Sep 16, 2026
By Victor Teran
Your Paid Acquisition Is Subsidising a Leaky Product
At median activation, 62 of every 100 people you buy never reach the thing your product is for. Spending more does not fix that. It buys the same leak at a larger scale, and it hides the leak behind a growth number that keeps everyone comfortable.
Retention
Paid
Activation
Paid acquisition is the most reversible growth lever a founder has, which is exactly why it gets pulled first. Turn it up, the numbers move this week, and nobody has to admit anything about the product.
The trouble is arithmetic. Spend flows into the top of a funnel whose conversion rate you did not change. A fixed proportion of every additional pound therefore goes to people who were never going to arrive. At median activation that proportion is most of it.
Key takeaways
Median B2B SaaS activation is 38%. At that rate, 62% of paid spend buys users who never reach the value moment.
Fixing activation is a multiplier on every future pound of ad spend. Increasing spend is not a multiplier on anything.
Q.AI went from 38% to 67% activation with no new features. At constant spend that is roughly 1.76x the activated users for the same money.
Paid works, and it works far better second. The order is the whole argument, not the channel.
What does a leak cost at real numbers?
At 38% activation, more than half of every acquisition pound buys someone who leaves.
Take a founder spending $20,000 a month at a blended $40 per signup. That is 500 signups. At the B2B SaaS median activation rate of 38% (Perspective AI, May 2026), 190 of them reach the value moment and 310 do not.
The effective cost per activated user is not $40. It is $105.
Now hold spend flat and move activation to 67%, which is what happened at Q.AI. The same 500 signups produce 335 activated users at an effective $60 each. Same budget, same traffic, 145 more activated users, and a cost per activated user that fell by 43%.
At $20,000/month, 500 signups | 38% activation | 67% activation |
|---|---|---|
Activated users | 190 | 335 |
Effective cost per activated user | $105 | $60 |
Wasted spend on users who never activate | $12,400 | $6,600 |
Nothing in that table required a bigger budget, a better agency, or a new channel.
The reason cost per activated user is the honest denominator is that activation marks the point a new user first experiences the product's core value (Amplitude). Everything before it is a cost, not a customer.
Why does spending more feel like it is working?
Because the top-line number moves immediately and the damage arrives on a delay.
Signups respond to spend within days. Retention responds within months. So for one or two quarters, more spend genuinely looks like more growth, and the leak scales invisibly underneath it.
By the time the retention curve makes the problem undeniable, the spend is embedded in the plan, the team is built around the volume, and cutting it looks like going backwards. That is a considerably harder conversation than the one available at the start.
This is also why paid gets blamed unfairly. The channel usually worked. It delivered the traffic it promised. What it could not do was fix what happened after the click, and nothing in an ads dashboard makes that distinction visible.
FREE GUIDE
The 4 product leaks costing you growth
A short audit guide for founders. Find the four places your product leaks revenue, and what to fix first.
How do you tell whether the problem is traffic or product?
Compare activation rates by source. If they are all bad, it is not the traffic.
This is the diagnostic that settles most arguments between founders and whoever is running their ads, and it takes an afternoon.
Segment new users by acquisition source and look at activation for each cohort separately:
What you see | What it means |
|---|---|
Every source activates at a similar low rate | The product is the constraint. Fix activation before scaling spend |
One source is far worse than the rest | Targeting or message mismatch on that source. Fix or cut it |
Paid far below organic | You are buying people with less intent. Either the promise is off, or the audience is |
All sources healthy, retention still poor | Activation is fine. The problem is habit formation, not acquisition |
a16z's growth metrics guide makes the general version of this point, segmenting benchmarks by revenue scale, software type and go-to-market motion rather than publishing one number (a16z). A blended figure across sources hides exactly the difference you need to see.
Only the second and third rows are advertising problems. The first and fourth are product problems that no amount of budget or creative testing will solve, and they are the more common findings.
Does this mean don't run paid?
No. It means run it second, once each new user is worth what they should be.
Paid acquisition is genuinely the fastest way to buy learning and volume, and waiting for organic to compound is a luxury most funded companies cannot afford. The argument here is about order, not about the channel.
