His Data Clients Include the NBA, UFC, and Princess Cruise Lines. He Does It All Himself. And He's Fine With That.
Tim Shea has been in data analytics for 25 years. He's started five companies. He's worked with Nike-scale brands and garage startups and he'll tell you they all have the same problem: their data is stuck in 20 or 30 different places and nobody can answer a simple question on Monday morning without spending half the day pulling spreadsheets that don't quite fit together.
He built a company called Latticework to fix that. His clients include the NBA, UFC, Princess Cruise Lines, and Reddit. He works with retail and DTC brands in the $50 to $100 million range. He had five employees during Covid, went back to solo, and has no plans to hire a team.
He calls VC-backed startups "another form of poverty." He calls his current setup "crushing it."
He's not wrong.
The Data Problem Nobody Outgrows
The moment you start a business, your data starts scattering. Facebook ad results go one place. Google analytics go somewhere else. Shopify has its own dashboard. Amazon has another. You add Salesforce, a supply chain tool, a returns platform, a subscription system. Before long you have 30 logins and a Monday morning ritual of downloading spreadsheets that were never designed to talk to each other, jamming them together, and hoping the answer that falls out is somewhere close to real.
Then you do it again next Monday.
Tim's whole thesis is that the insight is in there — it's just buried under the complexity of how modern businesses actually operate. His job is to build what he calls the latticework first: get all the data into one place, clean it, give it a coherent structure. Once that's done, the questions answer themselves.
For early-stage founders, he's practical about when this actually matters. If you're a million-dollar business selling on one channel, Shopify analytics is enough. Don't call Tim. But the moment you're growing fast, saying yes to Target, experimenting with subscriptions, running drops, hiring a team that signed up for 15 different tools — that's when you need someone who can look at all of it together and say: here's what's actually working, here's what isn't, and here's the one metric that should be running your whole operation right now.
Moneyball for Retail Brands
The baseball analogy is Tim's favorite, and it holds up.
Billy Beane didn't win the World Series by optimizing for home runs. He found undervalued metrics — on-base percentage, walks, things other teams weren't paying for — and built a competitive team on a fraction of the budget. Tim applies the same logic to retail: there are high-leverage, counterintuitive metrics buried inside most companies' data that, if you double down on them, produce outsized results. The problem is you can't find them if your data is still stuck in 30 spreadsheets.
His version of the Moneyball wrinkle: lifetime value, broken into segments instead of averaged.
Most companies talk about LTV like it's a single number. It's not. You have Black Friday customers who maybe spend big on the first purchase but don't come back for a year. You have January customers who sign up during a promotion, pay a lot to acquire, but then buy every single month like clockwork. Those are two completely different businesses sharing the same P&L.
The Athletic Greens example is the clearest illustration. Their CMO told Tim the average customer stays on the product for two years at $99 a month. Run that math — and then ask yourself what you can afford to pay to acquire someone like that. The answer might surprise your finance team. It won't surprise your marketing team.
Tim's point: if you know your customer segments, you know your payback period, and if you know your payback period, you know how much to give marketing. A CEO with a finance background panics when acquisition cost hits $1,000. A CEO with a marketing background says: I know what that customer is worth over 36 months, and $1,000 is cheap. The data is what lets you have that argument with evidence instead of opinion.
The End of the BI Dashboard Era
Tim used to love Tableau. Would hug it, he said. Would cuddle it like a teddy bear.
He's done with it now.
The shift is AI. Instead of paying for Tableau or Looker or Sigma — platforms that impose their own limitations on what you can visualize and how — Tim points AI directly at a curated semantic layer in the database and generates interactive HTML dashboards custom-built for the specific question being asked. No platform in the way. No "Looker doesn't support that one thing." Just: here's the question, here's the data, here's a visualization that lets the CEO run scenarios in real time and watch the dollar numbers move.
The thing that was a huge pain two or three years ago — "what if this metric went up by just one more percent, could we show what that's worth?" — is now just a prompt. The $50,000 opportunity hiding inside a line that's going slightly the wrong direction on a chart is now findable without a full sprint from an engineering team.
He's not saying AI replaces the domain expertise. He's saying it removes the platform tax. You still need smart humans who know what questions to ask. You just don't need to pay Tableau for the privilege of asking them.
Five Companies, Back to Solo, Still Going to the Moon
Tim grew up in Boston, moved to LA in 1999, discovered that straight A's don't mean much when you're hustling in a new city, and stumbled into entrepreneurship the way most people do: by realizing he couldn't stay in a 9-to-5 job reading CNN until 5pm.
Five companies later, he's back to running Latticework solo. He had employees during Covid. It didn't work the way he wanted. A good friend gave him a piece of advice he took seriously: some people make the mistake of being really good at something and then hiring a cast of characters who end up doing the work at a lower standard. Tim delivers the work. If he needs contractors, he brings them in. But the client gets Tim.
Some people hear "solo consulting business" and say lifestyle business. Tim says fine, call it that. But he's buying toys, taking vacations, running side quests, and working with genuinely interesting brands on genuinely hard problems. What exactly is poverty about that?
He has a pointed take on VC, too. When someone hands you a $2 million seed round, they're also handing you a deadline: come back in 18 months, three times bigger, ready for the next round. That's a specific type of pressure with a specific type of outcome range. Tim isn't anti-VC in theory. He's just clear-eyed that debt is cheaper than equity if you're already making money, that buying businesses is an option, that grinding your service-based business into profitability is an option, and that none of these require a term sheet.
Crushing it, he says, is an option.
When the Right Idea Hits
The thing about building skills across five companies and 25 years of client work is that when the next big idea comes along, you have the toolkit for it. Tim doesn't have a specific next venture announced. He doesn't need to. He's built the financial base and the optionality. When something clicks, he can move.
Stu made the same observation about himself. His first business was a Google ads agency. His second was an edtech company he built to 170 people and sold to private equity in 2021 — with only friends-and-family seed money, never institutional, and the PE firm still called it "bootstrapped." Now he's building a baseball resort on 191 acres. None of those three businesses are obviously connected, except through the skills and relationships built doing the last one.
That's the actual moon, for some founders. Not the IPO. Not the unicorn valuation. The life where you've built enough and learned enough that you can actually choose what comes next — and go do it well.
What Stuck With Me
This episode is a conversation between two people who have both figured out what they want and built their way to it. There's no struggling founder moment. No dramatic near-failure. Just two guys who've been in the game long enough to be honest about what it actually looks like.
Tim's most useful line is probably the one about True North metrics. Every company at every stage has 1-2 numbers that should be running the show. Most companies have 30. The work of building well is the work of simplifying: what does this business actually optimize toward, and is all the data we're drowning in serving that goal or obscuring it?
He's been asking that question for 25 years. He's still finding people who need to hear the answer.
Now you have a friend in the data analytics business.
Recorded live at a Startups with Stu retreat. More at startupswithstu.com.
Subscribe for Updates
Subscribe to the Startups With Stu newsletter and get advice and stories sent straight to your inbox to help launch your business to the moon!




Comments