Importance of Quality Assurance

8 weeks. That’s all we had. In 2021, just before Thanksgiving, I was brought in to help a major retailer. The year before, they had lost nearly £1 million in only 1.5 hours of downtime. The failure was so severe that the Development Head was fired. This time, there were no non-functional requirements in place. The QA Head and I had to prevent history from repeating itself. The pressure was enormous: SMEs were too busy to help. No performance benchmarks existed. The biggest shopping season was approaching fast. So we started from scratch: 1. Used production data to identify real-world patterns 2. Focused on 4 critical workflows (like the gift card surge) 3. Built load and endurance tests around those flows 4. Partnered with developers to fix bottlenecks quickly 5. Created a live monitoring team to catch issues early The result? 1. Near five-nines availability 2. No major outages 3. And the QA Head got promoted The lesson: You don’t need perfect requirements. You need urgency, focus, and cross-functional action. If you’re heading into a risky season, ask yourself: What’s your most fragile revenue path? Do you know how it behaves under stress? Are you waiting for failure to tell you where to look? 📣 What would you do if you had 8 weeks to stop a million-pound mistake?

Founders often think that QA is just an added cost to be managed. But in reality, QA and software testing are all about revenue. Here’s how: 1 - Preventing downtime Downtime costs $6k+ PER MINUTE for the average software company. If your software goes down, it might cost you revenue directly… or it might be more indirect. Where customers call to complain, you pay for customer service and block customers from using your product. They don’t renew. They don’t upgrade. They don’t give testimonials. 2 - Preventing defects/bugs Defects are costly, too. We’ve heard stories of companies losing hundreds of thousands of dollars in revenue PER DAY due to defects. And other stories of massive customer churn due to big defects. You can’t completely eliminate defects, but you shouldn’t ignore the fact that they are costly, and investing in testing finds them so that you can solve them and improve your overall QA process. 3 - Preventing roadmap delays I have yet to meet a product leader who’s not frustrated (or getting burned) by roadmap delays. These delays are almost *always* connected to last-minute testing and bugfixes. Which can delay new features by days, weeks, or more. This is a revenue problem: The “Cost of Delay” is a metric we should all take more seriously. It answers the question, “How much revenue do we lose by shipping late?” With good, fast testing, product teams can actually hit their roadmap goals and gain insights on how to improve the overall quality. “Cost of Delay” is a great way to get the organization to invest in testing + QA!

