Business Insights and Analysis

Amazon kept getting complaints – but the executive team didn’t know why, so Jeff Bezos called customer services on speaker in front of everyone… In the late 1990’s, Amazon was growing fast. Every week, Jeff Bezos gathered his leadership team for their most important ritual: the Weekly Business Review. One day, the head of customer service proudly presented a slide: “Average phone wait times: 59 seconds.” Tick - move on to the next item. Jeff paused. A number of Amazon’s customers were not happy. He knew this because he maintained and read a public email account. Customers would email him directly, Jeff would forward on to the appropriate executive with a simple “?” for follow up. But his customer service leader was saying everything was rosy. Something didn’t add up… So right there, in the middle of the meeting, Jeff did something radical: Placing the call on speaker, with the entire executive team watching, he picked up the phone and called Amazon’s customer support line… The room went silent. 60 seconds passed. Then 2 minutes. Then 5 minutes. Still no answer. After 10 minutes of hold music - still nothing. “It was a really long time,” Jeff recalled. “More than 10 minutes.” In a flash, the metric they’d been using to reassure investors and guide operations collapsed. The problem? The data wasn’t wrong – but it was measuring the wrong thing. The metric measured average wait time for answered calls, ignoring the calls that never got picked up. That one moment rewired Amazon’s entire approach to measurement, feedback, and truth. Jeff didn’t just want favorable data, he wanted reality. The result? Amazon rebuilt its customer service from the ground up - and made customer service a core part of its moat. Reflecting on that meeting, Jeff said: “When the data and the anecdotes disagree, the anecdotes are usually right.” 👉 Enjoyed this story? Subscribe for one great real life finance story a week: BizStory.co

Africa has too many small businesses, and too little business."  A few months ago, The Economist said this and it stuck with me.   #Africa is the only continent with no company on the Forbes Global 2000. We have just 60% of the large firms you’d expect, given the size of our economies. Why? Are we less talented? Less ambitious? No. But we do face structural roadblocks. Here’s what we hear all the time:  Lack of access to finance. Lack of access to markets. #Finance Want to start a business? Collateral + 20% interest rates. Manage to grow? You’re lucky if you get paid within 90 days. And yet, Kenyan banks are thriving.  NCBA Group just reported profits of KSh 61.8B (~$460M). So we have money — just not for businesses that create jobs and value? #Markets: -Flights within Africa cost $400–$1000. Cheaper to fly to Dubai or Europe. - It’s easier to ship goods from China to Kenya or Uganda than between our own countries. -54 countries = 54 licenses. One continent, but no real single market. -Even our trade payments go through the US dollar, costing us $5B every year. And yet, we have the tools: AfCFTA – continental free trade SAATM – single African air transport PAPSS – Pan-African payment system But we haven’t activated them at scale. So what now? Global systems are skewed. Capital is more expensive. Risk is exaggerated. But we’re not powerless. Individually, we are weak. But collectively, Africa is strong. Let’s stop waiting for  international governments or donors to save us.  Can we build trust, work as pan-Africa, drop the ego, and prove The Economist wrong again?

Another unfortunate example of outsiders thinking you can fully automate accounting. Bench, and those similar, have an unscalable model. Why? To have a large enough TAM and attract VC investments, they must charge rock bottom prices. Otherwise they are unaffordable to many small businesses and TAM shrinks. To make money charging rock bottom prices you have to automate your way to profitability. But outsiders seemingly don’t understand it’s nearly impossible to automate accounting to the degree they need to (at least as of now). Accounting is heavily reliant on the inputs from the client. Garbage in = garbage out. Automations don’t fix this. So to ensure quality, you hire accountants, likely offshore. But labor costs significantly compress margins. A typical Bench engagement might yield $250/mo in revenue but cost, in direct labor, $80 to prepare and $40 to quality check. You then need someone to own the account and deliver the work… if you include that in your $40 quality check you’ll see significantly less quality output. So roughly 50% gross margin… but if you are Bench you also have a ton of embedded costs that most accounting firms don’t have (sales, marketing, technology dev, account managers, legal (quality issues), refunds (quality/service issues), internal finance, etc.) You just can’t make money in this model… at least not until AI can truly automate most of the accounting work. Interestingly I think a well staffed and solidly built CPA firm could figure out how to move down market and crush it. Why? Because unlike tech companies, the CPA firm’s goal is to be profitable from day 1. If a CPA firm goes down market, they’d prioritize profitability over top line growth. But that’s a boring VC story.

