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.
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
171
48,654
Matt Swulinski retweeted
We scaled Stan to $40M ARR on 100% organic growth.
Earlier today, I spent an hour picking the brain of @MattSwulinski, Head of Growth at viktor.com (formerly marketing hire #1 at Wispr Flow managing tens of millions in spend).
I went in asking the question most over:
When are your retention and product metrics actually ready to turn on paid ads?
Here were the 4 core takeaways that shifted how I think about acquisition:
1. The first 60 days will look terrifying
Most founders test paid ads for 2 weeks, see an initial CAC of $500 to $1,000 against a $50 product, panic, and shut it off.
His take: That initial spike is normal while pixel signals are cold and creatives are uncalibrated.
It takes roughly 2 months of relentless iteration across creative, landers, and positioning to bring that down into a healthy range ($15-$25). If you aren't prepared to dedicate a real test budget for 60 days to let the machine learn, don't start.
2. Do not optimize for the purchase if volume is low
If you aren't clearing 50+ purchases a week on a specific ad set, ad platforms do not have enough data density to optimize delivery.
Instead, find your upstream activation proxy.
At Wispr, they discovered that dictating 200 words was a near-perfect statistical predictor of long-term paid conversion. They optimized ad delivery toward that 200-word milestone, and let the product onboarding handle the rest.
3. Signal engineering matters as much as creative
What you withhold from the algorithm is as important as what you feed it.
If users can sign up with low-intent accounts or empty sandboxes, those still fire conversion pixels. The ad network thinks it won and feeds you more low-quality traffic.
Filter out trash signals before the pixel reports back to Meta or Google so the algorithm only hunts for users who actually retain.
4. Organic ceilings are real
Even the best word-of-mouth engines hit an asymptote.
Companies like Superhuman scaled massively on founder referrals, but eventually had to layer on paid distribution, creator partnerships, and sponsors to break through the plateau.
Paid acquisition is not a replacement for a great product or organic distribution.
It is the operational engine you build once you have something people already love, to guarantee you control your own pipeline rather than hoping the algorithm smiles on you tomorrow.
Earlier today, I spent an hour picking the brain of @MattSwulinski, Head of Growth at viktor.com (formerly marketing hire #1 at Wispr Flow managing tens of millions in spend).
I went in asking the question most over:
When are your retention and product metrics actually ready to turn on paid ads?
Here were the 4 core takeaways that shifted how I think about acquisition:
1. The first 60 days will look terrifying
Most founders test paid ads for 2 weeks, see an initial CAC of $500 to $1,000 against a $50 product, panic, and shut it off.
His take: That initial spike is normal while pixel signals are cold and creatives are uncalibrated.
It takes roughly 2 months of relentless iteration across creative, landers, and positioning to bring that down into a healthy range ($15-$25). If you aren't prepared to dedicate a real test budget for 60 days to let the machine learn, don't start.
2. Do not optimize for the purchase if volume is low
If you aren't clearing 50+ purchases a week on a specific ad set, ad platforms do not have enough data density to optimize delivery.
Instead, find your upstream activation proxy.
At Wispr, they discovered that dictating 200 words was a near-perfect statistical predictor of long-term paid conversion. They optimized ad delivery toward that 200-word milestone, and let the product onboarding handle the rest.
3. Signal engineering matters as much as creative
What you withhold from the algorithm is as important as what you feed it.
If users can sign up with low-intent accounts or empty sandboxes, those still fire conversion pixels. The ad network thinks it won and feeds you more low-quality traffic.
Filter out trash signals before the pixel reports back to Meta or Google so the algorithm only hunts for users who actually retain.
4. Organic ceilings are real
Even the best word-of-mouth engines hit an asymptote.
Companies like Superhuman scaled massively on founder referrals, but eventually had to layer on paid distribution, creator partnerships, and sponsors to break through the plateau.
Paid acquisition is not a replacement for a great product or organic distribution.
It is the operational engine you build once you have something people already love, to guarantee you control your own pipeline rather than hoping the algorithm smiles on you tomorrow.
