Robert Doxey

Deinfluencing you as an ex-Senior Data Analyst in Tech: 1. I never want to work at FAANG and think it's highly overrated.     2. I don't code or do data projects for fun on the weekends.     3. I love putting tables on dashboards so stakeholders can export data to Excel.     4. People can learn data skills on their own & be super successful without a formal education.     5. Overly data-driven cultures can be toxic and lead to analysis paralysis.     6. Instead of obsessing over new trends like LLMs and Gen AI, I've specialized in the bread and butter of analytics (SQL & data viz!)     7. Data storytelling can be manipulative and deceiving.     8. Data isn't always as secure and protected as you think it is.

1,005

62 Comments

Deinfluencing you as an ex-Senior Data Analyst in Tech: 1. I never want to work at FAANG and think it's highly overrated.  2. I don't code or do data projects for fun on the weekends.  3. I love putting tables on dashboards so stakeholders can export data to Excel. 4. People can learn data skills on their own & be super successful without a formal education. 5. Overly data-driven cultures can be toxic and lead to analysis paralysis. 6. Instead of obsessing over new trends like LLMs and Gen AI, I've specialized in the bread and butter of analytics.  7. Data storytelling can be manipulative and deceiving.  8. Data isn't always as secure and protected as you think it is.

764

54 Comments

In 11 months, I negotiated $71K to $153K as a data analyst in tech ⬇️ By NOT saying any numbers in interviews and underselling myself. I didn’t think I was worth $140K or $153K and I probably would’ve asked for less money (and been happy with it too!). But instead of naming my price too low, I asked the recruiter their budget. 9 times out of 10 they’ll tell you, and if they don’t… kind of a red flag. I don’t like to go forward without that transparency. Here’s how you respond to the dreaded “what are your salary expectations?” question: I’d love to learn more about the role and meet more members of the team so I can better assess the market value. But if you have a budget, it would be helpful to hear. Then SIT IN THE SILENCE. Do not over explain or justify. Do not fill the awkwardness. Just sit and make them talk. BOOM 💥 Whoever speaks first loses 👏🏼

2,532

197 Comments

Secret analyst habits: - Will rewrite an entire SQL process just because they didn't like the way it was formatted. - Say "it depends" more often than a therapist with a philosophy degree. - Believe "clean data" lives in the same place as unicorns, fairies and leprechauns. - Still genuinely thinks they'll get around to "doing the documentation later."

1,573

108 Comments

Databricks just hit $5.4B run rate at a 65% YoY growth. That doesn't happen. I read a few articles, but the one written by Jason M. Lemkin breaks it down the best. Here were some of the stats he calls out: 1. Databricks growth is ACCELERATING at a $5.4B run rate. Usually as companies get larger, growth slows -- Databricks is doing the opposite which is nearly unheard of. 2. Databricks growth is absurd at 65% YoY (he uses Snowflake for comparison at ~25-30%) 3. There are zero other public enterprise companies other than Palantir that is even over 40% growth. 4. Just the Databricks AI business alone would be a top 20 Public software company ($1.4B run rate) And he nails why this happened. Databricks built a data platform specifically for data engineering and data science workloads. Then once we became the de-facto standard for that, we kept adding products. Data Warehousing (DBSQL) -- $0 - $1B in under 4 years AI Products (Agent Bricks,etc.) -- $0 - $1.4B in 18 months Lakebase -- 2x the revenue trajectory of Data Warehousing at this stage AI / BI Dashboards and Genie -- insane adoption (free product!) and because our founding team (who are all still at the company) are PhDs (including Ali, our CEO).. we simply deliver incredible product time and time again. It is a recipe for continued upward trajectory...and from the inside, it doesn't feel like it's slowing down any time soon. https://lnkd.in/eWxww5gf

