It Can't Be Messy, It Has to Work: A Conversation with Kas Szatylowicz

On a podcast literally called Messy Growth, Kas Szatylowicz opened with a line that pushes right back: it can't be messy, it has to work.

Kas is a fractional head of marketing and go-to-market advisor, and one of those rare operators who pairs deep product thinking with hard revenue accountability. She's been the first marketing hire at multiple growing startups, led growth at AI and cybersecurity companies like Lakera and V7, and brought a chunk of the Fortune 100 into an inbound funnel, all with lean teams and zero fluff. She brings an engineering mindset to marketing, thinks in systems, and cares about exactly one thing: pipeline.

The personal brand carries, the job resets

Ask why she's been headhunted for every role since 2018, and the honest answer is her LinkedIn presence, built long before the AI-content flood. Not just posting what she's doing but how, staying actionable, and openly sharing failures and work in progress to get input back. The offers don't come through recruiters, they come through VCs reaching out about portfolio companies, because she consistently made the founders she worked with look great to their boards, and because she networked. As she and I agreed, your personal brand carries from job to job. The job itself resets, you have to prove your worth again, but the brand comes with you.

Foundations before experiments, and killing your darlings

Walking in as the first marketer, the psychological shift is curiosity, being willing to feel like the stupidest person in the room and ask everything, especially in a heavily technical field like computer vision she knew nothing about at first. The second is experimenting hard and killing your darlings, because marketers tend to kill things too late, staying with a strategy because they've sunk time into it.

But she explicitly rejects "move fast and break things," which just creates chaos. Her split is roughly 40% building foundations and processes, 60% experimenting. When she arrives, the CRM is usually a mess, there are no sales or marketing processes, nothing lives at the product-marketing intersection, and everyone's working in silos while everything burns. And there's rarely resistance to fixing it, early-stage teams don't know any better and are grateful someone takes the reins.

What the engineering mindset actually means

Kas did Python and JavaScript courses, not to write production code but to stop being scared of it, to set things up via code and not freeze when she sees a terminal or a VS Code window. Beyond that it's thinking in systems and being relentlessly data-driven. It doesn't clash with creative marketing, she says, it's a competitive edge, and she's naturally analytical rather than creative. Which feeds her core stance: she's the marketer who focuses on making things work, not making things pretty. At early stage, pretty doesn't matter, A/B testing your landing page when a thousand people visit a month isn't moving the needle. A lot of brand-obsessed marketers, she notes, do badly at early-stage startups for exactly this reason.

Lazy and lucky: 3,500 to 90,000 in 11 months

Her honest read on scaling organic traffic from 3,500 to 90,000 in under a year is "lazy and lucky." Lazy because instead of pretending to be a subject-matter expert, she went to her founder, admitted she couldn't write about computer vision, and outsourced content to ML engineering students on Medium, real experts who could handle the code and equations that generalist writers can't. Lucky because the timing predated the LLM boom and competitors weren't doing SEO. There were some tricks she's not proud of, a broad ML glossary of thin content covering many terms, but the traffic converted and generated the majority of the company's inbound pipeline.

The bigger lesson: outsourcing to smart specialists isn't lazy, it's smart, but you still have to know what "great" looks like to manage them. She's watched countless founders pour money into design, PPC, SEO, and PR agencies and get almost nothing back, because they couldn't evaluate the work and got sold the dream. Outsource, yes, but be deliberate about to whom.

Stop obsessing over traffic

The metrics growth marketers obsess over that barely move revenue? Traffic, which she calls a vanity metric especially in B2B, and social media, obsessing over whether every post is perfect when nobody remembers it tomorrow. She only reports pipeline generated. The single dashboard number that tells her a team understands growth is qualified meetings, then deals. Nothing before, nothing after.

On MQLs, she thinks the term is loaded and that holding onto one fixed MQL definition is a way of lying to yourself, especially when you're industry-agnostic and the same job title means completely different problems in agriculture versus defense. Better to consider the whole account's health and come back to the North Star: do they have a use case, are they willing to buy, do they have the money.

