Head of Growth @viktor_com Previously at @wisprflow & @superhuman
Warsaw, Poland
- Tweets 299
- Following 288
- Followers 2,773
- Likes 352
Pinned Tweet
Most guests tell Harry to wait on paid. I told him the opposite.
Start day one. Paid is the fastest test of whether your PLG actually works.
Also in here: why $100K on Meta needs 400 creatives a month, plus the teenagers earning $20-30K making ours.
I have interviewed 100 of the best growth leaders in the world.
@MattSwulinski is easily top 3. (alongside @alexschultz and Brian Hale)
He scaled Wispr Flow to over $100M in ARR and built a UGC machine.
He scaled Superhuman from founder personally onboarding every customer to a growth machine with $50M ARR.
If you are an early stage founder or growth leader, this will be the best episode you will listen to this year!
I condensed my biggest lessons from the discussion below:
1. The E-Commerce Playbook Is the Right Playbook for SaaS
The e-commerce playbook, where every dollar spent ties directly to a purchase or conversion, is the right model for modern SaaS. With distribution becoming a critical moat in a crowded AI market, SaaS companies should deploy UGC creators, constantly test new creative, and diversify channels to build their brand.
2. Paid Acquisition Is the Fastest Way to Validate PLG
Relying solely on organic content and word-of-mouth takes too long to validate product-market fit. Paid acquisition creates the fastest feedback loop for proving a PLG funnel works, allowing teams to test positioning, refine messaging, and optimize conversion within a single week.
3. You Only Need Three Core Channels to Scale to $10M ARR
Startups often ruin their acquisition engines by trying to run ten channels poorly at once. Reaching the first $10M in ARR only requires mastering three core channels: video intent on Meta, search intent on Google, and lifecycle retention through email and SMS.
4. Scaling Paid Ads Requires 500 New Creatives Every Single Month
On platforms like Meta, creative increasingly acts as the targeting algorithm. Scaling spend without hitting audience fatigue requires 400 to 500 new creative assets every month, produced through UGC revenue-share programs, specialized agencies, and internal teams.
5. How the Best Growth Leaders Test for True Spend Incrementally
Blindly increasing ad spend wastes money on conversions that may have happened organically. The best growth leaders measure spend elasticity against ARR growth and run strict holdout tests to determine whether each additional dollar generates genuinely incremental revenue.
6. In Three Years, Companies Will Operate Like a Board of Directors
Tech organizations are shifting away from manual execution. Within three years, lean human teams could operate more like boards of directors, spending 20% of their time on strategy while autonomous AI agents handle 80% of operational execution.
7. Fire Your Marketing Team if They Aren’t Systems Thinkers
Marketers focused on repetitive manual tasks are becoming increasingly replaceable. High-performing teams need systems thinkers who can break their work into inputs and outputs, then build self-improving AI workflows that multiply their personal leverage by 10x.
(links in comments)
14
1
158
44,188
Every model launch is a COGS event now.
Opus 5.5 went into Viktor today. We ran it on the real tasks people hand him in Slack + Teams. It passes more of them than Opus 5 and it costs less per task.
A better model used to mean a better product. Now it also means better margins at the same price. If AI tokens sit in your COGS, cost per task is the line you watch.
@claudeai Opus 5.5 is now available in Viktor.
We ran it through ViktorBench, real tasks people hand Viktor in Slack and Teams. It scores higher than Opus 5 on usefulness. It costs less per task. The best value point we have measured for an AI employee that works all day.
1
1
17
1,198
Matt Swulinski retweeted
Fun fact: @wisprflow was our first-ever customer for Dub Partners back when it was still in private beta.
@tankots personally set up conversion tracking with Dub and gave us a ton of valuable feedback.
Today, they've scaled to millions of users with @dubdotco and have sent over $300K in payouts.
Forever grateful to Tanay for the early support + @MattSwulinski for making the intro 💪
With Dub, Wispr Flow efficiently scaled their AI voice dictation app to millions of new users, at a ~4x average ROAS.
"We probably wouldn't have grown as fast without Dub." ~ Tanay Kothari, co-founder & CEO of Wispr Flow
Learn how they did it: d.to/wisprflow
3
7
42
6,116
The TV by my desk says $52M. The box under it says 0.0018%.
$52M is @viktor_com's annualized revenue, 7 months after launch. 0.0018% is the share of the world's 1 billion knowledge workers who pay for him. 17,698 people. I spent 3 weeks in SF and New York meeting some of them. I came back to Warsaw today to a team that shipped the whole time I was gone.
The big number got its announcement last week. The small one is what I'm back at my desk for.
If you have a Viktor on your team, what was the first job you gave him?
11
2
55
4,621
Matt Swulinski retweeted
Introducing the @dubdotco Program Marketplace → dub.co/marketplace
Discover the best affiliate programs in AI, Fitness, Marketing, Fintech, Design, DevTools, and more.
Featuring world-class companies like Beehiiv, Wispr Flow, Framer, Granola, Polymarket, Viktor and more.
25
11
128
19,341
I was at @lennysan's Summit all day. Half the people I'd hire were in that room, I talked to maybe 20 of them.