The sequence that makes the money work:
1. Define the value moment, so activation means something. Most products we audit are measuring signup completion and calling it activation, and every number downstream inherits that error. 2. Find where new users stop, with a funnel and session replays, before changing anything. 3. Fix the earliest expensive step. In our engagements the biggest lever sits in the first minutes of the experience nearly every time. 4. Then scale spend, against a funnel where each acquired user is worth substantially more.
Doing step four first is not wrong because paid is bad. It is wrong because you are buying at the worst exchange rate you will ever have, and you are locking in a cost per activated user that the product could have halved.
What is the number to watch?
Cost per activated user, not cost per signup.
Cost per signup is the number your ads platform shows you, and it is the number that improves when you buy cheaper, less interested traffic. It can fall while your business gets worse.
Cost per activated user is spend divided by people who reached the value moment. It moves when the product improves, when targeting improves, and when the promise in your ad matches what the first session delivers. That is the number worth putting on the wall.
It is also the number that connects paid spend to the rest of the funnel, which is why a redesign that ships without moving it has not earned anything.
If you are running paid today and cannot produce that figure, that gap is the finding. The root cause is usually the same one as everything else in this post. The value moment was never defined, so nothing downstream of it can be measured. That is an afternoon of work, and it will change what you do with next month's budget.
WHAT NEXT
Want this fixed in your product, not just explained?
Sep 16, 2026
By Victor Teran
Your Paid Acquisition Is Subsidising a Leaky Product
At median activation, 62 of every 100 people you buy never reach the thing your product is for. Spending more does not fix that. It buys the same leak at a larger scale, and it hides the leak behind a growth number that keeps everyone comfortable.
Retention
Paid
Activation
Paid acquisition is the most reversible growth lever a founder has, which is exactly why it gets pulled first. Turn it up, the numbers move this week, and nobody has to admit anything about the product.
The trouble is arithmetic. Spend flows into the top of a funnel whose conversion rate you did not change. A fixed proportion of every additional pound therefore goes to people who were never going to arrive. At median activation that proportion is most of it.
Key takeaways
Median B2B SaaS activation is 38%. At that rate, 62% of paid spend buys users who never reach the value moment.
Fixing activation is a multiplier on every future pound of ad spend. Increasing spend is not a multiplier on anything.
Q.AI went from 38% to 67% activation with no new features. At constant spend that is roughly 1.76x the activated users for the same money.
Paid works, and it works far better second. The order is the whole argument, not the channel.
What does a leak cost at real numbers?
At 38% activation, more than half of every acquisition pound buys someone who leaves.
Take a founder spending $20,000 a month at a blended $40 per signup. That is 500 signups. At the B2B SaaS median activation rate of 38% (Perspective AI, May 2026), 190 of them reach the value moment and 310 do not.
The effective cost per activated user is not $40. It is $105.
Now hold spend flat and move activation to 67%, which is what happened at Q.AI. The same 500 signups produce 335 activated users at an effective $60 each. Same budget, same traffic, 145 more activated users, and a cost per activated user that fell by 43%.
At $20,000/month, 500 signups | 38% activation | 67% activation |
|---|---|---|
Activated users | 190 | 335 |
Effective cost per activated user | $105 | $60 |
Wasted spend on users who never activate | $12,400 | $6,600 |
Nothing in that table required a bigger budget, a better agency, or a new channel.
The reason cost per activated user is the honest denominator is that activation marks the point a new user first experiences the product's core value (Amplitude). Everything before it is a cost, not a customer.
Why does spending more feel like it is working?
Because the top-line number moves immediately and the damage arrives on a delay.
Signups respond to spend within days. Retention responds within months. So for one or two quarters, more spend genuinely looks like more growth, and the leak scales invisibly underneath it.
By the time the retention curve makes the problem undeniable, the spend is embedded in the plan, the team is built around the volume, and cutting it looks like going backwards. That is a considerably harder conversation than the one available at the start.
This is also why paid gets blamed unfairly. The channel usually worked. It delivered the traffic it promised. What it could not do was fix what happened after the click, and nothing in an ads dashboard makes that distinction visible.
