01 / Operator · Board Advisor · Warsaw
Fifteen years inside DTC growth — Meta, Google, programmatic, and the measurement stack underneath them. Margin-weighted bidding, cohort retention curves, incrementality testing. New customer economics that pay back inside your runway, and scaling decisions that show up in the P&L, not just the dashboard.
02 / What I Do
Most teams make growth decisions on data that was never built to answer business questions. Pixel attribution, platform reports, agency dashboards — useful for running campaigns, useless for running a P&L. I rebuild tracking, attribution, and the data pipeline underneath so the number the CEO sees matches the number in the bank.
ROAS is a reporting metric. Profit is a business outcome. I move the bidding surface from platform-reported revenue to nCAC, POAS, and payback period. The platforms keep working the same way. What changes is what they're told to optimize for — and which SKUs, cohorts, and channels survive the next quarter.
New customer volume is easy to buy. New customer volume that pays back inside your runway is harder. I separate the two — incrementality by channel, margin by SKU, retention by cohort — and scale only where those three agree. Growth that doesn't compound is just a burn rate with better PR.
Two board seats, three advisory engagements. The job is holding growth to the same standard as finance — same rigor, same cadence, same willingness to say no to the CMO. One day a month, quarterly packs, zero pitch decks.
03 / How I Think
Most marketing decisions are made on data that was never meant to answer business questions. Five principles I've defended against real P&Ls — and against the agencies that push back on them.
If a channel doesn't generate incremental revenue, it's noise — regardless of what the platform reports. Attribution tells you which touchpoint got credit. Incrementality tells you which spend you could kill tomorrow without losing revenue. In my experience the gap between those two numbers is 30 to 50 percent. On brand search and retargeting, routinely past 70. Geo-lift, PSA tests, CUPED on onsite data — pick one and run it quarterly. The first cycle pays for the engagement.
ROAS is a reporting metric. Profit is a business outcome. They are not the same number and they rarely move together. I optimize to POAS, nCAC, and payback — not to whatever the platform happens to export. The mechanic is unglamorous: pull COGS and fulfilment from the ERP, compute net contribution per line item, push it into the conversion API as the value parameter. Two weeks of engineering. Not a strategy.
Blended performance is an average of averages. It hides more than it reveals. Paid cohorts almost always retain worse than organic — often 30 to 40 percent worse by month twelve. "LTV to CAC of three" is a number without a denominator until you say what horizon it's measured over and which acquisition source produced it. I build retention decay by cohort month and channel, then compute payback against the twelve-month curve — not the theoretical lifetime the spreadsheet suggests.
You can't scale what doesn't exist. Most "growth hacking" is harvesting demand that was already there and claiming credit for the conversion. Real scaling means knowing which demand is latent, which is declining, and which you'll have to build. Search volume, category growth, share of voice, purchase frequency by cohort — those tell you whether the ceiling is actually where you think it is. Spending harder against a shrinking pool isn't a strategy. It's a burn rate in disguise.
Campaigns don't scale businesses. Systems do. A tactic is something that worked once and gets retired. A system is a measurement stack, a decision cadence, and a reallocation rule that keeps working after I leave. I don't hand over playbooks of what to do. I hand over rules for how to decide — which is the only thing that compounds.
04 / Selected Work
Names under NDA. Numbers are audited — annualized where needed, measured against holdout where the design allowed it. Where it didn't, I say so.
DTC Footwear · $60M revenue · US
Fractional Head of Growth · 9 months
Blended ROAS looked fine. New customer acquisition had quietly flatlined for four quarters while the platform kept serving ads to existing customers who would have bought anyway. We separated new-customer conversions at the pixel layer, switched bidding to nCAC and first-order POAS, and rebuilt prospecting creative against product-level contribution margin. Reported ROAS dropped. Profit didn't. That's the distinction the board eventually cared about.
DTC Home Goods · €25M revenue · 6 markets
Strategy Lead · 6 months
Platform reports said one number. The warehouse said another. Finance said a third. Every weekly meeting started with reconciliation and ended without a decision. We built a cohort-based LTV forecasting layer on top of the warehouse, calibrated against twelve months of closed finance data, and made it the only source the board used. Decisions started getting made in the first half of the meeting.
DTC Apparel · €80M revenue · 9 markets
Advisor · 8 months
Forty-plus creative assets a month, organized by platform. High volume, low signal, no learning carrying across channels. We restructured testing around demand themes instead: audience × angle × product, as a three-way matrix with enough power per cell to actually read the result. Winner rate roughly doubled. More importantly, the team could explain why a winner won — which is the only way the knowledge compounds.
05 / Approach
I work at the intersection of marketing, data, and business strategy. The deliverable isn't a campaign plan. It's a decision system your team runs after I'm gone.
Phase 01
Two weeks in the data. Account access, margin by SKU, the gap between what the platforms say and what finance actually books. End of week two I either have a thesis worth acting on — a specific constraint costing you money — or I tell you there isn't one and we both move on. That happens about one time in six. Fee is fixed either way.
Phase 02
Three to four weeks. Written target economics: POAS floor per channel, payback horizon, a six-month test calendar, what we're killing and why. Signed by CEO and CFO — not just the CMO. Most "growth strategies" are activity plans. This one is a rule: when X happens, we do Y. Finance signs it because it maps to the P&L they're accountable for.
Phase 03
Hands in the accounts, not adjacent to them. Bidding restructured, conversion API rebuilt with margin data, cohort LTV layer calibrated, weekly reporting mapped to board metrics. Always done with your team, never around them. I don't build systems your team can't operate — that's how good work quietly dies the quarter after I leave.
Phase 04
Playbook written, internal owner trained, quarterly check-in set. My goal is to make the retainer unnecessary. About half my clients keep me on advisory afterward. The other half don't need to. Both outcomes are fine. What I won't do is stay on to justify a retainer.
→ No retainers without a thesis I'd bet my own money on.
→ No engagements below €500k annual media. Nothing to optimize.
→ No agency kickbacks, referral fees, or rev-share with platforms.
→ Direct communication, fast iteration, decisions backed by data — not opinions.
06 / Contact
Best fit: DTC brands past early traction, teams where measurement and profitability numbers don't agree, and situations where spending harder has stopped improving the outcome. Email is best. Tell me what's broken and what you've already tried. Bring the P&L, or a rough approximation of it.