This is the most underrated way to use Claude: (and it has nothing to do with writing or coding) It's competitive intelligence. Using data that's free, public, and updated every single week. Here's my extract step by step guide: Step 1. Go to claude .ai. Step 2. Select the new Claude "Opus 4.6." Step 3. Turn on "Extended Thinking." Step 4. Pick a competitor. Go to their careers page. Step 5. Copy every open job listing into one doc. (Title. Team name. Location. Full description) Step 6. Save it as one .txt or .docx file. Step 7. Search the company at EDGAR (sec .gov) Step 8. Download its recent 10-K or 10-Q filing. (Official strategy, risks, and financials - all public.) Step 9. Upload both files to Claude Opus 4.6. Step 10. Paste this exact prompt: "You are a competitive intelligence analyst at a rival company. I've uploaded [Company]'s complete current job listings and their most recent SEC filing. Perform a strategic intelligence analysis: → Cluster these roles by what they suggest is being built. Don't use the team names they've listed. Infer the actual product initiatives from the skills, tools, and responsibilities described. → Identify capabilities or teams that appear entirely new — not mentioned anywhere in the SEC filing. These are unreleased bets. → Find roles where seniority is disproportionately high for a new team. This signals executive-level priority. → Cross-reference the SEC filing's Risk Factors and Strategy sections with hiring patterns. Where are they investing against a stated risk? Where did they flag a risk but have zero hiring to address it? → Predict 3 product launches or strategic moves this company will make in the next 6-12 months. State your confidence level and cite specific job titles and filing sections as evidence. Format this as a 1-page competitive intelligence briefing for a CMO." What you'll find: → Products that don't exist yet but will in 6 months. → Priorities that contradict what the CEO said. → Risks they told the SEC but aren't addressing. This is what consulting firms charge $200K for. It took me 10 minutes. I used the new Claude 'Opus 4.6' for a reason: ✦ It read 60 job listing & a 200-page filing together.  ✦ And connects dots across both. ✦ It is superior in thinking and context retrieval. That's why I didn't use ChatGPT for this.

Nobody cares in FP&A how accurate your forecasts are if you're late. Example 1 - Monthly revenue forecasting Avoid this mindset: "We need 100% accurate data before submitting the forecast." Instead, think: "Our forecast, based on 90% of the data, shows we're 5% above target. This early insight allows sales to adjust tactics now, potentially increasing results by 2-3%. We'll refine later, but this snapshot drives immediate action." Example 2 - Annual Budgeting Process Move past: "We'll start budgeting in Q4 for the most accurate year-end data." Be proactive: "By using a rolling forecast, we adapt quarterly. Our Q2 review flagged a potential 15% increase in material costs. Early action let operations identify alternate suppliers, saving $2M next year. Waiting for year-end would've been too late." Takeaway? Speed in FP&A matters more than accuracy. Balancing speed with precision makes FP&A a true strategic partner. Time > everything

Mike Maples, Jr is one of the most successful startup investors in history. He's worked with more early-stage startups than almost anyone alive, and with his fund, Floodgate, helped pioneer seed-stage investing as a category. He's been on the Forbes Midas List eight times and has made early bets on transformative companies like Twitter, Lyft, Twitch, and Okta. In his new book (coming out this Tuesday!), Pattern Breakers: Why Some Start-Ups Change the Future, he shares the three common elements he's uncovered that separate startups (and founders) that break through and change the world from those that don’t. This research is rooted in his decades of notes, decks, and founder relationships, and is unlike anything I've seen elsewhere. In our conversation, Mike shares: 🔸 The three elements of breakthrough startup ideas 🔸 The importance of founder disagreeableness 🔸 Why you need to both *think* and *act* differently 🔸 How to avoid the “comparison trap” and “conformity trap” 🔸 How to apply pattern-breaking principles within large companies 🔸 Mike’s one piece of advice for founders 🔸 Much more Listen now 👇 - YouTube: https://lnkd.in/gPjw8RXk - Spotify: https://lnkd.in/gAUbgGxz - Apple: https://lnkd.in/g6t367uY Some key takeaways: 1. Get out of the present: Instead of just thinking about solving current problems, you need to immerse yourself in the future. Look for emerging trends, technologies, or shifts in behavior that suggest where the world is heading. Great innovations often stem from individuals who immerse themselves deeply in a niche or cutting-edge area. 2. Great startup ideas share three elements: a. Inflections: External shifts that create potential for radical change in how people think, feel, and behave. b. Insights: A unique understanding of how to harness these inflections and enable a future that the company believes in. c. Founder-future fit: An alignment between the founders and the future they envision, including their skills, motivations, and network. 3. Three things that successful founders do differently: a. Movements: A movement aligns early believers around a higher purpose, leveraging their emotional commitment rather than just pragmatic benefits. b. Storytelling: Frame your startup’s story as a hero’s journey. Position yourself not as the hero but as the guide (like Obi-Wan Kenobi) who invites customers (the heroes) to embark on a transformative journey toward a better future. Tailor your narrative to resonate with different stakeholders—investors, customers, employees—by emphasizing how they can achieve their aspirations through your vision. c. Disagreeableness: Founders who challenge the status quo often appear disagreeable because they defy conventional norms. They drive change by questioning existing patterns and persuading others to embrace new ways of thinking and acting.