FREE GUIDE
The 4 product leaks costing you growth
A short audit guide for founders. Find the four places your product leaks revenue, and what to fix first.
How do you tell whether the problem is traffic or product?
Compare activation rates by source. If they are all bad, it is not the traffic.
This is the diagnostic that settles most arguments between founders and whoever is running their ads, and it takes an afternoon.
Segment new users by acquisition source and look at activation for each cohort separately:
What you see | What it means |
|---|---|
Every source activates at a similar low rate | The product is the constraint. Fix activation before scaling spend |
One source is far worse than the rest | Targeting or message mismatch on that source. Fix or cut it |
Paid far below organic | You are buying people with less intent. Either the promise is off, or the audience is |
All sources healthy, retention still poor | Activation is fine. The problem is habit formation, not acquisition |
a16z's growth metrics guide makes the general version of this point, segmenting benchmarks by revenue scale, software type and go-to-market motion rather than publishing one number (a16z). A blended figure across sources hides exactly the difference you need to see.
Only the second and third rows are advertising problems. The first and fourth are product problems that no amount of budget or creative testing will solve, and they are the more common findings.
Does this mean don't run paid?
No. It means run it second, once each new user is worth what they should be.
Paid acquisition is genuinely the fastest way to buy learning and volume, and waiting for organic to compound is a luxury most funded companies cannot afford. The argument here is about order, not about the channel.
The sequence that makes the money work:
1. Define the value moment, so activation means something. Most products we audit are measuring signup completion and calling it activation, and every number downstream inherits that error. 2. Find where new users stop, with a funnel and session replays, before changing anything. 3. Fix the earliest expensive step. In our engagements the biggest lever sits in the first minutes of the experience nearly every time. 4. Then scale spend, against a funnel where each acquired user is worth substantially more.
Doing step four first is not wrong because paid is bad. It is wrong because you are buying at the worst exchange rate you will ever have, and you are locking in a cost per activated user that the product could have halved.
What is the number to watch?
Cost per activated user, not cost per signup.
Cost per signup is the number your ads platform shows you, and it is the number that improves when you buy cheaper, less interested traffic. It can fall while your business gets worse.
Cost per activated user is spend divided by people who reached the value moment. It moves when the product improves, when targeting improves, and when the promise in your ad matches what the first session delivers. That is the number worth putting on the wall.
It is also the number that connects paid spend to the rest of the funnel, which is why a redesign that ships without moving it has not earned anything.
If you are running paid today and cannot produce that figure, that gap is the finding. The root cause is usually the same one as everything else in this post. The value moment was never defined, so nothing downstream of it can be measured. That is an afternoon of work, and it will change what you do with next month's budget.
WHAT NEXT
Want this fixed in your product, not just explained?
Sep 16, 2026
By Victor Teran
Your Paid Acquisition Is Subsidising a Leaky Product
At median activation, 62 of every 100 people you buy never reach the thing your product is for. Spending more does not fix that. It buys the same leak at a larger scale, and it hides the leak behind a growth number that keeps everyone comfortable.
Retention
Paid
Activation
Paid acquisition is the most reversible growth lever a founder has, which is exactly why it gets pulled first. Turn it up, the numbers move this week, and nobody has to admit anything about the product.
The trouble is arithmetic. Spend flows into the top of a funnel whose conversion rate you did not change. A fixed proportion of every additional pound therefore goes to people who were never going to arrive. At median activation that proportion is most of it.
Key takeaways
Median B2B SaaS activation is 38%. At that rate, 62% of paid spend buys users who never reach the value moment.
Fixing activation is a multiplier on every future pound of ad spend. Increasing spend is not a multiplier on anything.
Q.AI went from 38% to 67% activation with no new features. At constant spend that is roughly 1.76x the activated users for the same money.
Paid works, and it works far better second. The order is the whole argument, not the channel.
What does a leak cost at real numbers?
At 38% activation, more than half of every acquisition pound buys someone who leaves.
Take a founder spending $20,000 a month at a blended $40 per signup. That is 500 signups. At the B2B SaaS median activation rate of 38% (Perspective AI, May 2026), 190 of them reach the value moment and 310 do not.
