong.</li></ol><p>Remember: You don’t have to tackle this alone. Many libraries offer free tech security classes, and family members are often happy to help set up security measures. The most crucial step is getting started, even if you take a small step at a time.</p><h3>Key Takeaways</h3><p>The digital world doesn’t have to be scary. With these basic protections in place, you can enjoy the benefits of technology while keeping yourself and your information secure:</p><ol><li>Enable two-factor authentication on important accounts, preferably using an authentication app.</li><li>Consider freezing your credit — it’s a free and robust protection.</li><li>Use unique passwords for each account, ideally with a password manager.</li><li>Be skeptical of urgent requests and unexpected contacts.</li><li>Know where to get help if something goes wrong</li></ol><h3>The Power of Community Support</h3><p>Remember that everyone speaking up about being scammed helps others avoid the same fate. Your experience, though painful, can become a powerful tool for protecting others. Many of today’s most effective anti-scam advocates started as fraud victims themselves.</p><h3>Resources for Further Learning and Support</h3><p>Here are some trusted sources for information and assistance:</p><ol><li>Federal Trade Commission’s Consumer Information: ftc.gov/consumer</li><li>National Cyber Security Alliance: staysafeonline.org</li><li>AARP’s Fraud Watch Network: aarp.org/fraudwatchnetwork</li><li>USA.gov’s Identity Theft Resources: usa.gov/identity-theft</li><li>National Center for Victims of Crime: victimsofcrime.org</li><li>Identity Theft Resource Center (free support): idtheftcenter.org</li><li>FBI’s IC3 (Internet Crime Complaint Center): ic3.gov</li></ol><p>Remember: Reaching out for help isn’t a sign of weakness — it’s a sign of wisdom.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4e157e244319" width="1" height="1" alt=""><hr><p><a href="https://medium.com/excella-labs-engineering/protecting-your-digital-life-4e157e244319">Protecting Your Digital Life</a> was originally published in <a href="https://medium.com/excella-labs-engineering">Excella Labs / Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Product Entropy Cycle Framework]]></title>
            <link>https://medium.com/excella-labs-engineering/product-entropy-cycle-framework-3e7a80412860?source=rss----cdbe8ef7f205---4</link>
            <guid isPermaLink="false">https://medium.com/p/3e7a80412860</guid>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Cameron Hopkin]]></dc:creator>
            <pubDate>Mon, 09 Dec 2024 19:34:00 GMT</pubDate>
            <atom:updated>2024-12-10T20:13:27.965Z</atom:updated>
            <content:encoded><![CDATA[<h3>Measuring the Decay of our Digital World</h3><p>Have you noticed how your favorite apps and websites seem to get worse over time? What starts as a clean, useful service gradually becomes cluttered with ads, paywalls, and frustrating changes. This isn’t just bad luck — it’s a predictable pattern I call the Product Entropy Cycle (PEC).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8ergZkMREWm37-5hv1Csuw.jpeg" /></figure><h3>The Journey from Excellence to Entropy</h3><p>Digital products follow a fascinating path of decay, similar to how a pristine garden becomes overgrown without constant care. Cory Doctorow first identified this pattern in 2022, calling it “enshitification.” Building on his insight, we can now understand this decline as a measurable process.</p><p>Think about Facebook’s evolution. Remember when your feed showed mostly updates from friends and family? Now it’s a maze of sponsored posts, recommended content, and advertisements. This transformation didn’t happen overnight — it occurred through gradual changes that prioritized profit over user experience.</p><h3>Understanding the Cycle</h3><p>The Product Entropy Cycle moves through five distinct stages, each taking us further from the original user-friendly product:</p><p>First comes Initial Excellence, when the product truly serves its users. Think early YouTube, with its simple video sharing and minimal ads. Then comes Early Monetization, where reasonable revenue streams appear — like YouTube adding occasional pre-roll ads.</p><p>The real decline starts with the Optimization Pivot, where investor returns become the priority. This leads to Aggressive Extraction, where short-term profits drive decisions. Finally, Terminal Decline sets in, when the product becomes so compromised that users begin leaving en masse.