Gartner MQ 2025: Enterprise Networking Disrupted, and Cisco’s No Longer Leading The new Magic Quadrant for Wired & Wireless LAN is here, and it marks a true turning point. Having spent nearly three decades in this industry, I can say without hesitation: this is the most significant shift in enterprise networking we’ve seen in years. Cisco is officially out of the Leaders quadrant after dominating for over a decade. Their fragmented strategy, rising complexity, and disjointed cloud/on-prem platforms are no longer being overlooked. Extreme Networks also drops, now behind Arista, which continues to climb with a solid AI and fabric-driven approach. Nile and Meter make powerful debuts as Visionaries, proving that Network as a Service (NaaS) is fundamentally changing the market dynamic. Juniper Networks leads the pack, delivering exceptional execution and vision powered by Mist AI, automation, and cloud-native operations. HPE Aruba remains a Leader, and what’s even more interesting is the pending acquisition of Juniper by HPE, expected to be reviewed in court later this year, could create an absolute #powerhouse in networking. Imagine Aruba’s robust edge and campus portfolio combined with Juniper’s AI and cloud-native strength. Note: The full Gartner report is behind a paywall https://lnkd.in/dYCYnzsP #Networking #EnterpriseIT #GartnerMQ #NaaS #AIops #Cisco #Aruba #Juniper #Arista #DigitalTransformation

How to Practice Data Analysis: A Step-by-Step Guide 1. Understand the Basics Before diving into practice, ensure you have a strong foundation in key concepts: What is data analysis? It is the process of inspecting, cleansing, transforming, and modeling data to discover useful information, draw conclusions, and support decision-making. Types of data analysis: Descriptive Analysis: Summarizes historical data. Diagnostic Analysis: Explains why something happened. Predictive Analysis: Uses data to predict future outcomes. Prescriptive Analysis: Suggests possible courses of action. Key Skills to Learn: Statistics: Fundamental concepts like mean, median, variance, correlation, and regression. Data cleaning: Handling missing values, outliers, and ensuring data consistency. Visualization: Tools like matplotlib, seaborn, and Tableau for data presentation. Programming: Proficiency in Python or R for data manipulation and analysis. 2. Choose a Programming Language Data analysis is often performed using programming languages, especially Python and R. Choose one based on your interest or job requirements: Python: Preferred for beginners due to its readability and a large community. Libraries such as pandas, NumPy, scikit-learn, and matplotlib make data analysis in Python smooth. R: Specifically designed for statistical computing and data visualization. It’s widely used in academia and among statisticians. 3. Set Up the Right Tools You'll need the right environment to practice data analysis. Some common tools include: Jupyter Notebook: Interactive environment for Python that’s great for experimenting and visualizing code outputs. RStudio: An IDE for R, providing a great interface for data analysis and visualization. Google Colab: Cloud-based Jupyter notebooks that allow you to run Python code without needing to install anything on your local machine. 4. Work with Real Datasets Practicing with real datasets is one of the most effective ways to learn data analysis. There are numerous free sources where you can download datasets: Kaggle: A platform where you can find datasets, participate in competitions, and learn from other analysts. UCI Machine Learning Repository: A collection of databases for the empirical analysis of machine learning algorithms. Google Dataset Search: Helps find datasets stored across the web. Datasets to Start With: Iris Dataset: A small, classic dataset for classification. Titanic Dataset: Often used for practicing predictive modeling. NYC Taxi Trip Data: For time series and exploratory analysis. 5. Develop a Problem-Solving Approach When you have a dataset in hand, follow a structured approach to analyze it: 1. Define Your Objective What questions are you trying to answer? Clearly define the problem you're attempting to solve, as this will shape how you approach the data. 2. Data Collection If you are not given a dataset, you can scrape or collect your own data using APIs or web scraping tools like BeautifulSoup or Scrapy.