🎯 Let's Talk Linguistic Precision in the Age of AI As generative AI becomes embedded in writing programs and literature search databases, I've noticed something concerning: the blurring of critical linguistic distinctions that signal evidence strength. Consider the consequences of an AI outputting "proves" for correlational findings, or "suggests" for experimental results. 🙀 Here's a practical guide to maintain precision in research writing: 🧪 𝐂𝐚𝐮𝐬𝐚𝐥 𝐋𝐚𝐧𝐠𝐮𝐚𝐠𝐞 (𝐒𝐭𝐫𝐨𝐧𝐠 𝐄𝐯𝐢𝐝𝐞𝐧𝐜𝐞) 𝘥𝘦𝘮𝘰𝘯𝘴𝘵𝘳𝘢𝘵𝘦𝘴 𝘦𝘴𝘵𝘢𝘣𝘭𝘪𝘴𝘩𝘦𝘴 𝘳𝘦𝘴𝘶𝘭𝘵𝘴 𝘪𝘯 𝘥𝘪𝘳𝘦𝘤𝘵𝘭𝘺 𝘤𝘢𝘶𝘴𝘦𝘴 𝘭𝘦𝘢𝘥𝘴 𝘵𝘰 🔍 𝐐𝐮𝐚𝐬𝐢-𝐂𝐚𝐮𝐬𝐚𝐥 (𝐒𝐭𝐫𝐨𝐧𝐠 𝐎𝐛𝐬𝐞𝐫𝐯𝐚𝐭𝐢𝐨𝐧𝐚𝐥) 𝘴𝘵𝘳𝘰𝘯𝘨𝘭𝘺 𝘴𝘶𝘨𝘨𝘦𝘴𝘵𝘴 𝘤𝘰𝘯𝘴𝘪𝘴𝘵𝘦𝘯𝘵𝘭𝘺 𝘱𝘳𝘦𝘥𝘪𝘤𝘵𝘴 𝘪𝘴 𝘭𝘪𝘬𝘦𝘭𝘺 𝘵𝘰 𝘤𝘢𝘶𝘴𝘦 𝘵𝘺𝘱𝘪𝘤𝘢𝘭𝘭𝘺 𝘱𝘳𝘦𝘤𝘦𝘥𝘦𝘴 𝘳𝘦𝘨𝘶𝘭𝘢𝘳𝘭𝘺 𝘢𝘤𝘤𝘰𝘮𝘱𝘢𝘯𝘪𝘦𝘴 📊 𝐀𝐬𝐬𝐨𝐜𝐢𝐚𝐭𝐢𝐯𝐞 (𝐒𝐭𝐚𝐭𝐢𝐬𝐭𝐢𝐜𝐚𝐥 𝐂𝐨𝐫𝐫𝐞𝐥𝐚𝐭𝐢𝐨𝐧) 𝘪𝘴 𝘢𝘴𝘴𝘰𝘤𝘪𝘢𝘵𝘦𝘥 𝘸𝘪𝘵𝘩 𝘤𝘰𝘳𝘳𝘦𝘭𝘢𝘵𝘦𝘴 𝘸𝘪𝘵𝘩 𝘤𝘰𝘳𝘳𝘦𝘴𝘱𝘰𝘯𝘥𝘴 𝘵𝘰 𝘤𝘰-𝘰𝘤𝘤𝘶𝘳𝘴 𝘸𝘪𝘵𝘩 𝘵𝘦𝘯𝘥𝘴 𝘵𝘰 𝘷𝘢𝘳𝘺 𝘸𝘪𝘵𝘩 ⚖️ 𝐓𝐞𝐧𝐭𝐚𝐭𝐢𝐯𝐞 (𝐋𝐢𝐦𝐢𝐭𝐞𝐝 𝐄𝐯𝐢𝐝𝐞𝐧𝐜𝐞) 𝘮𝘢𝘺 𝘴𝘶𝘨𝘨𝘦𝘴𝘵 𝘢𝘱𝘱𝘦𝘢𝘳𝘴 𝘵𝘰 𝘱𝘳𝘦𝘭𝘪𝘮𝘪𝘯𝘢𝘳𝘺 𝘦𝘷𝘪𝘥𝘦𝘯𝘤𝘦 𝘪𝘯𝘥𝘪𝘤𝘢𝘵𝘦𝘴 𝘴𝘦𝘦𝘮𝘴 𝘵𝘰 𝘩𝘪𝘯𝘵𝘴 𝘢𝘵 💭 𝐒𝐩𝐞𝐜𝐮𝐥𝐚𝐭𝐢𝐯𝐞 (𝐓𝐡𝐞𝐨𝐫𝐞𝐭𝐢𝐜𝐚𝐥) 𝘮𝘪𝘨𝘩𝘵 𝘵𝘩𝘦𝘰𝘳𝘦𝘵𝘪𝘤𝘢𝘭𝘭𝘺 𝘤𝘰𝘶𝘭𝘥 𝘱𝘰𝘵𝘦𝘯𝘵𝘪𝘢𝘭𝘭𝘺 𝘩𝘺𝘱𝘰𝘵𝘩𝘦𝘵𝘪𝘤𝘢𝘭𝘭𝘺 𝘮𝘢𝘺 𝘤𝘰𝘯𝘤𝘦𝘪𝘷𝘢𝘣𝘭𝘺 𝘤𝘰𝘶𝘭𝘥 𝘱𝘰𝘴𝘴𝘪𝘣𝘭𝘺 𝘮𝘪𝘨𝘩𝘵 Think: Theoretical frameworks, hypotheses 🤔 Why This Matters: When AI tools use "proves" instead of "correlates with," they blur the lines between correlation and causation. When they say "demonstrates" for preliminary findings, they oversell uncertainty. These distinctions aren't just academic - they're fundamental to scientific integrity. #ResearchMethods #AcademicWriting #AI #DataScience #ResearchCommunity #Science Note: Visual made with napkin.ai.