</p><h3>The Mathematics of Digital Decay</h3><p>The Product Entropy Cycle isn’t just an observation — it’s a measurable phenomenon. We can express it mathematically as:</p><p>PE = Σ(MD × FD) / UV</p><p>Where:</p><ul><li>MD (Monetization Density) = Revenue Generation Events / User Interactions</li><li>FD (Feature Degradation) = Inaccessible Features / Total Features</li><li>UV (User Value) = User Engagement × User Satisfaction</li></ul><p>Total Product Entropy (TPE) = PE × EAF (where EAF is the Entropy Acceleration Factor based on investor pressure)</p><p>This formula helps us quantify the five stages:</p><ol><li>Initial Excellence: TPE &lt; 0.2</li><li>Early Monetization: TPE 0.2–0.4</li><li>Optimization Pivot: TPE 0.4–0.6</li><li>Aggressive Extraction: TPE 0.6–0.8</li><li>Terminal Decline: TPE ≥ 0.8</li></ol><h3>Real-World Application: Comparing AI Chatbots</h3><p>Let’s apply this formula to two leading AI platforms:</p><p>ChatGPT Current Analysis:</p><ul><li>MD: 0.6 (High — multiple subscription tiers, usage limits, planned ads)</li><li>FD: 0.5 (Medium-High — GPT-4 paywalled, API restrictions)</li><li>UV: 0.7 (Still Strong but Declining)</li><li>EAF: 1.2 (High investor pressure)</li></ul><p>TPE = (0.6 × 0.5) / 0.7 × 1.2 = 0.51</p><p>This places ChatGPT firmly in Stage 3 (Optimization Pivot), showing rapid progression through the cycle in just over a year.</p><p>Claude Current Analysis:</p><ul><li>MD: 0.3 (Lower — single premium tier, generous free access)</li><li>FD: 0.2 (Low — most features available to free users)</li><li>UV: 0.8 (Strong user value maintained)</li><li>EAF: 0.9 (Moderate investor pressure)</li></ul><p>TPE = (0.3 × 0.2) / 0.8 × 0.9 = 0.068</p><p>This places Claude in late Stage 1, transitioning to Stage 2, suggesting a more measured approach to monetization while preserving user value.</p><h3>Why This Matters for Everyone</h3><p>This isn’t just about inconvenience. As products decay, they often compromise our privacy and security. Free security features get locked behind paywalls. Privacy settings become increasingly complex and harder to find. Our personal data gets shared with more third parties for profit.</p><p>Consider how Google Search has evolved. Once it delivered clean, relevant results. Now the first screen is often filled with ads, and your search data is used for targeted advertising across the internet. This shift affects not just our user experience but our digital safety and privacy.</p><h3>The Resistance Factor</h3><p>Products can resist entropy through what we call the Resistance Factor: R = (Open Source Factor + User Control + Revenue Diversification) / Investor Pressure</p><p>Looking at our AI examples:</p><ul><li>ChatGPT’s R = 0.4 (Lower resistance due to high investor pressure)</li><li>Claude’s R = 0.7 (Higher resistance through balanced approach)</li></ul><h3>Hope for the Future</h3><p>Understanding the Product Entropy Cycle gives us power. We can recognize early warning signs and make informed choices about the services we use. We can support alternatives that resist this decay through open-source development, user-focused business models, and transparent practices.</p><p>As users, we can:</p><ul><li>Recognize when services enter dangerous phases of decay</li><li>Support platforms that prioritize user privacy and security</li><li>Choose open-source alternatives when available</li><li>Advocate for better digital privacy laws</li><li>Vote with our attention and wallets</li></ul><h3>Looking Forward</h3><p>The Product Entropy Cycle isn’t inevitable. By understanding this pattern, we can work together to create and support digital spaces that remain focused on user value rather than pure profit. The future of our digital experience depends on recognizing these patterns and actively choosing better alternatives.</p><p>This mathematical framework helps us not just understand but predict how our digital services might evolve. By watching these numbers change over time, we can make informed decisions about which platforms deserve our trust and attention.</p><p>What changes have you noticed in your favorite apps? Understanding these patterns helps us all make better choices about the digital tools we use every day.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3e7a80412860" width="1" height="1" alt=""><hr><p><a href="https://medium.com/excella-labs-engineering/product-entropy-cycle-framework-3e7a80412860">Product Entropy Cycle Framework</a> was originally published in <a href="https://medium.com/excella-labs-engineering">Excella Labs / Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why You Need 2-Factor Authentication]]></title>
            <link>https://medium.com/excella-labs-engineering/why-you-need-2-factor-authentication-18ce44b89b8f?source=rss----cdbe8ef7f205---4</link>