Reporting is NOT delivering insights. Unfortunately, many data & analytics professionals think it is. Reporting dashboards show WHAT's happening and enable basic slicing and dicing, but fail to deliver WHY. Example - "Performance is down 15% WoW" This is just stating the obvious. It's not a real insight. It's not actionable. This leaves many business leaders frustrated. When business stakeholders ask for more dashboards, what they are ultimately trying to achieve is "I need to know what's impacting my key business metrics and what I should do to improve it". Adding 15 more charts/views/slices won't help much to understand what's impacting the key business metrics and which actions should be taken. The key to REAL INSIGHTS that can move the needle? ROOT-CAUSE ANALYSIS to find the WHY (i.e., DIAGNOSTIC analytics) This is the most effective way to drive change with data & analytics. This can make the data & analytics team a TRUSTED ADVISOR and get a seat at the leadership and decision-making table. Insights need to be: 🟢SPEEDY: business stakeholders need quick insights into performance changes to make decisions before it's too late 🟢PROACTIVE: don't wait for business stakeholders to ask. Monitor key metrics and proactively share insights to become that trusted advisor 🟢IMPACT-ORIENTED: focus on the key drivers that drove most of the change and communicate accordingly 🟢EFFECTIVELY COMMUNICATED to drive the right action #data #analytics #impact #diagnosticanalytics

Big Tech is still throwing punches - even with one hand tied behind its back. 👊 If you’re wondering why markets are up today, here’s the honest answer: no one ever truly knows. But here’s the next-best answer: Microsoft and Meta blew the roof off earnings last night. META ($1.5T) is up 4.5%, MSFT ($3.2T) is up 8.5% -and when companies of that size move, the whole market moves with them.  So what happened? Let’s start with Microsoft, the $3.2 trillion ✨Cloud-and-Copilot✨ empire: 📈 $70.1B in revenue (+13% YoY) and $25.8B in profit (+18%). Azure up 33%. 🏗️ After 10 straight quarters of pedal-to-the-metal AI capex, Microsoft tapped the brakes, just slightly. Capex declined QoQ from $22.6B to $21.4B. It’s not a slowdown; it’s a strategic shift. From land grab to landscaping. From “build it all” to “build where usage is real.” 🤖 Copilot usage is scaling (55% growth in enterprise customers QoQ, DAUs +2x). Azure AI workloads are contributing 16 points of its 33% growth. Infra spend is translating to customer pull-through. 🐶 Satya Nadella says 30% of Microsoft’s own code is now AI-generated. Kevin Scott predicts that number hits 95% within five years. Microsoft is dogfooding AI harder than anyone, and in doing so, reshaping how software is built and sold. ➡️ From silicon (Maia), to models (OpenAI), to infra (Asure AI), to end-user interfaces (Copilot), Microsoft is executing an Apple-style vertical strategy, only scaled for global enterprise. Now over to Meta, the $1.5 trillion ✨Ad-and-Algorithm✨ machine: 📈 Revenue: $42.3B (+16% YoY). Net income: $16.6B (+35%). $50B stock buyback announced (yes, billion with a B) 🏗️ Capex guide raised to $64–72B for the year. A big chunk is going to custom silicon and data centers optimized for AI. Meta has been relatively quiet about chips, but they’re building a full vertical AI pipeline as well. Think Microsoft, but for consumers. 📦 While everyone else is chasing monetization, Meta is chasing usage. They’re not trying to sell AI tools. They’re trying to make sure you use Meta platforms 10% more per day. So yes: Meta is spending aggressively. But they’re doing it with a consumer flywheel and ad engine that already converts attention into dollars. They don’t need to prove that AI has economic value. They just need to prove that AI can deepen engagement. The rest takes care of itself. ➡️ We’re used to thinking of Meta as an ads business wrapped in a social network. But increasingly, it’s an LLM company with a multi-billion-user training pipeline. Even with constraints - AI infra bottlenecks, regulatory landmines, macro uncertainty - Big Tech is playing from a structurally advantaged position. They have data, distribution, capital, user base, and the luxury of not needing to monetize AI today to justify the spend. These aren’t your grandma’s incumbents. They move like startups, but with trillion-dollar tailwinds. 🌪️