Key Takeaways
- Wispr Flow launches 500 to 700 new UGC ads a month, but works with only 40 to 60 creators. They chose fewer, better creators and scaled how much each one makes.
- Creators are paid 7% of the ad spend behind their ads, uncapped, paid weekly. New creators get a $1,000 monthly minimum for 20 to 25 videos.
- Each creator is assigned a gap: a customer segment, format or country no other ad is covering.
- No two ads are allowed to look alike. Their view is that Meta's Andromeda groups similar ads and treats them as one.
- Their Meta lead optimizes for customers who build a habit, not just downloads, and gives every ad one concept, one angle and one awareness level.
- Paying a percentage of spend means creative costs rise every time you scale. That works for some companies. Others need a fixed budget.
In October 2026, Hector, Wispr Flow's head of content, posted a 14-minute video on X titled "i built @WisprFlow's paid ads program, here's how i'd rebuild it in 30 days." It passed 1.6 million views in a few days.
To be clear up front: we don't work with Wispr Flow, and none of this is our work. We run creative for brands spending $100K to $5M a month on Meta, including tech brands like Uber and Constant Contact, so we watched the video the way most ad people did, taking notes. Below is how the program works, in their own framing, and what we think other brands should take from it.
1. They Went All In on Paid UGC
About a year ago, Wispr Flow stopped organic UGC completely and moved everything into paid.
Their reasoning: organic only works if you have real talent on the team. Paid works for anyone who can ship enough different ads. You don't need a genius. You need a lot of ads that each look and sound different, and a system that keeps them coming.
The output today is 500 to 700 new UGC ads a month, with a target of 1,000.
2. 500+ Ads a Month Doesn't Mean Hundreds of Creators
This is the part we'd underline.
Wispr Flow gets to 500 to 700 ads a month with 40 to 60 creators. That's roughly 10 to 15 ads per creator per month, which lines up with the 20 to 25 videos new creators are asked for.
They didn't get there by signing everyone. They accepted about 50 creators out of more than 5,000 applications, and scaled in steps: 5 creators, then 10, 20, 40, and stopped at 40 to 60. Going from 5 to 50 took four months.
Keeping it small is deliberate. In Hector's words, $100,000 paid to 20 creators looks a lot better to a creator than $200,000 paid to 1,000. A small group means real earnings, which means the best creators stay and keep producing.
Quality of creators, quantity of ads. You don't need a roster of hundreds. You need a few dozen good UGC creators who can each make a lot of different ads.
It's also why Hector says most creator programs fail. Brands accept everyone who applies, nobody coaches the creators, and creators join without understanding they only get paid when their ads perform.
How they found them: Instagram keyword search ("UGC creator", "tech", "SaaS"), then mining the followers and comments of every good creator they found, sending 100 to 200 DMs a week. Creators also get $100 for referring someone at their level or above.
3. Creators Get Paid a Cut of Ad Spend
Ecommerce brands often pay creators a cut of each sale. That doesn't work well for software, where a sale can take weeks to close. So Wispr pays creators a percentage of the ad spend behind their ads instead.
The terms:
- 7% of ad spend, uncapped, paid out every 7 days.
- A guaranteed minimum for new creators: $1,000 a month for 20 to 25 videos. If a creator earns $500 in commissions, Wispr pays the other $500. Once they earn more than $1,000 on their own, the guarantee goes away.
The minimum is how they signed good creators before the program had a track record. The commission is what keeps the good ones around, since spend only goes behind ads that perform.
According to the video, the program has paid out $220,000 so far, and the top creator makes $12,000 a month.
4. The Tradeoff of Paying a Percentage of Spend
Paying on spend lines up incentives nicely. It also means your creative bill grows every time you scale.
The math is simple. At 7%:
- $500K a month behind creator ads is $35K a month in commissions.
- $2M a month is $140K a month.
- $5M a month is $350K a month.
A brand spending at Wispr's level could easily be paying multi-six figures a month in creator commissions. And the better an ad does, the more it costs you to keep running it.