            <guid isPermaLink="false">https://medium.com/p/18ce44b89b8f</guid>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[J. Austin Hughey]]></dc:creator>
            <pubDate>Fri, 11 Oct 2024 13:06:58 GMT</pubDate>
            <atom:updated>2024-10-11T13:06:58.169Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KwfaKhhVmrtpER5Q" /><figcaption>Photo by <a href="https://unsplash.com/@towfiqu999999?utm_source=medium&amp;utm_medium=referral">Towfiqu barbhuiya</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>One day a few years ago, I checked my spam email box to ensure nothing legitimate was misclassified, as I usually do every now-and-then. As I casually perused the usual laughable messages about cheap pharmaceuticals, last-minute “deals” and the like, one message did grab my attention in the most terrible of ways. Its subject sent a cold chill down my spine:</p><p>It was my password. And it was real.</p><p>Well, it was <em>one</em> of them. And it wasn’t one I had actively used in years, and it was only used for small things I didn’t much care about. Still, the fact that this particular message bore with it the very real, not-a-typo, plain-text password that I had (stupidly) re-used all over the place in my youth absolutely shocked me.</p><p>So naturally I went on to read the message. The contents were your usual extortion scam, but the author relied on the fear generated by one seeing their not-so-confidential-anymore password to make them wonder: “Could they really have/know what they say they do?” Of course, I never responded, but <em>what interested me was the fact that they knew the password itself</em>.</p><p>And that begs the question: Where’d they get it? Which brings me to our first topic:</p><h3>How Passwords are Stored</h3><p>In the early days of password storage, the simplest thing was to just store the letters of your password directly as you typed them in — as “plain text”. So if your password was “Andr0med4” (<a href="https://en.wikipedia.org/wiki/Andromeda_Galaxy">Andromeda</a>), that’s exactly, directly, what might have been stored in the database.</p><p>This was known as “plain text” storage, and while it was the simplest way to go, it has a serious problem.</p><p>What happens if a bad actor happens to steal a copy of the database and disseminate that on the dark web?</p><p>With plain-text passwords, it’s trivial for even a high-school kid to get a copy of that database and spin it up on their laptop and harvest account names and passwords. And since it’s a common practice for people to re-use the same email address and passwords in multiple places, once someone has that combination from one place, chances are pretty good they’ll be able to get into many accounts all over the internet for the given victim in question.</p><p>You might think this sort of thing doesn’t happen anymore, that storing passwords as plain-text is surely an archaic approach, long extinct, not practiced by any reputable company; <a href="https://arstechnica.com/?p=2053118">and you’d be wrong</a>.</p><p>The fact is that when you create an account somewhere, you have no visibility into how they secure your information — or even <em>if</em> they do so at all.</p><h4>A step better: Password Hashing</h4><p>Once it was clear that storing passwords as plain-text was not a good idea, an approach called “hashing” became a practice that was commonly applied.</p><p>With a hashing function, some bunch of data — whether it’s a string of plain text, or even a bunch of binary data, doesn’t matter — is converted via a series of mathematical algorithms into a static string representing that big “chunk” of information you fed the function.</p><p>The beauty of this approach is two-fold:</p><ol><li>It’s one-way only; once you have the hash value, there is no possible way to reverse the hash into its original form!</li><li>If you change any part of the original value and re-run the function, you get an entirely different output value!</li></ol><p>It doesn’t matter how short or long the input to the hashing function is, either. If I hash the letter “a”, I get the same length string. If I hash an entire <em>phone book</em>, I get the same length string — just a very different one. But if I so much as change a single dot in that phone book, on any page, for the alignment between a person’s name and their phone number, then re-hash it? Totally different string output!</p><blockquote>I know that someone out there has literally never seen a phone book before…</blockquote><figure><img alt="Photograph of a page from a phone book. Credit: Library of Congress" src="https://cdn-images-1.medium.com/max/1024/1*skvEgW0NZnTIchJVnhpPCg.jpeg" /><figcaption>(1986) California — White Pages — San Diego North County Coastal — July. California. [Image] Retrieved from the Library of Congress, <a href="https://www.loc.gov/item/usteledirec03354/">https://www.loc.gov/item/usteledirec03354/</a>.