Thread

Most teams pick a problem the way they pick a restaurant — whoever speaks loudest, or whoever's idea is least embarrassing to disagree with. We didn't want the loudest idea; we wanted the best one. So we structured the choosing.

Each of us generated ideas alone first, following Art Markman's argument in "Your Team Is Brainstorming All Wrong" — that group brainstorms quietly converge on the first person to speak and the most socially safe idea, and that independent generation produces more and more diverse options. Only then did we come together to build on each other's ideas IDEO-style: yes-and instead of yes-but. Then we voted on our top five.

Dating won decisively. Not because it was the most fun problem to say out loud at a party, but because it was the one with the widest gap between how much it matters to people and how badly the existing solutions serve them.

Before we trusted our own conviction, we went to Pew's 2020 research on dating in America. It confirmed the size of the problem and, more usefully, the shape of it.

The frustration wasn't an edge case — it was the median experience. And the why split sharply by gender, which told us one nudge for everyone would fail.

If the apps were growing the market without improving the outcome, the question wasn't "how do we build a better app." It was "what are all the existing services actually optimizing for, and what's nobody serving?"

We mapped the landscape on two axes: how gamified an experience was, and how much real connection it produced. The picture was clean. The swiping apps were fun but not rewarding — engineered for time-on-app, not relationships. The matchmakers and serious services were rewarding but not fun — effective for some, but a chore. And a long tail was neither.

The top-right quadrant — rewarding and fun, high on both — was empty. That white space was Thread's position. A coach makes the process genuinely rewarding; behavioral-science design makes it feel like progress instead of a second job.

"Single people" is a demographic, not an audience — it tells you nothing about what someone wants or how to reach them. We narrowed to a posture, not just a status: single working adults who live in urban areas, value experiences, and are dating with intention.

Secondary research told us the market was real. It couldn't tell us what someone actually feels at 11pm after a date that went nowhere. So we ran 12 semi-structured interviews — eight women, four men — 30 minutes to two hours each, going deep rather than wide on purpose.

Four themes carried across every conversation — and together they pointed straight at the product.

Everything above pointed somewhere specific, and our first design followed it: an app for low-stakes, in-person group dating, built to take the pressure off the one-on-one and put people in front of each other doing something real.

Before committing, I ran the whole category through the Ten Types of Innovation — generating ideas against all ten types (the board below is the unedited version of that thinking), then asking where we could actually win. The incumbents turned out to be generic: they competed on brand and on the size of their network — who else was on the app — and almost nothing else. We set out to differentiate across six of the ten types, four of them wide open.

The group-dating app was the obvious build. We didn't build it — and walking away was the best design decision we made. Three things were honest dealbreakers.

Coaching inverted every one of those risks. It needed no network — we could bootstrap it one member at a time. It needed no engineering org — a service launches with people and a thin software layer. And it was riding a real shift: coaching was going mainstream through books like Logan Ury's "How to Not Die Alone," which reframed dating as a skill you get better at, not a pool you optimize. Apps were saturated; coaching was fragmented and open. So we went and built the thing the category had left empty.

Choosing coaching forced a sharper question: who actually pays for it? The answer was a phrase we'd been circling all along — people dating with intention.

That last phrase did the most work. Dating with intention separates the person who wants a relationship and is willing to invest in getting better at it from the person who's content to swipe for entertainment. The first person will pay for a coach. The second never will. Designing for intention is what made the rest of the product — onboarding tasks, self-assessment, a human coach — feel like help instead of homework.

Thread is like having a dating coach in your pocket. We paired each member with a coach who delivered customized insight, rooted in behavioral science, to help them become the best version of themselves while improving their dating outcomes. It was coaching-as-a-service. It was deliberately not a swiping app — there was no deck of faces to burn through, because the product's job was to change the dater, not the dating pool.

Two surfaces carried it. A marketing landing page led with the thesis — Date Smarter — and made the coach-in-your-pocket promise legible to someone who'd never heard of us in the three seconds before they bounce.

Behind the login was the member dashboard, and it opened on a name, not a feed: "Hello Rob!" The work was structured as onboarding tasks that doubled as the coach's intake — a self-assessment and a dating-profile quiz that turned the member's own answers into the raw material for personalized guidance. From there: an assigned human coach, session booking and appointments, a referral credit system, and an "article of the week" — "Understanding Attachment Theory" being exactly the kind of thing — so the relationship kept teaching between sessions instead of going quiet.

Thread started as four MBAs — me, Sarah Jane Tong, Charles Moody, and Samantha Sutton — though for most of the run it was three: Sarah Jane left first, Charles after about two years, and Samantha and I finished it out. The core stayed deliberately tiny.

Everything that core wasn't, we hired in. Behavioral scientist Matt Wallaert designed the behavioral study, so "rooted in behavioral science" was a method, not a tagline. Dating coaches Samantha Burns and then Stephanie Salem — both contractors — made the coaching real, and KC Westbrook contracted on design.

The engineering was rented too. I built the first cuts myself in Webflow, and when the integrations outgrew me, I brought on Farewell Media and managed them through building the final product end to end — landing page, payment flow, onboarding, and dashboard — with Outseta and Stripe handling billing and payments underneath.

We pitched twice. The first got us into CreateX, Georgia Tech's accelerator — a $5K SAFE, no cap, no equity, effectively non-dilutive money to begin. Once we'd incorporated, we pitched again to the incubator's sponsor, Fusen Capital, who backed us with a tranched SAFE worth up to $505K, released against growth milestones.

We earned the first $25K tranche and didn't hit the thresholds for the $125K and $350K that would have followed. That's the honest shape of it: enough conviction from real investors to start, and a clear bar we didn't clear. Both are true, and I'd rather show you both.

Every dating app looks the same: a gradient, a sans-serif, a heart. That sameness is a message — it says "we are a utility, swipe efficiently." We needed the opposite signal. Thread had to feel warm, human, and a little playful the instant you saw it, because we were selling a relationship with a person, not an algorithm.

But before any of the marks, type, or color, there was a single image that fixed the feeling the whole brand had to carry.

Two graphic elements carried the brand's rhythm, movement, and emphasis across every surface.

Goofy, playful, and emotive, juxtaposed against relatable and trustworthy — two registers of imagery, and the contrast was the point.

A two-surface product still has to show up across a dozen contexts, each with its own rules — a search ad has three seconds, an Instagram grid has nine tiles, a blog post has to earn a ranking. So I built the brand as a system that generates on-message creative at volume, not a set of one-off layouts.

Over roughly three years, four of us got to build something real — a marketing site, a logged-in dashboard with human coaches and behavioral-science guidance, and a full go-to-market system across ads, social, and SEO. We were fortunate twice over: investors backed the problem with $505K, and real people paid to try the service. It didn't reach the scale we were after, which is part of the deal with trying something hard.

Mostly, I'm grateful. I learned more in three years of building Thread — about fundraising, leading a team, shipping without an engineering org, and reading a market — than any role could have taught me, and I'd take the same swing again.