I'm Trevor Fox. I build growth systems for technical companies.

I run Elegant Atomics, a growth practice for early-stage B2B SaaS. Mostly data and AI infrastructure, developer tools, and technical products. There are three of us.

I figure out how things actually work, then build systems that take advantage of it. That applies to a search algorithm, a buying committee, and a category that doesn't have a name yet.

Who I work with

Founders between seed and Series B with a technical product and a market that doesn't quite have a name yet. Two situations come up most: you've built something real and can't yet explain it to the market, or you can explain it and can't yet reach enough of the right people.

The other way in is the problem itself. Growth and GTM strategy that has to be reasoned out rather than copied. Growth engineering, where the answer is a system somebody has to build. Experimentation programs that need designing before they need running. If it's tangled and there's no playbook for it, that's the work I want.

If your category already exists, your playbook is written, and what you need is hands to run it, I'm not the best value you can buy.

What I do

Full service, discovery through measurement.

Strategy first: positioning, the category argument, where demand is going to come from. Then the build: SEO and content, inbound and outbound demand generation, web, marketing analytics, AI-driven automation. Then the measurement that shows whether any of it worked.

I'm in it end to end, from the first diagnostic to the numbers that prove it. You get me, not a strategy deck handed to someone else to run.

I helped Definite find product-market fit and generate $1.5M+ in pipeline in its first year. Before that I scaled Panoply's inbound trials 3x organically at Series A, and spent two years at Census on market development and category creation for reverse ETL, Series A to Series B. I've built identity and positioning from nothing for seed-stage startups, and I've done the unglamorous version too: audits, instrumentation, and the reporting that finally makes a number trustworthy.

Data and AI infrastructure is where I'm sharpest, but the approach transfers. I like the hard problems most, and I move fast through the rest.

If that's the shape of your problem, book 30 minutes.

How I work

I don't start with playbooks. I start with how the system actually works, then build on top of it.

That means taking things apart. Tearing down how AI search pipelines retrieve and cite information. Finding the layer a platform didn't mean to expose: a 41-bit timestamp buried in every LinkedIn post ID, an entity ontology sitting inside ChatGPT's frontend payloads. Then building the thing that uses it.

Chrome extensions, tools, automated outbound pipelines, content systems driven by AI agents. I build what I need to learn, and I learn what I need to build.

Growth work grounded in how markets, systems, and buyers function, not how a blog post says they should. Testable systems that show what's working, why, and how to scale it.

What people say

I've mentored on GrowthMentor since 2019. That's 83 sessions with founders and operators, and the one thing I've never stopped doing. There are 59 public reviews. The pattern in them is the part I care about: go deep enough to find the real problem, then leave people something to do about it.

"Spent over an hour diving into our GSC data and proposed a number of clear next steps to our challenges that we're excited to implement."

Max Fram-Schwartz, 2023

"We got into the weeds and back out. His range of marketing and technical skills helped me focus on how to leverage my engineering resources to help with marketing initiatives."

Oz Merchant, 2019

"He didn't just give surface-level feedback. He pushed me to think more critically about the idea and how to actually validate it in the real world. The conversation helped clarify what matters versus what's just 'nice to have.'"

Alexander Franklin, 2026

The other pattern is subtraction: cut the strategies that won't work before anyone spends on them, so the effort goes where it moves something. Same instinct I bring to client work.

What I build

The practice funds products I own. Building them is how I keep my hands in the work I sell.

Flux is LinkedIn analytics built for individual creators instead of company pages. No login, no extension, and it works on anyone's public profile. Spandrel is an open-source framework that turns folders of markdown into knowledge graphs an agent can navigate, and Context Cannon hosts them. Both are growing. Flux adds customers every week, and Spandrel's early adopters are building it into their GTM operations.

What I believe

AI can execute faster than any team. So vision, planning, and architecture are the real leverage now. Design the systems, leave the mechanical work to the machines.

Which makes context the whole game. Model capability and pricing are set by someone else. How you assemble the context you feed a model is the part you control. Context engineering is what humans do for models, and autonomous agents have to do it for themselves. The same discipline, applied recursively. That's the idea Spandrel came out of.

The goal isn't more traffic or more leads. It's a durable growth engine that's sustainable for the business (and hopefully for people and the planet).

How I got here

I read Moneyball on the injured list, in the middle of a short pro baseball career. It captured how I thought about competition. The edge comes from seeing patterns others overlook. I started a blog while playing overseas, accidentally fell into web marketing and analytics, and devoured everything I could find about web technology, data, UX, and marketing.

That took me through Postmates and Panoply to Census, where I spent two years arguing reverse ETL into existence as a category. In every startup job I'd had, about half the work was the part I wanted, the hard problems and the big things worth building. So I started Elegant Atomics to do that half all the time, and aimed it at early-stage data infrastructure, the exact market I'd spent two years building demand for.

The rest

I grew up in Oregon and Puerto Rico. I studied Communication and Philosophy at UC Davis. I played professional baseball in the US, the Netherlands, and Australia. I co-founded November Project Hong Kong, a free twice-weekly fitness community that still runs today. I've been writing this blog about marketing, analytics, and web technologies since 2013. I'm a husband and father, and I live in beautiful Portland, Oregon.

Reach out on LinkedIn or GitHub, or by email.

To talk business: calendly.com/trevor-fox/30min

To untangle growth and career challenges: app.growthmentor.com/mentors/trevor-fox

Thanks for stopping by!