So, here: Viktor is hiring PMs + PMMs. I also want one senior product person to own the vision. 7,000+ paying companies and a vision to conquer all knowledge work.
DM me or drop a name.
3
1
23
1,697
60,000+ teams have a Viktor on their team. As of today you can talk to him, and the work lands while you speak.
What are you going to ask him to do first?
ref.viktor.com/infinity-matt…
Meet Viktor Infinity ∞
Codename: Jarvis.
Talk to your AI employee live, powered by @OpenAI's GPT-Live, and the work lands on a canvas as you speak. Ask for the recap, read it, say yes, it's sent.
Voice AI finally got a job.
Live today: ref.viktor.com/infinity-x
2
25
4,331
No talks, one room, open bar. Sept 8 in NYC we're hosting a cocktail hour for people who run GTM at growing companies.
Founders, heads of growth, PMs, AEs. If you're at the bleeding edge of your field I'd love to meet you.
Register here: luma.com/gtmcocktails
5
26
1,363
Almost nobody can tell you which channel is incremental and which one to cut.
Most teams answer this by hiring a Meta expert or a Google expert. Wrong bet. You never turn those channels off anyway.
The hire is the orchestrator. One person who sits on total spend vs the growth rate that spend actually buys. Meta / Google / CTV and so on at the top. Lifecycle as the net underneath. Everyone needs someone who knows how the channels feed each other.
Get that person first. Then add channel experts and build the toggles inside each channel.
You’ll thank me later.
11
2
42
2,330
I thought I loved my home office.
I ran remote for 6 years and the last few weeks being in person at Viktor's Warsaw HQ have definitely been an adjustment. There is something truly magical about building a company in person and actually interacting with the office. I forgot how much I loved it + didn't realize what I was missing for so long!
Next week we're off-site in San Francisco. If you think there are people / companies that we need to meet, shoot me a reply or send me a DM.
12
47
2,697
Most teams still use AI like a chatbot.
Question in. Answer out. Tab closed.
The ones pulling ahead treat it like an employee. Real work goes through it. Hard feedback comes back. That muscle gets stretched every week until the agent stops sounding like a search box.
Stop prompting. Start managing.
6
22
1,403
Matt Swulinski retweeted
I first discovered @MattSwulinski through his podcast with @HarryStebbings.
Today, I watched my first YouTube livestream at @knocklabs because he would join a conversation with @scottjstrand.
Matt is indeed a world-class growth master:
At @WisprFlow, Matt owned nearly every KPI from impression to app download while monthly ad spend reached $4M to $5M.
He built an AI marketing OS to handle that workload, then kept rebuilding it:
1️⃣ Automate repeatable decisions. The system analyzed creative and moved ads from testing to scale.
2️⃣ Rebuild for the team. When Wispr hired more marketers, Matt realized the first version only worked for him. He spent 36 hours turning it into a Git-controlled multiplayer system.
3️⃣ Force fresh data. An investor deck pulled three-week-old data five minutes before a meeting. Matt added timestamps and required fresh queries from Hex or the database.
4️⃣ Design for adoption. About half of the marketing team never adopted the Git, folder, and pull-request workflow. That limitation pushed him toward AI systems the whole team could use.
Each failure gave Matt a design rule for the next version.
That is the discipline I want to bring into my own work: treat every failed workflow as input for a better system.
It’s a real treat to be interviewing @MattSwulinski next week.
He built one of the most advanced agentic growth systems in SaaS as the first marketing hire at @WisprFlow.
Now he’s doing the same thing at @viktor_com.
RSVP here: luma.com/xphob5wa
ALT Register for the session, "Humans + agents: The next era of growth teams" with Matt Swulinski, Head of Growth at Viktor.com.
4
1
17
1,145
Matt Swulinski retweeted
Harry Stebbings said "@MattSwulinski is easily a top 3 growth leader in the world," putting him next to Meta CMO, Alex Schultz.
Proud to say we've partnered with Matt and he's now offering consulting on @MentorPass.
First 5 sessions are 50% off 👇🏼
I have interviewed 100 of the best growth leaders in the world.
@MattSwulinski is easily top 3. (alongside @alexschultz and Brian Hale)
He scaled Wispr Flow to over $100M in ARR and built a UGC machine.
He scaled Superhuman from founder personally onboarding every customer to a growth machine with $50M ARR.
If you are an early stage founder or growth leader, this will be the best episode you will listen to this year!
I condensed my biggest lessons from the discussion below:
1. The E-Commerce Playbook Is the Right Playbook for SaaS
The e-commerce playbook, where every dollar spent ties directly to a purchase or conversion, is the right model for modern SaaS. With distribution becoming a critical moat in a crowded AI market, SaaS companies should deploy UGC creators, constantly test new creative, and diversify channels to build their brand.
2. Paid Acquisition Is the Fastest Way to Validate PLG
Relying solely on organic content and word-of-mouth takes too long to validate product-market fit. Paid acquisition creates the fastest feedback loop for proving a PLG funnel works, allowing teams to test positioning, refine messaging, and optimize conversion within a single week.