Why Testing is critical even when the issue looks very minor?! Contextual information might be helpful to understand it better. Case 1: A survey platform in your CRM has a laggy frontend. It seems manageable, but slow performance can silently drive users away, leading to potential business loss. Case 2: A simple file-generation software had an alignment issue. The output? Misaligned barcodes on printed cheques, causing automated printing machinery to halt—leading to costly delays. Case 3: Accessing 5-year-old data is slow and unreliable. While it may seem insignificant, consider its impact in critical industries. Imagine an audit where old transaction records fail to load or compliance data is inaccessible, leading to regulatory issues or legal consequences. Case 4: Healthcare domain. Never imagine any cases here. Strictly follow every process in the software cycle. Period. Testing activity should also think about ensuring reliability, business continuity, and preventing real-world consequences. #SoftwareTesting #QualityAssurance #BusinessImpact #TestingMatters

Most of the time easy test conditions are created and checks performed at lower cost, more quickly, more safely, and easier to repeat via tools and automation than doing them yourself. Certain conditions, such as data permutations, state transitions, sequences and flow of business logic are well suited for checking with automated scripts. Sometimes we wait until the most expensive, most complicated scenarios to cover some tests not because have analyzed the problem and decided that is the best way to cover them. Sometimes we wait because we didn't bother to analyze and are hoping that with all that complexity going on in the real world, we will stumble across something. We roll the dice against ourselves when we do that. The real world doesn't decide what to do just because we want it to when we want it to. The better approach is to roll up our sleeves, do the grunt work of thinking through a problem, and come up with test approaches that run faster at lower cost and scale better. Save your experiential and interactive testing for problems that require your capacity to notice things difficult to encode into a script. Save yourself for the deeper, hard problems that demand analysis and contemplation. Use that activity as a complement to earlier analysis to find conditions you didn't anticipate, and then immediately augment the lower-level testing tools with that discovered test scenario. #softwaretesting #softwaredevelopment You can find more ideas like this about testing, test design, development and test strategy in my book Drawn to Testing, available in Kindle and paperback format. https://lnkd.in/gB4NS4BS

I tried every software development approach and finally found a secret weapon that guarantees both quality and speed. I used to think QA as the final checkpoint. But this approach heavily slowed down the process. No founders want slower delivery. They demand both speed and quality. That's why I decide to integrate QA from the start: 1, Joins product discussions - Shapes requirements before coding - Identifies risks upfront - Prevents costly rebuilds 2, Continuous testing cycles - Daily feedback loops - Real-time bug catching - Offering real user perspective 3, Prevention over fixes - Risk assessment pre-sprint - Automated testing from start - Clear acceptance criteria The results surprise me and satisfy all of my clients: - 50% faster development cycles - 90% fewer post-release issues - Zero critical production bugs QA teams shouldn't be firefighters. Treat them as a secret weapon.