409

12 Comments

I've watched companies spend $5M on dashboards that collect dust. I've also seen a $0 spreadsheet drive a $50M decision. After two decades of building data products, I've stopped asking "what should we build?" and started asking "will anyone act on this?" Most dashboards fail the same six tests. 1. Will anyone look at it? I've seen million-dollar BI implementations where usage drops to zero after a few months. The exec sponsor moved on. The team got reorganized. Nobody opens it. Before building anything: Who specifically will view this, and what triggers them to look? 2. Will they understand it? I've sat in rooms where senior leaders nod along at charts they can't interpret. They won't admit it. They'll just ignore it later. The test: Can someone grasp the insight in under 10 seconds without you explaining it? 3. Will they act on it? This is where most dashboards fail. I've reviewed hundreds of dashboards. Ninety percent answer questions nobody is acting on. The data is accurate. The visuals are clean. It doesn't matter because no decision is attached. What specific action should this trigger, and does the viewer have authority to take it? 4. Can you measure the impact? If you can't prove it drove results, you can't prove it was worth building. Every dashboard should have a metric that improves if people use it. No measurement, no proof of value, no budget next year. 5. Does it replace something? Most dashboards get added to the pile. Nobody asks what they eliminate. I've seen companies with 400+ dashboards where maybe 20 get used. The rest are noise. What existing report or process does this replace? If nothing, you're adding clutter. 6. Who owns it? Dashboards without owners become orphans. Six months from now, who's accountable for keeping this accurate and relevant? If the answer is "the data team," it's already dying. The pattern I see over and over: Leadership requests a dashboard. Team builds it. Someone looks at it once. Leadership asks why data isn't delivering value. The team delivered. The organization failed the other five. What's a dashboard you've seen that actually changed how a company operated? I'm advising companies on AI strategy and exploring my next executive role (CDO, CPO, VP of Product/AI). If you're building something interesting, or hiring, DMs are open. #DataStrategy #Analytics #BusinessIntelligence #DataLeadership #Dashboard

84

25 Comments

How I Networked with Amazon Hiring Managers and received 10+ informational calls in 60 days before getting hired as Data Engineer... I used to send 100+ LinkedIn connection requests to get maybe 1 response. The messages that got accepted? Radio silence. It felt like shouting into the void. Until an accidental encounter on a flight that changed everything. Early in my job search, I found myself sitting next to a Senior Manager from a FAANG company. After some small talk, I shared my networking struggles. "Show me your messages," he said. After reading them, he smiled. "No wonder you're not getting responses." His feedback hit hard but changed my entire approach: "Stop talking about what you built. Start talking about problems you solved." 𝘏𝘦 𝘨𝘢𝘷𝘦 𝘮𝘦 𝘵𝘩𝘳𝘦𝘦 𝘱𝘳𝘪𝘯𝘤𝘪𝘱𝘭𝘦𝘴 𝘵𝘩𝘢𝘵 𝘵𝘳𝘢𝘯𝘴𝘧𝘰𝘳𝘮𝘦𝘥 𝘮𝘺 1-𝘪𝘯-100 𝘳𝘦𝘴𝘱𝘰𝘯𝘴𝘦 𝘳𝘢𝘵𝘦 𝘵𝘰 1-𝘪𝘯-5: 1. 𝙇𝙚𝙖𝙙 𝙬𝙞𝙩𝙝 𝙫𝙖𝙡𝙪𝙚, 𝙣𝙤𝙩 𝙩𝙤𝙤𝙡𝙨 ❌ "I have built ETL pipelines using Spark" ✅ "I solved SLA problems of generating regulatory reports by migrating them to Spark" 2. 𝙁𝙤𝙘𝙪𝙨 𝙤𝙣 𝙗𝙪𝙨𝙞𝙣𝙚𝙨𝙨 𝙞𝙢𝙥𝙖𝙘𝙩 Instead of listing technologies, I highlighted outcomes. Not "reduced time by 40%" but "ensured compliance deadlines were met when manual processes were failing." 3. 𝙏𝙖𝙧𝙜𝙚𝙩 𝙞𝙣𝙩𝙚𝙣𝙩𝙞𝙤𝙣𝙖𝙡𝙡𝙮 Only connect with people actively hiring or building teams. Check for recent job postings on their profile. The results in the next 60 days: - 10+ calls with Amazon hiring managers - 20+ more from other tech companies   - Multiple interview opportunities - And yes, landed my current Data Engineer role The biggest mindset shift: Hiring managers at FAANG don't care about your tool stack. They care about problems you can solve for their team. Every message I sent after that flight focused on one thing: How my experience solving specific problems could help their team's challenges. --- Comment "Template" and I'll share the exact message structure that got me those 10+ Amazon calls. #DataEngineering #NetworkingStrategy #JobSearch #Amazon #CareerAdvice #BigTech