</figcaption></figure><p>To bring the theoretical into the real world, let’s see this in practice.</p><p>Using our theoretical password from earlier, let’s turn this into an MD5 hash using Ruby (assuming one has the ‘digest’ extension installed):</p><pre># Assuming &#39;gem install digest&#39; has already been run<br>require &#39;digest&#39;<br>pw = &quot;Andr0med4&quot;<br>puts Digest::MD5.hexdigest(pw)<br><br># Result: &quot;912ec803b2ce49e4a541068d495ab570&quot;</pre><blockquote>Note: We do not recommend the use of <a href="https://en.wikipedia.org/wiki/MD5">MD5</a> as a password hashing algorithm in any possible case whatsoever. It has many known security vulnerabilities and is not nearly computationally expensive enough — which is precisely why it makes a great teaching tool, hence its inclusion here.</blockquote><p>The above would show you an MD5 sum of “912ec803b2ce49e4a541068d495ab570”. This is a one-way representation of what the password above would be when “hashed” by the MD5 algorithm, which is a fairly fast hashing algorithm.</p><p>…and that’s precisely the problem. It’s fast. <em>Too fast</em>.</p><p>When a computer generates a hash, the most efficient path isn’t always the best one in every case. Consider the attacker’s scenario: you have a big database of passwords that are hashed, so to figure out the actual password that you can re-use to break in elsewhere, you have to manually guess ad infinitum, hash it computationally, and compare the output. If the guess is wrong, you have to try another guess.</p><p>If the attacker (who bought that leaked database) can hash a guessed value 100,000 times each minute with a fast algorithm, that means they can guess roughly 100,000 times every minute, or about 6 million potential password guesses (letters of both cases, numbers, special characters, possibly non-English unicode characters, potentially even emoji!) every hour.</p><p>Now, imagine if the raw math behind the scenes were more complex. Imagine that it was complex enough that it slowed the computer down to just 5,000 per minute. That cuts guesses down from 6 million per hour to only 300,000.</p><blockquote>These numbers are entirely fabricated for purposes of simplicity. Real-world scenarios will differ drastically depending on a myriad of factors and context. For most practical purposes, however, “real” numbers are irrelevant when best practices are followed and strong cryptographic algorithms are used.</blockquote><p>So what’s the answer then, just use a different algorithm? Which one?</p><p>While there isn’t one algorithm to rule them all, one that is commonly considered a reasonable choice is called <a href="https://en.wikipedia.org/wiki/Bcrypt">bcrypt</a>. Its <a href="https://github.com/bcrypt-ruby/bcrypt-ruby">Ruby implementation</a> does allow the developer to optionally tune how computationally expensive they want to make the hashing step, but it has another killer feature that, while any developer could write themselves given sufficient knowledge, is a modern necessity.</p><p>But first, let’s describe the problem in further detail.</p><h4>Gettin’ Salty: Why Hashing Isn’t Enough</h4><p>So we know a few things by this point:</p><ol><li>Regardless of what hashing algorithm we use, if you shove any string into the function, you get a static string back;</li><li>If you alter the contents of the input in any possible way in the slightest, you get a radically different output from the hashing function;</li><li>All these hashing functions are one-way only, meaning you can’t get the password from just the hash value alone.</li></ol><p>Put a different way:</p><pre>require &#39;digest&#39;<br>Digest::MD5.hexdigest(&quot;Andr0med4&quot;)<br># &quot;d8e22b19c7785c642e4e3f3b0b3f2252&quot;<br>Digest::MD5.hexdigest(&quot;Andr0med4&quot;)<br># &quot;d8e22b19c7785c642e4e3f3b0b3f2252&quot;<br>Digest::MD5.hexdigest(&quot;Andr0med4&quot;)<br># &quot;d8e22b19c7785c642e4e3f3b0b3f2252&quot;</pre><p>No matter what you put in, if it’s the same thing, you always get the same thing out, it never changes, and you can’t figure out what it was based on the output alone.</p><p>…unless you had a reference table.</p><p>And indeed, such tables do exist. Known as <a href="https://en.wikipedia.org/wiki/Rainbow_table">Rainbow Tables</a>, many organizations and software projects do keep precomputed tabular values of known inputs and outputs for various algorithms so that, given any hashed value, all one needs to do is, much like a dictionary or phone book, just go look it up in the Rainbow Table and you’ve instantly got the input value — in this case, the user’s password!