The effective cost per activated user is not $40. It is $105.
Now hold spend flat and move activation to 67%, which is what happened at Q.AI. The same 500 signups produce 335 activated users at an effective $60 each. Same budget, same traffic, 145 more activated users, and a cost per activated user that fell by 43%.
At $20,000/month, 500 signups | 38% activation | 67% activation |
|---|---|---|
Activated users | 190 | 335 |
Effective cost per activated user | $105 | $60 |
Wasted spend on users who never activate | $12,400 | $6,600 |
Nothing in that table required a bigger budget, a better agency, or a new channel.
The reason cost per activated user is the honest denominator is that activation marks the point a new user first experiences the product's core value (Amplitude). Everything before it is a cost, not a customer.
Why does spending more feel like it is working?
Because the top-line number moves immediately and the damage arrives on a delay.
Signups respond to spend within days. Retention responds within months. So for one or two quarters, more spend genuinely looks like more growth, and the leak scales invisibly underneath it.
By the time the retention curve makes the problem undeniable, the spend is embedded in the plan, the team is built around the volume, and cutting it looks like going backwards. That is a considerably harder conversation than the one available at the start.
This is also why paid gets blamed unfairly. The channel usually worked. It delivered the traffic it promised. What it could not do was fix what happened after the click, and nothing in an ads dashboard makes that distinction visible.
FREE GUIDE
The 4 product leaks costing you growth
A short audit guide for founders. Find the four places your product leaks revenue, and what to fix first.
How do you tell whether the problem is traffic or product?
Compare activation rates by source. If they are all bad, it is not the traffic.
This is the diagnostic that settles most arguments between founders and whoever is running their ads, and it takes an afternoon.
Segment new users by acquisition source and look at activation for each cohort separately:
What you see | What it means |
|---|---|
Every source activates at a similar low rate | The product is the constraint. Fix activation before scaling spend |
One source is far worse than the rest | Targeting or message mismatch on that source. Fix or cut it |
Paid far below organic | You are buying people with less intent. Either the promise is off, or the audience is |
All sources healthy, retention still poor | Activation is fine. The problem is habit formation, not acquisition |
a16z's growth metrics guide makes the general version of this point, segmenting benchmarks by revenue scale, software type and go-to-market motion rather than publishing one number (a16z). A blended figure across sources hides exactly the difference you need to see.
Only the second and third rows are advertising problems. The first and fourth are product problems that no amount of budget or creative testing will solve, and they are the more common findings.
Does this mean don't run paid?
No. It means run it second, once each new user is worth what they should be.
Paid acquisition is genuinely the fastest way to buy learning and volume, and waiting for organic to compound is a luxury most funded companies cannot afford. The argument here is about order, not about the channel.
The sequence that makes the money work:
1. Define the value moment, so activation means something. Most products we audit are measuring signup completion and calling it activation, and every number downstream inherits that error. 2. Find where new users stop, with a funnel and session replays, before changing anything. 3. Fix the earliest expensive step. In our engagements the biggest lever sits in the first minutes of the experience nearly every time. 4. Then scale spend, against a funnel where each acquired user is worth substantially more.
Doing step four first is not wrong because paid is bad. It is wrong because you are buying at the worst exchange rate you will ever have, and you are locking in a cost per activated user that the product could have halved.
What is the number to watch?
Cost per activated user, not cost per signup.
Cost per signup is the number your ads platform shows you, and it is the number that improves when you buy cheaper, less interested traffic. It can fall while your business gets worse.
Cost per activated user is spend divided by people who reached the value moment. It moves when the product improves, when targeting improves, and when the promise in your ad matches what the first session delivers. That is the number worth putting on the wall.
It is also the number that connects paid spend to the rest of the funnel, which is why a redesign that ships without moving it has not earned anything.
If you are running paid today and cannot produce that figure, that gap is the finding. The root cause is usually the same one as everything else in this post. The value moment was never defined, so nothing downstream of it can be measured. That is an afternoon of work, and it will change what you do with next month's budget.
WHAT NEXT