18
5
103
5,818
We run 140+ scheduled growth jobs on agents. A person still signs off on four calls.
If you give agents real work, you have to decide where a person still says yes. Here's where I draw that line running growth at @viktor_com.
1. Every dollar of spend needs a yes, because money's hard to pull back.
2. I sign partner agreements myself.
3. Anything a partner or customer sees gets a person's yes, since it carries our name.
4. I sign off on the messages and claims in our ads.
Viktor drafts and flags all of it, and a named owner makes the call. You can hold your own team up against this list. If an agent can change one of these without a named person saying yes, that's your gap.
Our security page says money moves and customer emails wait for explicit approval.
If you give agents real work, you have to decide where a person still says yes. Here's where I draw that line running growth at @viktor_com.
1. Every dollar of spend needs a yes, because money's hard to pull back.
2. I sign partner agreements myself.
3. Anything a partner or customer sees gets a person's yes, since it carries our name.
4. I sign off on the messages and claims in our ads.
Viktor drafts and flags all of it, and a named owner makes the call. You can hold your own team up against this list. If an agent can change one of these without a named person saying yes, that's your gap.
Our security page says money moves and customer emails wait for explicit approval.
5
0
21
1,269
When one of the four comes up, Viktor sends the owner a draft with cost and side effects. He doesn't act until that person says yes.
0
0
1
236
If your agency owns the tracking, you don't own the channel.
I've run an agency across ecom, SaaS and B2B. Agencies are great at execution. The learning compounds in how you count and when you act, so that part stays with you.
What sits in your own accounts:
1. The ad accounts, in your company's name.
2. The pixel and server-side conversion setup.
3. A table in your own data that counts paying customers.
4. The rules for when spend goes up or down.
Say the agency quits tomorrow. You pass if you can still pull last month's cost per paying customer yourself.
I've run an agency across ecom, SaaS and B2B. Agencies are great at execution. The learning compounds in how you count and when you act, so that part stays with you.
What sits in your own accounts:
1. The ad accounts, in your company's name.
2. The pixel and server-side conversion setup.
3. A table in your own data that counts paying customers.
4. The rules for when spend goes up or down.
Say the agency quits tomorrow. You pass if you can still pull last month's cost per paying customer yourself.
7
0
21
1,213
The ad platforms learn from your account history and your conversion data. If the accounts stay yours, a new agency doesn't reset any of it.
0
0
3
240
Last week we paused two Google campaigns that looked strong in Google's numbers.
We gave them a 30-day read in our own data. None of their self-serve signups paid.
A conversion in Google's report doesn't mean anyone paid. So budget only goes up once we count payers ourselves.
The rules we run paid on at @viktor_com:
1. A full read window, set before launch.
2. Zero paying customers through that window means paused, however the dashboard looks.
3. Budget goes up only after 2 weeks holding target, and in steps.
My own rule of thumb on creative: an ad at 2 to 3x target cost per result with nothing back gets cut.
Once it's written down, anyone on the team (or an agent) makes the same call.
Tonight, pull last month's campaigns from your own data. Any campaign with a full read window and zero payers fails.
We gave them a 30-day read in our own data. None of their self-serve signups paid.
A conversion in Google's report doesn't mean anyone paid. So budget only goes up once we count payers ourselves.
The rules we run paid on at @viktor_com:
1. A full read window, set before launch.
2. Zero paying customers through that window means paused, however the dashboard looks.
3. Budget goes up only after 2 weeks holding target, and in steps.
My own rule of thumb on creative: an ad at 2 to 3x target cost per result with nothing back gets cut.
Once it's written down, anyone on the team (or an agent) makes the same call.
Tonight, pull last month's campaigns from your own data. Any campaign with a full read window and zero payers fails.
9
0
25
1,833
We fix the read window before any money goes out. Once it's spent, it's way too easy to give a campaign one more week.
0
0
1
209
Write the kill rule before you write the prompt.
Agents are cheap to start. That's the problem. Nobody plans the ending.
Every scheduled job we run at @viktor_com gets four things:
One owner.
One question it answers.
A review date.
A kill date.