612

899 Comments

Today we're announcing Chalk's $50M Series A at a $500M valuation, led by Felicis When I first started working on data infrastructure as an intern, the developer experience was an absolute nightmare. The slow dev cycles and lack of feedback turned me off the entire concept, and I switched my undergrad specialization from AI to Systems. That’s more-or-less how the field has remained for the past decade. Anyone who works in AI knows that when you need real-time inference, you get stuck with horrible tradeoffs between execution speed, breadth of functionality, and developer experience. But, I don’t think this is an immutable fact of life. At Chalk, we’ve dedicated ourselves to spending the next 10 years building a system that delivers all three dimensions at once. We still have so much to build, but we’ve come a long way – customers like Whatnot, Socure, and Sunrun use Chalk to easily ship complex models that used to require weeks or months of painful infra work. Turns out when you give smart people the right tools, they build incredible things. This funding gives us the resources to keep doing crazy things to help our customers succeed. We're growing the team fast across engineering, go-to-market, and ops in SF and NY. Come join us :-) Thanks Anna from Reuters for the great coverage! Article link in comments. 

266

20 Comments

We are obsessed with contact data quality. It'll never be as good as we want, but that's OK. First off, let go of 100% accuracy. Never happening. Impossible. I've talked to several LARGE data company founders. They aren't hoping you won't notice some data is bad. They are working hard to improve it. Data accuracy is NOT what is holding you back. Sure...there is a bar of acceptability that needs to be met. You need to be around 70% or you waste too much time. There's a bigger problem than data accuracy. Your targeting and messaging strategies are outdated. If your ICP is based on firmographic and technographic traits, you are stuck in the 2010s. Just because someone looks like your customer doesn't mean they have the same problems. If you base your messaging on the jobs to be done and general pains of a persona, you are stuck in the 2010s. People don't have "general" jobs and pains. They have specific ones. The ones that will either get them promoted or hired for fixing or not fixing them. ICP should be based on the pains you can prove you can solve for a company. Messaging should be for people whose job it is to stop the company from measruably suffering from that observable pain. If I wash windows, I go to the houses with dirty windows and ask who cleans the windows. If the person that answers the door doesn't, I ask for the person who does. That's how it's supposed to work in sales. You don't say "2 story brick houses usually have dirty windows and the teenage son in the house is most often tasked with cleaning them on Saturdays. Therefore, I will send 100 reps to hunt down teenage window washers in 2 story brick houses. Ignore all other houses." Yet, we do that when we say "ABC, Inc has Salesforce, Outlook, and an SDR team, so they probably have a pipeline problem and the SDR manager's job is to solve that." While that may partially be true, being partially true is why your reply rate is 1%. If your data isn't at least 65% accurate, your reps are kicking rocks. One thing I've found that works great for data is to have a primary provider that a rep uses, but then to have a second provider (like Findymail) who can cleanse and enrich that data. Having RevOps create a report of discrepancies allows you to see where accuracy is and is not so you can get rid of providers who can't meet the bar of acceptability.