3. You Only Need Three Core Channels to Scale to $10M ARR
Startups often ruin their acquisition engines by trying to run ten channels poorly at once. Reaching the first $10M in ARR only requires mastering three core channels: video intent on Meta, search intent on Google, and lifecycle retention through email and SMS.
4. Scaling Paid Ads Requires 500 New Creatives Every Single Month
On platforms like Meta, creative increasingly acts as the targeting algorithm. Scaling spend without hitting audience fatigue requires 400 to 500 new creative assets every month, produced through UGC revenue-share programs, specialized agencies, and internal teams.
5. How the Best Growth Leaders Test for True Spend Incrementally
Blindly increasing ad spend wastes money on conversions that may have happened organically. The best growth leaders measure spend elasticity against ARR growth and run strict holdout tests to determine whether each additional dollar generates genuinely incremental revenue.
6. In Three Years, Companies Will Operate Like a Board of Directors
Tech organizations are shifting away from manual execution. Within three years, lean human teams could operate more like boards of directors, spending 20% of their time on strategy while autonomous AI agents handle 80% of operational execution.
7. Fire Your Marketing Team if They Aren’t Systems Thinkers
Marketers focused on repetitive manual tasks are becoming increasingly replaceable. High-performing teams need systems thinkers who can break their work into inputs and outputs, then build self-improving AI workflows that multiply their personal leverage by 10x.
(links in comments)
3
3
18
5,700
30% of our paid budget never runs on our own ads.
You can also run partnership ads, and someone posts in collaboration with you, because the same ad from a creator handle beats it from the brand handle.
That's why that spend sits on their handle, not the brand handle.
How do you run the creator program? What's UGC you put spend behind on your own ads, what gets posted to their channels, and what do you actually run as partnerships?
6
29
2,297
I thought I was leaving the best startup scene in the world.
I spent my whole life in the US. Now the whole company is in the Warsaw HQ and things are cooking. We're building one of the fastest-growing companies in the world from this room.
It's a pleasure to be on this ride.
This is just the start.
22
3
97
7,520
YouTube taught people and they converted on Search weeks later.
Last click says Search did it and YouTube looks like it did nothing. If you pull YouTube, search dies, because the channels work together.
There's a lag where I spend today and they convert weeks later. Last click, first click, all of that is not enough. You have to look at multi-touch on a pretty hard funnel like Viktor.
7
1
38
6,698
Meta and Google are not optimized for B2B SaaS usage-based products.
They go nuts when you tell them one person could be a $50,000 annual contract and another is $50. Huge variance. A single person can be a high user and the algorithm has no idea what to do with that.
And each platform's algorithm does what you tell it. So let's say we have poor conversion tracking that doesn't track when someone subscribes and it's like a 50/50 match. If I optimize for that subscription, Meta will give me more of those people. Bad tracking and Meta doesn't know who those people are.
The main thing you optimize for is the most important event to you as a business. So SaaS, it's the subscription or the start trial.
At enrichment we put a predicted cash payback on the creative. We have been working with Meta on long journey optimization so that number can train. The bar for us is predicted payback under six months.
12
4
39
3,752
The model providers are handing out nuts comped plans. $200 on paper and thousands of dollars of tokens if you burn the limits.
I've seen companies tell employees to use the enterprise plan until they hit a credit limit, then switch to a Max plan to keep the usage benefits. So much subsidization and gaming.
We tested Anthropic's $200/mo Claude plan, which can yield up to $8,000/mo💰️💰️, while OpenAI's $200/month plan could yield up to $14,000/month worth of tokens if you exhausted all the weekly limits on long-horizon tasks. This test was done back in June.
3
15
2,829
A 3:1 LTV to CAC on revenue can be 1:1 on margin.
If AI tokens sit in your COGS, those are different companies. Compute LTV on gross profit, not revenue.
AI put real COGS back into software. Most SaaS models still pretend it did not.
6
2
54
5,087
We used to pay referrals on signup.
Both sides got credits the moment someone created an account. Volume looked fine. Paying customers did not.
We moved the trigger. Credits only land when the referral becomes a paying Viktor customer. The reward is a rev-share cut in credits on what they bring in.
Paying customers from referrals went 3x. Volume followed the quality.
The most valuable thing you can give a power user is not swag. It is a cut of the value they create.
9
4
85
8,321
We seeded Superhuman's affiliate program from one list: our top referrers.
The bet was simple. People who send you users for free become your best paid partners.
That list is how I met @nicklafferty.
His Superhuman review still sits at the top of search. Six years of updates. Superhuman has never made a page that outranks it.
None of it works without the design under it. Pay in the thing power users run out of. For a mail product, that was free months. Fire the ask at the moment of felt value, not at signup. Keep a leaderboard, because the leaderboard is the real asset.
A referral program does not just make signups. It makes a list of people who love your product enough to sell it. The leadership move is to go meet every name on it.
Years later I joined Nick on the Marketing Engineer Podcast, thanks to him and @profound. Same guy from the top of the leaderboard.
Wild how things come full circle.
2
1
14
1,470