No owner, and nobody notices when it goes stale. No kill date, and it runs forever because switching it off feels risky.
How many of your agents could name their owner today?
Agents are cheap to start. That's the problem. Nobody plans the ending.
Every scheduled job we run at @viktor_com gets four things:
One owner.
One question it answers.
A review date.
A kill date.
No owner, and nobody notices when it goes stale. No kill date, and it runs forever because switching it off feels risky.
How many of your agents could name their owner today?
2
0
9
993
The kill date is the one people skip. It flips the default. At review, the job has to earn more time. Otherwise it survives because nobody wants to be the one who switches it off.
0
0
2
168
Silence is not success.
We replayed a September weekend when our tracking broke. To be clear, this was a replay, not a live catch.
In the replay, Viktor flagged it Friday evening.
In real life, humans found it Sunday.
He also raised one false alarm. I'll take that trade.
A quiet dashboard on a weekend means nothing broke, or nobody was looking. Do you know which one yours is?
We replayed a September weekend when our tracking broke. To be clear, this was a replay, not a live catch.
In the replay, Viktor flagged it Friday evening.
In real life, humans found it Sunday.
He also raised one false alarm. I'll take that trade.
A quiet dashboard on a weekend means nothing broke, or nobody was looking. Do you know which one yours is?
2
0
8
1,191
On scope: one replayed weekend is a test, not proof. The false alarm is the real cost of an always-on watcher. One noisy ping against two days of broken tracking nobody saw? I'd make that trade.
0
0
2
164
If people chat with your agent a lot, worry.
More messages can mean more hand-holding.
Someone explaining the task again.
Fixing the output.
Nudging him to finish.
At @viktor_com we don't count chats.
We count tasks he finished.
A report sent.
A lead updated.
A ticket closed.
So when someone tells me their agent is doing great, I ask one thing:
show me the work he finished last week.
A quiet agent with a long list of done tasks beats a chatty one every time.
More messages can mean more hand-holding.
Someone explaining the task again.
Fixing the output.
Nudging him to finish.
At @viktor_com we don't count chats.
We count tasks he finished.
A report sent.
A lead updated.
A ticket closed.
So when someone tells me their agent is doing great, I ask one thing:
show me the work he finished last week.
A quiet agent with a long list of done tasks beats a chatty one every time.
1
0
12
1,128
If you're rolling out an agent on your team, ask this before anything else:
What did he finish last week that nobody had to redo?
If nobody can answer with a list, you have a chatbot.
What did he finish last week that nobody had to redo?
If nobody can answer with a list, you have a chatbot.
0
0
1
197
Our cost per conversion got worse the same week we got smarter.
At @viktor_com we stopped telling our ads to chase signups.
We switched to the moment someone adds Viktor to Slack or Microsoft Teams.
On paper it looked awful.
Every conversion suddenly cost way more.
I still think it was the right call.
A signup is a click.
An install means someone put him to work.
The mistake was skipping one check first.
We never asked if the new event would hit about 50 conversions a week.
Below that, the ad platform can't learn.
So for a while we paid more and the algorithm was guessing.
We picked the right event and switched in the wrong order.
At @viktor_com we stopped telling our ads to chase signups.
We switched to the moment someone adds Viktor to Slack or Microsoft Teams.
On paper it looked awful.
Every conversion suddenly cost way more.
I still think it was the right call.
A signup is a click.
An install means someone put him to work.
The mistake was skipping one check first.
We never asked if the new event would hit about 50 conversions a week.
Below that, the ad platform can't learn.
So for a while we paid more and the algorithm was guessing.
We picked the right event and switched in the wrong order.
12
0
35
2,297
Before you switch, do one boring check.
Pull last month's count for the new event.
Divide by four.
If it's under about 50 a week, go one step shallower.
Pick the deepest event that still clears 50 and go deeper once volume catches up.
Pull last month's count for the new event.
Divide by four.
If it's under about 50 a week, go one step shallower.
Pick the deepest event that still clears 50 and go deeper once volume catches up.
1
0
2
243
On Tuesday I walked product people at The_Product through how I pick channels.
