I took one revenue line from $2M to $250M+
At PingPong I owned the cross-border payments line end to end: $76M in year one, past $250M+ in year two, on funnels I built and ran across North America, Europe, and Southeast Asia.
Senior Marketing Manager at ThinkingAI since 2026. Four years at PingPong before that, a 1,500-person company: Marketing Manager, Ads & Lifecycle, 2021–2023, then Senior Digital Marketing Manager, Growth & Operations, 2023–2025. And six years at ToLocal, a company of twenty-odd, the last of them as VP of Marketing there: seven people in the US and five in China, on a $10M annual media budget.
A revenue line and an 8-tool stack at PingPong, a team of 7 at ToLocal
When PingPong handed me the cross-border payments line, the number was revenue, and I treated every channel, workflow and tool as an input to it. Ten years across B2B fintech and SaaS, and I can still defend that number in a finance review line by line.
I chose an 8-tool GTM stack, Salesforce, HubSpot, Webflow, 6sense, Zapier, ZoomInfo, Salesloft and GA4 with Tag Manager, and governed it myself so no report depended on a vendor's version of the truth. I ran the funnels across North America, Europe, and Southeast Asia as one system with regional variants. And when the marketing org needed cover, I took over all of Marketing for 6+ months with no interruption to the business.
As VP of Marketing at ToLocal I ran a $10M annual media budget at 80%+ ROI while leading a team of 7. Since late 2025 I have shipped nine AI GTM tools solo, design through deploy, including CreativeOS for policy-aware ad copy and a fintech marketing compliance checker.
Three kinds of evidence, and what each one settles
Sorted by the kind of doubt each one answers. The case studies show the number and the conditions it happened under. The systems show what I can build when the stack will not do it. The writing shows how I think when nobody is grading me.
Shipped work
Revenue engineThe number itself: $2M to $76M in year one, past $250M in year two, on funnels I built and ran. Acquisition efficiencyCPL nearly halved on two channels with volume held. Proof the growth was bought efficiently rather than bought. Global web and trackingOne funnel across North America, Europe and Southeast Asia, with attribution that survived a finance review. Content enginesOrganic demand produced on a cadence a two-person team could actually hold.Systems I built for it
CreativeOSA later demo of the same idea: ad copy written inside platform and legal policy, so the creative queue stops being the launch bottleneck. Built independently, no customers. Competitor RadarA competitor's whole ad library read as one strategy, which is where a counter-position comes from. TechSpyAccount qualification from a live technology read, so a target list is built on evidence rather than firmographics.How I think, written down
Landing page testing without overthinking itThe testing cadence behind move 03, including the tests I decided were not worth running. Building an ABM list from technology signalsHow the enterprise target list gets built, and the point where I threw out a 2,671-company universe. Fintech keyword research: intent, trust, riskAcquisition in a category where the compliance constraint shapes the keyword set. A referral-led growth systemA growth loop designed from nothing for a new feature, with the trust limits written in. Running social campaigns with limited resourcesWhat the same method looks like with no budget, which is the version most teams need.Four moves I repeat
on every funnel I own
01 · Instrument the number
Before I scale anything I make the number trustworthy. At PingPong that meant freezing spend increases for a quarter and rebuilding tracking myself until cost per qualified lead was a figure finance could audit line by line. It is an unpopular first move, because it looks like nothing is happening. The alternative is scaling on a number nobody will defend in the review that follows, and then losing the budget anyway.
A number finance will not defend is not a number02 · Find the constraint
Every quarter the number is capped by one thing. Once it was acquisition cost, with LinkedIn leads at $300+ each. Later it was a twelve-hour lead handoff bleeding intent while reps slept. I pick the one constraint, size what removing it is worth, and put the rest of the roadmap down. The discipline is in what gets ignored: a quarter spent improving three things at once usually moves none of them far enough to notice.
One constraint a quarter; the rest of the roadmap waits03 · Buy efficiency with tests
I test campaign structure, audience, keyword and landing path on a fixed cadence, with the success metric written down before launch so the result cannot be reinterpreted afterwards. That run cut LinkedIn CPL to ~$160 and Google registration cost from $90+ to ~$35, with volume held on both. Holding volume is the part that matters. Efficiency bought by spending less is just arithmetic; the same leads at half the price is the result.
Efficiency with volume lost is just spending less04 · Automate the winners
Whatever works, I wire into infrastructure so it keeps working without me. The 200+ workflows I built took over the routing, enrichment and alerting people used to do by hand, which is how lead handoff went from ~12 hours to 1–2 minutes. I automate only what has already proven itself, because automating an unproven play makes a bad decision repeat faster. If the machine needs me daily, I have not finished building it.
Automate only what has already earned itThe inputs behind $2M → $76M → $250M+
Same pipeline, roughly half the price, then no waiting
I rebuilt the acquisition layer from the tracking up and tested my way down the cost curve: LinkedIn cost per lead fell from $300+ to ~$160 and Google registration cost fell from $90+ to ~$35, with volume held on both channels. That efficiency is what let the line I owned scale from $2M to $76M in year one without the budget scaling with it.
Then I removed the human bottleneck. The 200+ workflows I built collapsed lead handoff from ~12 hours to 1–2 minutes, so the demand I paid for stopped going cold in an inbox. Sales owned the close; my part was getting a warm lead into their hands in minutes instead of the next morning. I measured the machine, found its slowest part, and replaced it.
CPL nearly halved · handoff ~12h → 1–2 min · line at $250M+ by year twoHiring for growth?
Give me the number. I will build the machine that hits it and show you the math.