Surface Area Compounds: A Conversation with Melissa Withorn

Every industry has that one problem everybody complains about and nobody fixes. In the creative world it's freelance talent. Agencies depend on freelancers every single day, and the systems behind that dependency still look like they were built a decade ago. Talent lives in spreadsheets. Portfolios disappear into forgotten folders. Hiring starts with someone posting in Slack: does anybody remember that designer we worked with last year?

Melissa Withorn spent 17 years inside creative agencies, most recently as a VP group creative director, before deciding that pattern was worth building against. CAST, the platform she started from a hackathon five months ago, is her attempt to give agencies actual infrastructure for the bench they now depend on. She's six months into founding, building in public from Portugal, and refreshingly unwilling to pretend she has it figured out.

The industry that never fixed its own ops

Agencies live on the cutting edge of creative work and nowhere near it on their own operations. "Things take a long time to turn over," Melissa says. "It can be a beast to redirect the ship." For a long time a list somewhere was good enough, and that was a defensible position when the problem was small. You'd bring on freelancers for a big pitch, maybe five a year, and if it took someone all day to remember the name of the motion designer you loved five years ago, it took all day. Fine.

What changed is volume and speed. Full-time benches are shrinking while freelance usage becomes routine, which means the bench is no longer a backup plan. "Freelancers are the backbone of most agencies. It's how they win work." When you need to hire someone tomorrow and you need to do it constantly, an all-day search stops being a quirk and starts being a competitive liability. That's the urgency Melissa says she's finally seeing, and it's why she thinks it's high time the problem got treated with real infrastructure rather than duct tape.

The pattern she couldn't unsee

There was no lightbulb moment. There was accumulation. Melissa has been the freelancer, the fractional leader, and the agency person trying to track someone down, and she describes the system as not really working for anyone on either side. What made it undeniable was breadth: "When you work at one place, you can think that maybe this is just inherent to that one place." Freelancing across a lot of agencies removed that excuse. Everyone was dealing with the same shade of the same problem.

She built the early product with the people who actually run the work, producers, ops leads, studio managers, project managers, the ones doing the tracking down. A lot of what's now in CAST, particularly around onboarding and managing freelancer calendars, came directly out of those conversations.

Keep it small, keep it specific

The hard part wasn't gathering input. It was surviving it. Melissa is candid that a single great conversation with a big global agency, rattling off can-it-do-this after can-it-do-this, derailed her for two full days and sent her down a week-long rabbit hole. Her diagnosis of why it landed so hard is the useful part: "When I first started, I didn't really know who I was building it for. I was just building for the problem."

Building for a problem instead of a person is how you end up creating the world. The correction was to picture a specific person, a specific agency, and get genuinely niche. She's also honest about the emotional mechanics, which is where most founders actually lose the thread. People tell you it's great, they ask what if, and you get swept up in being showered with compliments. Her rule now is to stay off both extremes: don't build for the top 1% or the bottom 1%, keep it middle of the road.

The archetype problem

Walking away from a VP group creative director title costs something, and Melissa doesn't pretend otherwise. What she describes grappling with isn't mainly the money, it's the narrative. The founder story as it's usually told is 20-something, Bay Area, Stanford, MBA, and she is a 38-year-old mom in Portugal with a 17-year agency career behind her. "It can be hard sometimes not to let that impact you, to hear that talk track playing in the back of your head."

Her answer isn't to argue with the archetype. It's to price her own edge accurately. Someone with more money or more experience could probably build CAST, and she's decided that isn't hers to worry about. What she has is the relationships and a better understanding of the problem than anyone else in the room. "I'm qualified, and I know what the hell I'm talking about." The proof shows up in demos, and the compliment she takes most to heart is when an agency tells her she understands what they're up against. That's the competitive edge she's actually building on.

What Lovable gives you, and what it doesn't

CAST's MVP was built on Lovable, out of a hackathon, with no intention of becoming a business. Melissa is unambiguous about what that unlocked: "I was never going to have this nugget of an idea and then hire a dev and spend $20,000. That was just never going to happen." A business with beta customers is a reality she says she'd never have reached five years ago.

She's equally unambiguous about the failure mode, and this is the sharpest thing she says about the current moment. You can spend $500 accidentally in a month vibe coding, never talk to a customer, never interrogate the problem, and "build and build and build this gorgeous sandcastle and not really know if it's solving a real problem that people will pay for." For personal tooling, she thinks the tools are genuinely amazing. But the gap is real: "The transition from Lovable into a full-fledged profitable business is where a lot of these projects die, quite frankly." The tooling collapsed the cost of building. It did nothing to the cost of being right.

Building in public as a substitute for an office

Asked whether building in public is marketing or philosophy, Melissa's answer is that the line is extremely blurry and she has no idea what she's doing constantly. The honest driver is that she doesn't have an office anymore, and this is how she stays connected to people.

What she didn't expect was the response to being unpolished. The instinct is to present like there's a whole team behind you rather than one person masquerading as one. Instead, people liked the ground floor. They reach out to say this is me, I can't believe you're doing this, I ate a chicken nugget over the sink the other day and I feel like a feral founder. That recognition became the Feral Founders Club, and the broader lesson is one most builders underrate: "People are cheering you on more than you think."

One person logs it, ten people build it

With real agencies paying real money, the roadmap question gets sharper. Melissa's early instinct was to take everything back to the build, and she calls that a mistake outright, mostly because of the wasted time. The rule she landed on is clean enough to steal: one person asking for something gets logged and thanked. Ten people asking for the same thing gets designed and built.