</p><p>So how do modern applications deal with this? Enter the “salt”.</p><p>In modern cryptography, a “salt” is essentially just random bits of data tacked on the end of something before you run it through the one-way hashing function. When you do that, you’re guaranteed to get a radically different value each time because, as we know about one-way hashing algorithms, you get a different output if you change the input!</p><p>Back to algorithms for a second: remember how MD5 was too fast? That’s why we’re going to shift over to a different one known as <a href="https://en.wikipedia.org/wiki/Bcrypt">bcrypt</a>. There are others, however bcrypt is generally regarded as well-tested and secure against most attacks that are likely to happen in the real world. The interesting thing about its <a href="https://github.com/bcrypt-ruby/bcrypt-ruby">Ruby implementation</a> is that it allows you to fine-tune its computational expense, for even more difficulty cracking the password(s) if an attacker did get their hands on the hashes it generated! And to top it all off, it’ll even salt the password for you.</p><p>Check out this example:</p><pre># Assuming &#39;gem install bcrypt&#39; has been done<br>require &#39;bcrypt&#39;<br># You can alter the cost factor if you like, but the default is a fairly good balance<br># BCrypt::Engine.cost = 25 # 12 is the default in the Ruby implementation, only go higher if your hardware is wicked sick<br><br>p = BCrypt::Password.create(&quot;Andr0med4&quot;)<br># Result: &quot;$2a$12$YfkedrXWVyzkhRiSwiNup.a/gjAal2AKcYLgtpfBBv6RxLTtRulJi&quot;<br><br># Now, let&#39;s run the same thing again; see how adding a salt - random data - forces the output to CHANGE?<br>p = BCrypt::Password.create(&quot;Andr0med4&quot;)<br># Result: &quot;$2a$12$es0bveGJGVHphLCvGVw9MeG117fuP7C42ozCKQbZPp4V8sY6mpKni&quot;<br><br># Proving the point one more time:<br>p = BCrypt::Password.create(&quot;Andr0med4&quot;)<br># Result: &quot;$2a$12$LCLeP26kNnAARkbYIgGfx.ntU48.3M.Y6NdbDYVL5o5kn8an7K/.W&quot;<br><br># The BCrypt Gem (Ruby Implementation) gives us nice and easy comparison operators as well:<br>p == &quot;Andromeda&quot;<br># =&gt; false<br><br>p == &quot;Andr0med4&quot;<br># =&gt; true</pre><p>Note how I’ve shown above that, due to the presence of the salt automatically added by the bcrypt implementation in Ruby, the value changes each time that I create the password object above. The “gibberish” part is what a developer would otherwise store in their database — not a plain text password, and certainly not only the hash, but the full output: the hash and the salt, after running both, smushed together, through the bcrypt algorithm (which it does for us automatically thanks to the design of the implementation in Ruby; merely a developer convenience, a productivity enhancement).</p><p>This makes it much more difficult for an attacker to brute force a password even if they did get their hands on the database from whatever application or service might have stored it in this way.</p><h4>How Passwords are Stolen</h4><p>Despite all these measures, via a myriad of ways, various companies still manage to get broken into all the time, having their customer databases copied and disseminated on the <a href="https://en.wikipedia.org/wiki/Dark_web">dark web</a>.</p><p>Typically, the theft of customer (and other) data happens due to lax security measures at the company, failure to follow best practices, security flaws in software, insider threats, and other employee mistakes, including — believe it or not — compromised employee accounts that weren’t using 2-factor authentication!</p><p>Take a look at the “<a href="https://haveibeenpwned.com/PwnedWebsites">Who’s been pwned</a>” list (warning: it’s a <em>long</em> list, loading may take a minute) on <a href="https://haveibeenpwned.com/">Have I Been Pwned</a>, a website that tracks data breaches from various companies and websites worldwide. Some of the worlds largest, most well-capitalized companies are on that list and have been for over a decade; it’s no longer a matter of “if” information is stolen, but “when”.</p><p>From LinkedIn, to Microsoft, AT&amp;T, Equifax, T-Mobile, Facebook, Target, Marriott, eBay, Adobe, Home Depot, JPMorgan Chase, Capital One, Yahoo, and hundreds of others, the trend is clear: data breaches are a fact of life and your passwords will be breached, sold, and exploited at some point. It’s basically a certainty.</p><p>So knowing that, how do you protect yourself <em>anyway</em>?