How to Build a $100M Growth Engine: Lessons from Wispr Flow & Superhuman | Matt Swulinski

Why does it matter? Because waiting on paid ads may be the most expensive mistake SaaS founders make.

Matt Swulinski built the growth engine that put Wispr Flow on the map and drove Superhuman toward its Grammarly acquisition. His central claim is uncomfortable: the conventional wisdom telling SaaS founders to treat paid ads as a dangerous drug — reserved for later-stage scale — is exactly backwards. The companies winning distribution right now are running the ecommerce playbook, and the window to adopt it is narrowing every week.

• Paid ads are the fastest PMF validation tool available — messaging, funnel, and positioning tested in days, not quarters
• 90% of SaaS teams fail at paid because they skip measurement setup, not because the channel doesn't work
• Meta's Andromeda update made creative the targeting mechanism — 400–500 net-new creatives per month is the floor at $100K spend
• The talent required to run this playbook exists in less than 1% of current marketing candidates

Distribution is the only moat — and ecommerce figured this out a decade before SaaS did

Open X any day and five clones of your product launched overnight. Two copied your homepage. The only way to out-compete them isn't a better product — it's getting to customers faster. "Distribution to me is the only moat."

SaaS founders spent years watching ecommerce obsess over UGC programs, performance creative, and paid-first distribution — and dismissing it as a crutch for brands without organic advantages. Ecommerce ignored them and built an engine that compounded for a decade. The same founders are now competing in a market where distribution is the only variable that separates winners.

The ecommerce model treats every dollar as a purchase or add-to-cart proxy. Hundreds of UGC creators working across every creative dimension. A balance of channels that builds brand without depending on any single one. Swulinski applied this directly at Wispr Flow and again at Victor. "That's what put it on the map."

The resistance was real at Superhuman too — pure PLG, referral, word of mouth. When Fixer came along and ran the performance playbook, it scaled faster. The lesson cost time that couldn't be recovered.

Paid fails for most SaaS teams because they skip conversion tracking — then blame the channel

"I'd say 90% of companies don't do that as a first step. Before you spend your first cent, have everything set up."

The setup most SaaS teams skip: server-side conversion tracking, match rate maximized on Meta, enrichment score maxed on Google. These are not optimizations — they are the mechanism by which platform algorithms learn who your best customers are. Without them, Meta targets randomly. CAC spikes. The team concludes paid doesn't work.

Ecommerce solved this years ago with tools like Triple Whale and Elevar — out-of-the-box attribution in 15 minutes. SaaS has no equivalent. You need an analytics engineer building from scratch: database, BI layer, homegrown attribution. No shortcut exists.

The sequence: hire the analytics engineer before the media buyer. Confirm match rate and enrichment score before the first dollar of spend. Run paid without proper tracking and you're feeding garbage signals to the algorithm — every missed conversion is a customer profile it will never learn from. "If you have poor conversion tracking, Meta doesn't know who those people are. It just randomly targets people — and you'll have a super high CAC. You'll say paid doesn't work for me." Most of the time, they haven't done the setup correctly.

A $100K monthly Meta budget requires 400–500 new creatives — a volume most SaaS teams have never attempted

Four hundred to five hundred net-new creatives. Per month. That's the floor, not an aspiration.

Meta's Andromeda update pulled audience targeting controls away from media buyers. "The creative is the targeting." Meta analyzes video content and finds the people most likely to convert based on who those assets attract. Audience settings are no longer the lever. Creative diversity is.

Saturation arrives fast when the package lacks variety. Same hook, same demographic, same setting — the algorithm burns through the signal in days. "If you only focus on what works and you pump just that, performance will crater. You need to constantly be creating net-new, completely wild, weird ideas."

At Victor: hundreds of creators paid a percentage of ad spend, five agencies, and an in-house creative team. The company built its own tool to reformat vertical story ads into landscape YouTube templates. "We have kids that are like 17, 18, 19 that are making 20, 30k a month just making a couple ads for us."

The 80/20 rule applies — 20% of creators drive 80% of results — but the 20% is invisible until you have the volume to surface them. "The key thing around creative is not the one creative. It's the package of all of your assets together. They have to be different." This is a manufacturing problem, not a design problem.

The marketing talent you need exists in less than 1% of candidates — most teams should be rebuilt around that bar

"Probably fire most of your marketing team." Swulinski says it flat — and then explains exactly what he means.

The bar is systems thinking: the ability to step back from a task, map every input and output, identify the inter-relationships, and build a self-improving loop around the workflow. Not using ChatGPT to answer questions. A system that monitors, transforms, decides, and improves its own outputs over time without constant human input.

His interview signal: how do you use AI workflows in your personal life? The bad answer is a memory-enabled chat thread. The good answer describes a feedback loop — data comes in, something happens to it, the next iteration improves because of what happened this time.

At Wispr Flow, he was the only person executing against a $3–5 million monthly budget for a year and a half — managing a million dollars a month on Google, hundreds of thousands of keywords, 70–120 newsletter providers simultaneously. A Marketing OS built on Claude Code, connected to Obsidian, logged every session and decision into a compounding memory layer. Each day's work built on the last.

The bifurcation is accelerating. "The A players are becoming S-tier players and the people that were B players are becoming D players — and that rift is created." There's no stable middle. The systems thinker with moderate channel experience will out-compete the channel specialist who can't automate their own job. Less than 1% of candidates currently clear the bar.