Daily releases sound exciting - 𝐮𝐧𝐭𝐢𝐥 𝐛𝐮𝐠𝐬 𝐬𝐭𝐚𝐫𝐭 𝐥𝐞𝐚𝐤𝐢𝐧𝐠 𝐢𝐧𝐭𝐨 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐨𝐧. . Then it’s chaos. Recently, I met a Y Combinator-backed startup founder. Their goal? Release features daily and stay ahead of the competition. The problem? Bugs in production. Here’s what happened: As they scaled up and started shipping daily, they struggled to write and execute test cases. Bugs leaked into production, and the damage was immediate: • Frustrated customers, dropped NPS/CSAT scores. • A reputation hit every time something broke. • Slowing down to fix bugs felt like losing momentum. • Trying to balance it all led to rising costs. They were caught in the classic trade-off: 𝐬𝐩𝐞𝐞𝐝, 𝐪𝐮𝐚𝐥𝐢𝐭𝐲, 𝐨𝐫 𝐜𝐨𝐬𝐭—𝐩𝐢𝐜𝐤 𝐭𝐰𝐨, 𝐫𝐢𝐠𝐡𝐭? But it doesn’t have to be this way. Here’s I shared my perspective in nutshell : 1. Shift left testing: Move QA earlier in the development cycle. Catch bugs during design or coding—it’s cheaper and prevents last-minute surprises. 2. Automate regression tests: A strong test suite ensures fixes don’t break existing features. 3. Use feature toggles: Roll out changes incrementally to catch issues early. 4. Focus on critical issues. Not every bug will hurt your customers. 5. Learn from feedback. Let users guide your priorities without losing focus. When you integrate testing early and automate regressions, you’re not just fixing bugs—you’re preventing them. Less firefighting, more time to build features that matter. How do you handle this balance in your projects? #ProductDevelopment #ShiftLeft #QA #Leadership

For most of my career, I have been a champion of a model-based approach to software quality. It involves creating visual models of a software requirement. An often overlooked benefit is the simple act of forcing critical thinking, and collaborating on it. It sounds obvious, but specifically thinking about the positive and negative scenarios within a system. It never ceases to amaze me how significant the improvement in quality can be, simply by thoroughly thinking through a process. Questions like, "What happens when I enter a negative number?" or "What if I attempt to transfer more money than I have in my account?" prompt us to dive deeper. Modelling forces thinking about the different scenarios which ultimately leads to a greater understanding of how the system, but also gives visibility into the thoroughness of the testing approach. #SoftwareQuality #ModelBasedApproach #QualityEngineering #SoftwareTesting"

🚨New blog post📝: Your Study Is Too Small (If You Care About Practically Significant Effects) Seeing more discussions about MCIDs and effect size confidence intervals on here lately, so I wrote up a quick explainer on why this matters more than most people realise. I focus mostly on why your goal - effect estimation - can be in conflict with your approach - power analysis for hypothesis testing. Most studies are designed to detect effects, but not to precisely estimate them. This creates a blind spot that affects how we interpret research across medicine, psychology, business, and policy. Here's a concrete example: You run a study testing if a new training program improves performance. You get Cohen's d = 0.40 with p < 0.05. Success, right? Not so fast. Your confidence interval might be [0.12, 0.68]. What this actually tells us: ✅ "There's probably an effect" ❌ "But it could be anywhere from trivially small to quite large" If your research question is "Should we implement this training program population-wide?" you need to know whether the effect is meaningfully large, not just whether it exists. The solution is precision-focused sample sizes: Instead of asking "How many participants do I need to detect an effect?" ask "How many do I need to estimate it precisely?" For the same study: Standard approach: n = 100 per group → CI spans 0.56 units (total N = 200) Precision approach: n = 196 per group → CI spans 0.40 units (total N = 392!!!) Bigger samples aren't just about finding effects - they're about understanding them well enough to make informed decisions. Whether you're evaluating medical treatments, educational interventions, or business strategies, precision matters as much as statistical significance. See full breakdown at the link below. (Also included is a lovely illustration of the issue by Adrian Olszewski; go give him a follow also!) #Research #DataScience #Statistics #MCID #EffectSize #Evidence #DecisionMaking #PowerAnalysis #Equivalence #NonInferiority #EstimationStatistics #Precision #Psychology #SampleSize #NHST https://lnkd.in/eRAWHSCV