</p><h4>How Passwords are Sold</h4><p>In many cases, these hacks are pulled off by semi-professional criminal organizations. In other cases, these breaches are done by actual nation-state intelligence units that masquerade as organized crime. Either way, their objectives are either (a) to create fear and wreak havoc and chaos, (b) to make money by selling stolen information to those who will create that chaos, or usually (c) both.</p><p>Using various forms of encrypted communications over the dark web, various bad actors will arrange to sell this data. Usually, if the passwords in that data dump are encrypted, it will be up to the purchaser to do whatever they can do decrypt those passwords (if possible, or if of interest, to them). Usually, if the passwords are properly salted and hashed (we’ll get to that below), the passwords themselves are not what the purchaser cares about; usually these types of databases have other information in there of use to criminals and threat actors, things like social security numbers, purchase histories, banking or financial information, and so on. But if they aren’t, or worse — are stored in plain text — it can be an instant and easy way to start exploring what other accounts elsewhere, beyond the website or vendor hacked, the purchasing malicious actor could further compromise for each user in the list.</p><p>These bad actors usually execute their purchases via cryptocurrencies of various forms, through some form of money laundering via known, trusted parties, or through other means. Once the money is received by the seller, they transmit the database via encrypted means to the purchaser, who is then free to do whatever they please with it.</p><p>Generally, they would start by looking at a few dozen, to a few hundred entries and see if cracking the passwords in the file appears a viable endeavor. If they were stored in plain text (also called “clear text”), no cracking is needed — just write a bot to try breaking into each account you can think of using the same email and password and you’re good to go! — but if they are at least hashed, the criminal has to decide whether or not it’s worth trying to crack the passwords in the database.</p><h4>How Passwords are Cracked</h4><p>Now we get to the fun part: how do you crack a password when all you’ve got is a hash?</p><p>There are several tools that have been developed over the years for doing things like this. One such tool is <a href="https://github.com/openwall/john">John the Ripper</a>, which comes with support for using GPUs (Graphical Processing Units — basically video cards PC gamers use for smoother graphics and better framerates) and newer algorithms. (The reason for using GPUs is because they have massively parallel computing infrastructure in the video card, and they can be very useful for large-scale simultaneous computational tasks. This is also part of the reason they’re heavily used in AI datacenters.)</p><p>Tools like John the Ripper provide several functions, however we’ll focus here on the simplicity of what’s known as a “brute force” attack. In a “brute force” attack, a program operates in a loop, coming up with a “guess” of a password, will run it through a hashing algorithm, and compare the result against some form of data — a flat file on disk, or perhaps even the stolen database itself if the tool supports it (or the source code is altered to).</p><p>When all the application did to protect the password was just hash it, this is fairly straightforward: you’ve got only one dimension to guess against:</p><ol><li>Come up with a guess (random text of unknown length) as to the password.</li><li>Hash it through the algorithm you think they used.</li><li>Compare to the known value in the stolen database: Are they equal?<br>No? Go to 1.<br>Yes? Feed it to your account hijack automation pipeline.</li></ol><p>But when you add a salt, you add a huge layer of complexity:</p><ol><li>Come up with a guess as to the password.</li><li>Look up the salt (usually stored alongside the password as plaintext, which is generally fine)</li><li>Tack it on the beginning of the password and feed it to the hashing algorithm.</li><li>Compare to the stored value:<br>If not equal, go back to 3 and instead of putting at the beginning, stick it on the end and see if that works. If not, go back to 1.</li></ol><p>However, even with all this in place, not every vendor, company, or software system will follow these practices, and even those that do may still have weaknesses that are exploitable.</p><p>There’s yet one more way that you — as either a non-developer, or one that does not have the ability to alter some other party’s authentication system — can protect yourself.</p><p>And <em>that</em> is <strong>Two-Factor Authentication</strong>.