Start paid the moment you hit 50 organic conversions — it validates product-market fit faster than any other method

Fifty organic conversions. That's the threshold — enough for Meta's algorithm to identify what your best customer looks like. At that point, start paid.

"You can refine messaging, do creative testing, test your funnels, test your positioning all within a week on paid." Content writing, user interviews, podcast appearances — these are slow feedback loops that produce equivocal data over months. Run organic and paid in parallel. Paid is just the faster validation engine.

The core three channels: Meta, Google, lifecycle. Email, SMS, push if you have an app. "You can scale to your first million, ten million ARR just off of those three things." TikTok and Reddit have their moment — it isn't now. Three channels executed well beats eight channels done poorly.

Treat the first $100K as research, not acquisition spend. Track cost per critical event — download, trial start, install. Chart spend against ARR and watch the elasticity. If spend rises and ARR follows, the engine works. If the curve flattens, something is breaking: landing experience, conversion tracking, creative diversity, or message clarity. Fix it as you spend. Start earlier than feels comfortable.

The right way to find your spending ceiling is to deliberately blow past it

At Wispr Flow, the team 5x'd monthly budget in a single month. Not incrementally. Deliberately, to find where it would break.

"Go as hard as you absolutely can and then see it blow up — and then understand where and how it blew up. Pull back and use that as information."

Gradual scaling takes months to surface what a stress test reveals in weeks. When growth targets are 30–40% month-over-month, six months of cautious budget increments is a strategic loss against a competitor moving faster. Blow it up fast, map what failed — creative saturation, audience fatigue, funnel breakdown, channel inter-dependency — then rebuild a repeatable ramp with that map.

Wispr Flow learned Meta needed to be secondary to Google. That relationship only became legible under stress. "We then knew, okay, meta needs to be second to Google. Here's why. Here's how all these things worked together. These newsletters were [expletive]. This podcast doesn't work."

The stress test runs in reverse too. At maturity, dial paid back to find the organic floor. If growth collapses entirely when spend drops, the other half of the acquisition engine — SEO, AEO, affiliate — hasn't been built. The target at scale is 35–45% of acquisition from genuine organic and word-of-mouth. Paid should amplify that base, not substitute for it.

Affiliate drives 10–15% of Victor's monthly acquisition — and most SaaS companies have never seriously tried it

Ten to fifteen percent of Victor's monthly acquisition comes through affiliate. The program pays 10–15% revenue share. One affiliate who closes a company at $10K per month earns $1,500 recurring every month that company stays. Some affiliates earn $20–30K per month.

"It is the highest ROI channel." The structural reason: affiliates self-select as genuine advocates. They already have audiences. They create YouTube reviews, Reddit threads, and comparison content without additional spend — and that content feeds both organic discovery and AEO citation loops. Compensation is pure performance, so CAC is structurally capped.

These are the same people whose long-form YouTube content AI answer engines cite most heavily. The affiliate program and the AEO strategy compound into each other without any deliberate coordination between them.

Launch early, before significant distribution exists. A 10–15% revenue share, easy to discover and join, is enough to attract serious earners. The affiliates who start earning while the product is still maturing become the longest-tenured advocates when it scales. "Even before they're your customers, they will become your customers."

Most SaaS companies have a referral program. Almost none have a serious affiliate program. The gap in ROI is the gap in that sentence.

Traditional PR doesn't drive traffic anymore — it exists to feed AI answer engines with citations

TechCrunch still matters. Not for the traffic spike — for the citation.

"PR is made for citations right around AEO." External credible sources writing about your product are a primary input that determines how prominently AI answer engines feature you when someone asks a relevant question. The traffic wave that used to justify a whole PR strategy has largely disappeared. What remains is citation value, and it compounds over time.

YouTube does this better than press. Long-form reviews rank for longtail queries and are disproportionately cited by ChatGPT and similar engines. "YouTube reviews — because they're long-form — they're one of the main things that get picked up by the answer engine. There are really high citations on ChatGPT."

The practical sequence: seed long-form YouTube reviews before investing in traditional press. A detailed ten-minute walkthrough from a creator with 5,000 subscribers generates more AEO citation value than a brief product mention in a major publication. The longtail ranking compounds. And every affiliate who earns real money becomes a YouTube reviewer who becomes a citation source. The flywheel connects back to the affiliate program — most founders haven't started spinning it.

The growth-engineer is underpriced today — and the market is about to correct

The talent gap Swulinski describes isn't just a hiring challenge — it's a signal about where the profession is heading. Marketing as a function is being compressed: entire teams replaced by one systems thinker with an agentic stack. These operators are currently underpriced because most companies haven't yet felt the difference between having one and not. As the first wave of AI-native growth teams demonstrates what compounding distribution actually looks like at scale, the market will reprice fast. Swulinski thinks this will look like the early ML researcher bidding wars — the same scarcity, the same sudden premium. The founders who find the 1% now won't be competing for them later.

---

Topics: growth, SaaS, paid advertising, product-led growth, UGC, creator programs, performance marketing, Meta ads, Google ads, affiliate marketing, AEO, answer engine optimization, AI-native teams, systems thinking, Superhuman, Whisper Flow, conversion tracking, customer acquisition