PLG and enterprise at the same time

Running PLG and enterprise motions simultaneously is the most common thing she sees and close to the worst, two people on a marketing team, one on each, blasting everybody with everything. Enterprise needs time, six-to-nine-month sales cycles, and PLG is often over-engineered. Her cybersecurity example is sharp: engineers try the product with a Gmail, not a work email, so teams ignore those signups, when a quick LinkedIn check reveals a CISO at a Fortune 500. Don't discard those leads, enrich them, and sequence the two motions properly instead of doing everything at once.

For big accounts, the answer is value

Asked whether landing a six or seven-figure account is about messaging precision or sales alignment, Kas offered a third thing: providing value. Understand their pain, nurture them, build trust. Salespeople too often assume someone will hop on a call without knowing them. She likes The Challenger Sale's instinct, teach the prospect something, give them insight, gently challenge them, rather than leading with relationship alone, then sell.

Clarity is the whole game

Her teams held eNPS above 50 and a manager score around 95%, in pressure-cooker roles. The secret is one word: clarity. People burn out when there's no clarity, expectations get missed or misunderstood, and accountability disappears. Clarity means everyone genuinely agrees on what we're doing, why, and how. It sounds simple, but in an early-stage startup where everything feels like a priority, it's one of the hardest things to reach, and it has to come from the top. Her favorite tool for it is the not-to-do list: alongside the to-do list, an explicit list of what you're not going to do, because otherwise you end up doing everything.

The founders she loves, and the dangerous pattern

She loves working with technical founders, more structured, better at thinking about edge cases, versus business-background founders who "follow the money, we'll figure it out" and promise customers things without understanding the work required. And the dangerous pattern she sees repeatedly at seed and Series A: founders focusing on satisfying VCs and looking good for them, and obsessing over competitors, instead of obsessing over their customers and the problems they're solving.

Key takeaways

Foundations before experiments. Roughly 40% building processes and analytics, 60% experimenting. "Move fast and break things" just makes chaos.

Make it work, not pretty. At early stage, brand polish and landing-page A/B tests don't move the needle. Function first.

Traffic is a vanity metric. Report pipeline generated and qualified meetings. Nothing before, nothing after.

Don't run PLG and enterprise at once. Sequence them. Enterprise needs six-to-nine-month patience; blasting everybody with everything fails.

Clarity prevents burnout. People burn out from missed expectations and no accountability, both symptoms of no clarity. Keep a not-to-do list.

Frameworks worth stealing

The 40/60 split

Coming into an early-stage company, spend about 40% of your energy building the missing foundations, CRM hygiene, sales and marketing processes, unfragmented analytics tools, and 60% experimenting. Don't jump strategy to strategy without a base, and don't over-build process either.

The not-to-do list

Alongside your to-do list, maintain an explicit not-to-do list. In an environment where everything feels urgent, naming what you're deliberately not doing is what creates focus, protects the team from doing everything, and makes accountability possible.

Quotes worth keeping

The lines I wrote down.

It can't be messy. It has to work.

Traffic is a vanity metric. I only care about pipeline generated.

People burn out because there is no clarity.

Rapid fire round

Same questions every guest. Quick questions, quick answers.

Best advice you've ever received? Feedback to look more broadly and not stay narrow in her field, which at the time was outreach and backlinks.

Advice you ignored and wish you'd listened to? The advice to take equity and accept a below-market salary. She rejects it, take both, and choose a place where you can have both.

What would you tell your younger self? Learn to code.

Ongoing challenge that keeps you up at night? Keeping the CRM organized and up to date, because now that you can connect LLMs and build Claude skills for the CRM, garbage in really is garbage out. It all comes down to data quality.

Favorite spot? Youmiko, the vegan sushi place. Both she and I share a strict rule: never queue for food.

Tool you can't live without? Claude. She does everything in it, work and personal.

---

Kas Szatylowicz is a fractional head of marketing and go-to-market advisor who has been the founding marketer at multiple startups and led growth at AI and cybersecurity companies including Lakera and V7. She works mostly in deep tech and machine learning, and is building a parallel career in psychotherapy. Find her on LinkedIn.