</p><h3>Two-Factor Authentication (2FA): A One-Time Password for Your Password</h3><blockquote>“I heard you like passwords, so here’s a password for your password.”</blockquote><p>Imagine this: you go to some website, you login, and just after you submit your username and password, but before you see your “dashboard” or whatever, you’re presented a new intermediary step:</p><p>Now you’ll need to provide a second pass<em>code</em> of sorts, consisting of 6 digits.</p><p>You switch over to an app on your phone, select the website you’re trying to login to from a list, and see a countdown timer of 19 seconds: that’s how long you have until the next key rotation. Next to that, you see a code: 862–417. You type it into the form and you’re in.</p><p>That’s 2FA in a nutshell. It’s a rotating code, a “password for your password”, that rotates every 30 seconds on a pre-determined clock, that helps prevent your account from being broken into.</p><p>It relies on the idea of requiring “two factors” to get into your account:</p><ol><li>Something you <em>know</em>: the password; and,</li><li>Something you <em>have</em>: the device that generates the 2FA code.</li></ol><p>In some cases this could be a dedicated, physical device; I’ve worked in environments where RSA SecureID token generators were attached to people’s keyrings, for example.</p><figure><img alt="RSA SecureID Token Generator" src="https://cdn-images-1.medium.com/max/810/1*u2MNgHnY5aa9FdWfJmOguA.jpeg" /><figcaption>Some environments require a separate, physical device that generates the 2FA token used. Most folks would attach this to their keychain so it’s always with them. The depicted image is a commercial product sold by RSA; no endorsement or sponsorship is intended nor implied.</figcaption></figure><p>In the modern age of iThingies and Mobile-everything, some (many) environments have allowed systems like the above to be replaced with various mobile applications like <a href="https://apps.apple.com/us/app/microsoft-authenticator/id983156458">Microsoft Authenticator</a>, <a href="https://authy.com/">Authy</a>, and/or others. Some password managers also have features that allow you to set up 2FA integration with various websites, having that password management app generate the code for you so you can copy-and-paste, or in some cases, it may even auto-fill it for you!</p><figure><img alt="Microsoft Authenticator Screenshot (mobile app)" src="https://cdn-images-1.medium.com/max/1024/1*MkgTG6t7QiF9RZYg1CGtgQ.png" /><figcaption>This screenshot of Microsoft Authenticator (credit: Microsoft) is a bit dated (2020), but should give those new to the concept an idea as to how the overall workflow might function.</figcaption></figure><p>2FA is pretty easy to set up in most cases. In general:</p><ol><li>You set up your account with some website online;</li><li>After that’s done, you follow their steps to add 2FA;</li><li>As part of that, they’ll show you a QR code and/or a strange-looking text string to enter into an app or a security key device like the one described in the example above. You’d enter the code there, and it will generate a “response” of sorts — your first six-digit 2FA code as a sort of “confirmation” that it’s set up right.</li><li>You’ll enter that into the website to confirm you’ve set it up. If it checks out (it will), the website should show you what are called <strong>“recovery codes” </strong>— usually a QR code and/or some series of long weird looking IDs —<strong> that you can copy, print, and/or download and store somewhere. You should absolutely do this each and every time!</strong> <em>I am not kidding around here!</em></li></ol><figure><img alt="Screenshot: Twilio Authy on iOS" src="https://cdn-images-1.medium.com/max/454/1*qXSpOTNO5s7NMm3YuN3IUA@2x.png" /><figcaption>Twilio Authy on iOS (Credit: Twilio Authy, Apple App Store)</figcaption></figure><p>From now on, every time you go to login, you’ll be asked for a username, your password, and if those check out, your 2FA code. If that code does not check out, your login will fail.</p><figure><img alt="2FA Setup process with Slack (Credit: Twilio Authy)" src="https://cdn-images-1.medium.com/max/798/1*LWfELj1QY2F8j10F_L3QAQ.png" /><figcaption>An example of what the setup process might be like, this one from Slack once upon a time. Credit: Twilio Authy (<a href="https://authy.com/guides/slack/">https://authy.com/guides/slack/</a>)</figcaption></figure><p>A word of warning, however: some (too many) websites and companies want to rely on email and SMS as a “cheap” form of 2FA. <strong>DO NOT DO THIS.