Underneath that is a discipline about time horizon. She has big plans for CAST, and she's deliberately not building for them yet. "It can be really easy to get caught in the five-years-from-now and build for that." Today the entire focus is retaining beta customers and making them see the value.

Free for freelancers, forever

The marketplace has an obvious hard side. Freelancers have thin budgets, no patience, and strong opinions. Melissa's resolution is to stop trying to monetize them at all: free for freelancers, forever, with a paid tier that opens up features. Profiles and collaborators cost nothing.

She justifies it from both sides of the ledger. Personally, she's a freelancer and knows the stack, the Adobe subscription, the Calendly, the SaaS fees that pile up. Strategically, a dense freelancer population is exactly what makes the product valuable to the agency paying the bill, because discovery through an existing network is the whole point. And it sets up the piece she's most animated about: turning referrals into income. Freelancers get asked constantly whether they know an amazing DP or someone who can build a Shopify site, and she spent years not realizing that was compensated knowledge. "Who you know that can crush a project for someone is worth paying for." Referral networks already exist and already work. She wants to operationalize them and give them scale.

Reducing the friction to zero

CAST solves a real problem, but it asks for behavior change from an industry with enormous inertia, and the underlying pain is chronic rather than acute. That's the hardest category of product to sell, and Melissa says it's the problem she wakes up trying to solve every day.

Her current answer is unglamorous and correct: stop making the customer do the work. The original model was DIY, buy it and load your people in. She learned fast that adoption dies there. Now she imports rosters herself for beta clients, and she's seen every possible input, Google Sheets, Airtable, links, even a screenshot of a CRM. All the agency does is log in and find their bench already there. "If I can make the platform intuitive and ready to go day one, then I've done my job."

Surface area compounds

Asked what she hopes the public documentation of this pivot looks like in two years, Melissa deliberately declines to make it about CAST's metrics. What she wants it to say is that a career can take chapters, that accumulated knowledge can be turned into something new, and that trying is available to you. "It doesn't have to be perfect. It doesn't have to be the version you think."

Her framing for why that works is the line that gives this episode its title: surface area compounds. Even if CAST doesn't work out, even if she never touches creative direction again, constantly putting yourself in the arena generates serendipity, and she firmly believes only good things come from it. The toll is the part people skip. "You have to be willing to look like an idiot, and be cringe, and just do it." Her evidence is the past year, and the number of people who came out of the woodwork to support her, offer introductions, and cheer her on.

Key takeaways

The bench became the business. Freelancers are how agencies win work now. Managing them in a spreadsheet was defensible at five hires a year and is a competitive liability when you need someone tomorrow.

Build for a person, not a problem. Building for the problem is how you end up creating the world. Get niche enough to picture a specific agency, and stay off both the top 1% and the bottom 1%.

AI collapsed the cost of building, not the cost of being right. Lovable makes the MVP possible for someone who doesn't code. It does nothing to tell you whether anyone will pay, and that gap is where most of these projects die.

One is a data point, ten is a feature. Log the single request and thank them. Design only for the pattern you hear over and over.

Adoption is your job, not the customer's. If the tool needs a migration, do the migration. Ready to go on day one is the product.

Frameworks worth stealing

The one-versus-ten filter

Every founder with early customers drowns in requests, and the failure mode is treating each one as a mandate. Sort by frequency instead of volume or account size. A single request, even from a flagship logo, gets acknowledged and logged, nothing more. When roughly ten people independently ask for the same thing, you've found a pattern worth designing against. This is what keeps a great conversation with a big global agency from costing you a week of building.

Do the migration yourself

When your product requires behavior change and the pain it solves is chronic rather than acute, onboarding friction will kill you long before your feature set does. Take the migration off the customer entirely. Accept whatever format the data arrives in, spreadsheets, Airtable, links, screenshots, and do the import by hand until it doesn't scale. The customer's first session should show them their own data already loaded, because a tool that asks a busy team to populate it is a tool they'll never populate.

Quotes worth keeping

The lines I wrote down.

Freelancers are the backbone of most agencies. It's how they win work.

The transition from Lovable into a full-fledged profitable business is where a lot of these projects die, quite frankly.

You have to be willing to look like an idiot, and be cringe, and just do it.

Rapid fire round

Same questions every guest. Quick questions, quick answers.

Best advice you've ever received? Reframe the no. Instead of counting rejections, go collect 100 of them, and be excited each time because it moves you closer to the yes. A game changer for a recovering perfectionist.

Advice you ignored and wish you'd listened to? Start building in public earlier. Coming from a world where everything is buttoned up, a deck and a site, she spent too long perfecting instead of putting it out before it was ready.

What would you tell your younger self? That if you'd told her she'd leave America, live in Portugal, start a company, and leave advertising, she'd have laughed in your face. When you're young everything feels permanent. She's finally enjoying the unpredictability.

Ongoing challenge that keeps you up at night? That it's just her. She has a developer now, but balancing giving the business her all, being with her kids, and actually enjoying the process is the thing that doesn't resolve.

Favorite spot? Breakfast in Portugal: a pastry, a meia de leite, and a pastel de nata. Her husband has run a taste test nationwide, and the ones in Belém really are that good.

Tool you can't live without? Claude and Codex. She's recently moved to Codex for database and code changes, finding it gets things right the first time. She doesn't code, so the business runs through them.

---

Melissa Withorn is the founder of CAST, a platform for how creative agencies manage freelance talent, and a fractional creative leader with 17 years inside creative agencies, most recently as VP group creative director. She also runs the Feral Founders Club, a newsletter and free community for people building on their own.