47

21 Comments

Alvys just raised a $40M Series B, and for anyone who has wrestled with a legacy TMS, this feels like a long-overdue jailbreak. This isn't cosmetic software with a few fresh dashboards. This is Nick Darman and Leo Gorodinski turning battle scars into a platform that finally respects the operators running America's supply chains. Darman has lived both sides of the grind, dispatching trucks in 2004 while at Georgia State University, scaling Archerhub to $90M in recurring revenue, and slogging through every outdated system the industry still clings to. Gorodinski built his name at Jet.com, where he was the first executive hire, ran engineering, and helped grow it into Walmart's $3.3B acquisition. Together, they're rewriting what it means for a #trucking platform to be the core operating system of the business. The round was led by RTP Global, with Alpha Square Group, Titanium Ventures, Picus Capital, and Bonfire Ventures joining in. Investors don't back a $40M raise unless the numbers justify it. Tenfold growth in 2023. A customer base doubling since January of last year. Carriers reporting 28 percent more #loads, 9% revenue growth, 90% faster accounting, and 80% less #manualdata entry. Case studies speak louder than pitch decks: Bear Down Logistics Inc. cut twelve minutes per load, PBJ EXPRESS INC slashed $120K in overhead, and reconciliation times fell off a cliff. Those gains compound fast when you're moving freight at scale. Today Alvys powers about 700 customers, with 800 within reach, but the target is far bigger, 400K viable U.S. #trucking businesses. Its edge is in architecture and focus: a cloud-native TMS with more than 120 integrations, from Samsara and RMIS to DAT, Motive, Intuit QuickBooks, and Geotab. The hybrid carrier-broker design kills duplicate entry and stitches workflows together seamlessly. This isn't theory. It's proprietary tech born from Darman's own brokerage, sharpened for the entire market. The fresh capital fuels what comes next: AI-powered #dispatching, deeper #APIs, enhanced compliance, and a freight marketplace aggregating over a trillion dollars in opportunities. The team is stacked with heavy hitters, CRO Patrick M., VP of Marketing Jeremy Hermanns, VP of Brand and Community Ava Barnes, VP of Customer Success Emily Feliton, VP of Product John Carofano, all building toward one aim: make midsize carriers as efficient as giants while still serving the enterprises running 6,000 loads a month. Alvys isn't playing catch-up. It's pulling the industry forward. When technology can deliver time, margin, and trust in one system, it's not just another funding announcement, it's a signal that the future of logistics is already in motion. #Startups #StartupFunding #VentureCapital #SeriesB #AI #Automation #Freight #FreightTech #FreightOps #cloud #CloudComputing #Compliance #Technology #Innovation #TechEcosystem #StartupEcosystem If software engineering peace of mind is what you crave, Vention is your zen.

28

3 Comments

Merger of dbt Labs and Fivetran (which had recently acquired Census - a Fivetran company) announced today. The data layer is where AI succeeds or fails. I expect a lot of advancement in the data layer over the next 12-18 months. https://lnkd.in/ecwvX834 #martech

108

20 Comments

Been building something fun after hours lately using Cursor and Claude code. I'm building a SQL challenge platform—think Wordle meets LeetCode for current and aspiring data folks. As we grow the team South Shore Analytics, I've been thinking about how to keep everyone sharp and learning from each other. But I also remembered how interesting I thought it was that every new hire at Uber had access to SQL training on day 1 of the job when I started there. I'd love to be able to use this tool to help people that don't think of themselves as 'technical' or 'data' people to build the skills to self-service their data questions and to make smarter decisions quickly. Finance folks waiting 3 days for a "quick pull." Marketing teams making decisions on month-old exports. Ops managers who could answer their own questions if they just knew how to JOIN two tables, etc. What I'm Building: One SQL puzzle daily based on your ELO style rating. You see the target output, write a query to match it. Your score is dictated by problem difficulty, result accuracy, and time to complete. Compete within your organization or globally if you choose to. The goal is that it'll help our team stay sharp, but I'd love to see this product eventually unlock extremely sharp operators who are eager to move quicker while making data-driven decisions. SQL shouldn't be scary. It's just asking questions in a structured way. So far, have just been testing it myself, but excited to get a basic version released in the next few weeks. Anyone else working on side projects with Cursor, Claude Code, Replit, Lovable, etc.? What are you building just because it's fun?