</strong> <a href="https://www.vice.com/en/article/hacker-got-my-texts-16-dollars-sakari-netnumber/">Text messages — despite carrier claims to the contrary — are not secure at all, and in fact can be hijacked</a>. Email is also not guaranteed to be secure as it traverses the internet and is rarely, if ever, encrypted.</p><figure><img alt="Image depicting 2FA backup codes with codes redacted for safety" src="https://cdn-images-1.medium.com/max/616/1*VX0l-v0HeVaFr4LStTGZ3A.png" /><figcaption>ALWAYS save your backup codes! I am not even remotely kidding here! (Credit: Twilio Authy)</figcaption></figure><p>If for some reason you lose access to your email account, phone number, or device, you may find yourself permanently locked out of your account, and with customer service from most of these companies clearly not a priority, your chances of ever getting it back will likely be “slim to none”.</p><p>For any given website, company or vendor, insist on real 2FA or nothing.</p><h4>How 2FA Stops an Account Compromise</h4><p>By this point you may be wondering, “how does 2FA stop an account from being compromised?”</p><p>Let’s say I’ve successfully stolen your username and password for something and I try to login. But now I’m on step 3: the 2FA code.</p><p>I could guess any random 6 digits, but unless the 2FA implementation at that website, software or vendor is not correctly set up, it would be nearly guaranteed to be wrong.</p><p>Now, let’s say I’ve got massive computing power — perhaps I’m with military intelligence from an enemy nation state. How about I just brute force the right code with the computing resources we have?</p><p>Also not possible: it’ll rotate in 30 seconds, maximum. And you can only try so many generated codes in 30 seconds. On top of that, if the vendor in question has decent security policies, they’re likely to temporarily lock the account after several failed 2FA attempts in a short timeframe, so if an attacker starts flooding bad 2FA codes in a short timeframe — even with a valid username and password — at worst, your account would get locked, and you’ll likely get an email about it (if the company in question has things built well), notifying you to a problem you otherwise might not know about, but would certainly cause you harm.</p><h3>Set Up 2FA On Your Personal Accounts</h3><p>By now it should be clear that using Two-Factor Authentication (provided it’s real and not based on email or SMS) should be an absolute no-brainer default in every case. Chances are that your workplace has account policies pre-configured to require 2FA for sensitive accounts (hopefully for all user accounts), and such decisions are usually outside the locus of control for most employees anyway (if you can turn on 2FA for yourself, do it!).</p><p>Instead, look at your personal accounts outside of work. Start with the things you rely on most critically, things like your bank, cell phone account, and email provider. Do those companies offer authentic 2FA setup? If so, enable it immediately.</p><p>Next, if you use social media authentication — signing in through things like X (formerly Twitter), Facebook, Google and/or Apple — enable 2FA at those places as well, since an account compromise there enables an attacker to access anything you accessed <em>through</em> that account. (Note: Apple’s version of 2FA is generally secure-enough, but is a little…Apple-specific. Nonetheless it’s better than email/SMS, and better than nothing. Use if it you have an Apple account.)</p><p>It may seem daunting to keep track of all these usernames, passwords, 2FA codes and such. That’s where a good password manager comes in. There are many out there, and while Excella makes no endorsements, the author has had positive experiences using <a href="https://1password.com/">1Password</a> in the past, and has heard good things about <a href="https://bitwarden.com/">Bitwarden</a> as well.</p><p>Finally, if your children are of sufficient age to be managing their own internet accounts, and such a topic is appropriate for discussion in your household, consider discussing account security with them as well. Educating them on the importance of using 2FA (cumbersome though it may appear) can make a huge difference: by teaching them about the tools and technology they can use to protect themselves, you equip them with some of the knowledge and skills necessary to succeed in an ever-difficult and increasingly-complex world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=18ce44b89b8f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/excella-labs-engineering/why-you-need-2-factor-authentication-18ce44b89b8f">Why You Need 2-Factor Authentication</a> was originally published in <a href="https://medium.com/excella-labs-engineering">Excella Labs / Technology</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>