It's also how I run growth at @viktor_com
Your channels should fall out of four answers about your own product. And the magic moment answer decides whether paid makes sense at all.
1. Who has the problem? Go where they already spend time.
2. What's LTV, so what can CAC be? Push as much as you can into self-serve and wire up the data so every signup sends a predicted value signal you can optimize on. Demos only earn their place when you can't measure that value without a human.
3. How long to the magic moment? Minutes opens creators, paid social and self-serve. Hours means search and content, then email closes. Days or weeks means onboarding first. Paid only buys more churn.
4. What makes buyers trust you? B2B trusts similar companies. Developers trust working code.
Each answer needs a consequence. No consequence, no answer.
Write your four answers, then check the channels you run now. If they don't follow, you're running someone else's plan.
It's also how I run growth at @viktor_com
Your channels should fall out of four answers about your own product. And the magic moment answer decides whether paid makes sense at all.
1. Who has the problem? Go where they already spend time.
2. What's LTV, so what can CAC be? Push as much as you can into self-serve and wire up the data so every signup sends a predicted value signal you can optimize on. Demos only earn their place when you can't measure that value without a human.
3. How long to the magic moment? Minutes opens creators, paid social and self-serve. Hours means search and content, then email closes. Days or weeks means onboarding first. Paid only buys more churn.
4. What makes buyers trust you? B2B trusts similar companies. Developers trust working code.
Each answer needs a consequence. No consequence, no answer.
Write your four answers, then check the channels you run now. If they don't follow, you're running someone else's plan.
3
1
20
1,149
If I could only keep one, it's question 3.
Time to the magic moment tells you if paid can work, before you spend a cent on ads.
Same four questions I used at Wispr Flow. Now I run them while scaling Viktor.
Time to the magic moment tells you if paid can work, before you spend a cent on ads.
Same four questions I used at Wispr Flow. Now I run them while scaling Viktor.
1
0
1
198
I don't think the newsletter is dead. I think it got a gatekeeper.
Once an agent decides what reaches someone, your product has to be callable by that agent, not just emailable.
Most growth playbooks still assume a human opens the email.
Once an agent decides what reaches someone, your product has to be callable by that agent, not just emailable.
Most growth playbooks still assume a human opens the email.
a big take I've been noodling on and I’m ready to call:
the newsletter is dead
@every's core business is a newsletter with 120k+ subscribers so I don't say this lightly.
persistent agents mean i dont look at my inbox
with dots (GrokBots, Muse, etc) I almost never look at my inbox. if I do it's for a specific thing I need to reply to.
i think in a year I'll be doing all of my email this way - fully filtered by my dot / something like it.
this is not new: it's a continuation of the trend of having a “newsletter” folder in Gmail, but 10x more filtered.
what it means
It doesn’t mean people won't read newsletters, but it does mean that getting a person’s email is no longer the guarantee that you have a relationship with them that it used to.
luckily, with persistent agents (and AI coding more generally) there is an entirely new playing field to build relationships with readers. instead of fighting to get an email address, publishers and creators will need to make sure we're installed in our reader's agent harness of choice, so that they get directly notified when we do something new.
we're already starting to experiment with this @every with our MCP and the frontier alerts we send users of our slack agent (every.to/agent)
it's also more affordable than ever for publishers and creators to build their own platform ecosystem: you can have your own website, app etc. and you can use AI to publish 100x more high-quality stuff to your community—so we may see a world where many creators themselves are running their own little empires that have their own apps, feeds etc.
to be clear: It'll take a while before this actually happens, but the idea of getting someone's email creating a direct relationship with them is over.
here's a set of predictions I made 3 years ago about AI and media that panned out pretty well: every.to/chain-of-thoug…
3
1
17
1,445
We built for this at viktor.com. Viktor has an MCP server, so the agent you already use can hand him tasks in your existing workspace.
You need a paid plan and an API key you create in Settings first.
Docs: viktor.com/docs/mcp-server
You need a paid plan and an API key you create in Settings first.
Docs: viktor.com/docs/mcp-server
2
0
1
189