70

14 Comments

Where’s data viz headed in 2025? New tools, new formats, and new expectations are reshaping how we design with data. These are some of the trends I’m watching closely (and already leaning into). 1) 𝗔𝗜-𝗔𝘂𝗴𝗺𝗲𝗻𝘁𝗲𝗱 𝗗𝗲𝘀𝗶𝗴𝗻: There’s lots of hype right around tools like bolt.new and Lovable right now. But the results still feel too basic and formulaic to me. I’m honestly more interested in tools like Relume and Visual Electric, which aim to enhance the design process in specific ways, not replace it. 2) 𝗖𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝘃𝗲 𝗗𝗮𝘁𝗮 𝗔𝗻𝗮𝗹𝘆𝘀𝗶𝘀: When Observable announced the debut of Data Canvases last week, I was stoked. We’ve had real-time collaboration in design tools like Figma for years. I’ve always wanted the equivalent for data teams and this feels like a big step. 3) 𝗔𝗻𝗶𝗺𝗮𝘁𝗲𝗱 𝗗𝗮𝘁𝗮 𝗦𝘁𝗼𝗿𝗶𝗲𝘀: More and more clients are asking us to create animated videos with their data. Think short GIFs (<30 seconds) featuring a slick chart and a single, sharp insight, designed to stand out in a crowded social feed. In a world driven by TikTok and short-form video, this makes sense creatively and strategically. 4) 𝗪𝗲𝗯𝗚𝗟 𝗮𝗻𝗱 𝗦𝗵𝗮𝗱𝗲𝗿𝘀: If you're looking to make your visuals pop, WebGL and custom shaders let you push the browser to its creative limits. We’re using them to build visuals that feel dynamic, immersive, and a little bit magical. 5) 𝗡𝗼-𝗖𝗼𝗱𝗲 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺𝘀: Tools like Webflow and Framer are democratizing web development and taking dev timelines from months to weeks. We’re leaning into this on specific projects more and more. 6) 𝗜𝗺𝗺𝗲𝗿𝘀𝗶𝘃𝗲 𝟯𝗗 𝗩𝗶𝘀𝘂𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻𝘀: The rise of Three.js and Blender is helping 3D find its place in data storytelling. A few years ago, I would have dismissed this as a fad. But recent examples from NYT and Washington Post, where they’ve used 3D to great effect, have convinced me otherwise. Data people, what’s on your radar this year?

18

1 Comment

Paraphrasing about 3-4 Databricks customers i chatted with yesterday: "We like Databricks because they listen to our feedback and are transparent as to what of our feedback is likely to materialize into what we need or not."

176

4 Comments

As a data nerd, the rise of the GTM Engineer championed by Clay feels a lot like the emergence of the Analytics Engineer ushered in by dbt Labs. In both cases, a tool arrived at the perfect time and perfectly solved the problems teams were facing. Data people rallied around dbt, and it quickly became a core part of the modern data stack. It is now nearly ubiquitous. Now, Rev Ops, Marketing Ops, and Sales Ops teams are seeing something similar. The GTM Engineer, powered by Clay, is becoming the go-to solution for the complexity and fragmentation in modern go-to-market workflows. Today’s fundraising news is a testament to that momentum. Excited to share that Clay has raised $100M at a $3.1B valuation. Check out our announcement via the The New York Times in the comments below.

145

6 Comments

CDPs won't exist in five years. Let me explain why. CDPs were an answer to a real problem 10 years ago. But they were fundamentally the wrong answer. Why? CDPs were typically bought by marketing teams to solve what is actually a data problem... ...A problem that data teams SHOULD OWN. When you look at what a CDP does, it's handling customer data collection, identity resolution, making the data accessible to marketing and other business users (ideally in their existing tools), etc. As data teams finally get the charter to solve these problems, they're building solutions in a composable way and using best-of-breed tools for each part of the pipeline. This is why CDP as a category won't survive. Not because we don't need customer data platforms, but because what we really need are: → Data teams that understand marketing use cases → Marketing teams that understand data requirements → Composable architectures that respect data governance So if you're looking at buying a CDP right now, I'd ask: Have you talked to your data team first? Could they build you something more sustainable with the RIGHT tools? And if you're a data leader, are you proactively solving marketing's customer data challenges? The future is clear: customer data is a data problem. And data teams that solve it well will drive tremendous business impact. https://lnkd.in/eEXKNVR6

235

77 Comments