For a venture-backed company in growth mode, that can be a fine trade. For most mid-size and larger companies, it's a harder one. Their budgets get set ahead of time, and a creative line that moves with media spend is tough to plan around.
That's the reason we charge a flat fee. If an ad really takes off, you can let it rip and keep the margin, instead of paying a bigger invoice just because you spent more behind the creative. Both models work. You just need to decide which kind of risk fits your budget before you build the program around it.
5. One Creator Per Gap
Each creator is assigned a gap, not a product:
- a type of customer no current ad is speaking to
- a format nobody on the team can shoot
- a country with nothing running
One creator per gap. This keeps the ads from blurring together and pushes new spend toward people the account isn't already reaching.
You can see it in Wispr's ads. The same product gets sold through very different angles: accuracy and fixing mistakes mid-sentence, dictating into Slack and iMessage, speed vs. typing, and meeting notes.
6. No Two Ads Can Look Alike
This is the rule the whole system is built around.
Hector's view is that Meta's Andromeda update clusters videos that are 50 to 60% similar and suppresses them. So 1,000 talking heads saying roughly the same thing get treated as one ad. Meta hasn't published an exact threshold, but the direction matches what Meta has said about Andromeda rewarding creative diversity.
It explains a rule that sounds backwards. When creators see which ad is winning, they want to make their own version of it. Wispr stops them and pushes them to film something new instead. A copy of a winner gets grouped with the original, so it never gets shown.
That's the opposite of the "iterate on winners" advice most accounts run on. The useful distinction: a new hook, a new person, a new setting or a new angle is a new ad. The same script in the same format with a different face may not be.
7. Creators Are Run Like a Sales Team
Each creator gets:
- a 30-minute onboarding
- a weekly call
- 1:1s and creator coaches
- access to the ad account
Every Thursday, the group gets walked through what's working, which customers to go after next and what to test.
Showing creators real performance data is rare. Most brands send a brief and never tell the creator how the ad did. Here, creators can see what wins and why, which is a big part of how 50 people turn out hundreds of ads a month that don't all look the same.
8. How Their Meta Lead Runs the Account
Wispr Flow's performance marketing lead posted a follow-up video on how the Meta side runs. It's less about creators and more about everything around them. The parts worth stealing:
Optimize for good customers, not just downloads
For ecommerce, a good result is a purchase. For a lead gen business, it's a call. For an app like Wispr Flow, it's less obvious. A download is good. A purchase is good. But the signal that predicts a real customer sits in between.
For Wispr, it's habit: how many words someone dictates after a certain point. That's the event they rally around, and it's how they judge whether an ad brought in the right people.
Before you spend, work out the steps a customer takes from cold to paying, and which of those steps lines up with your best customers. That's what Meta should be optimizing for.
Find the objections first
Wispr pulls real questions from reviews, Reddit, support tickets and comments on its own ads. The big one: why pay for dictation when it's free on your phone and your Mac?
A lot of their ads answer that directly. Others just show the product making someone's day easier. Either way, the ad is built around a question real people are already asking.
Brief one person, one situation, one promise, one proof
Don't hand creators a full script. Their Meta lead calls that a recipe for disaster. Instead, the brief covers:
- the one person the ad is talking to
- the situation they're in
- the promise the product makes
- one piece of proof
Then hire creators who are already fans, so they don't need the product explained. Ideally they look like the customer: real estate agents, lawyers, developers, founders, students.
And don't hire on follower count. Some of their best creators have about a thousand followers.
Give every ad a concept, an angle and an awareness level
One problem, one person, one way. Every Wispr ad is defined by three things:
- Concept: the one product truth the ad is about. About 90% of the time, it's just one.
- Angle: how you come at it. Saving time at work, or working better with ChatGPT.
- Awareness level: who it's for. Someone who has never thought about dictation, or a Claude Code power user looking for something that works across apps.
Map those out and you have about 150 ads. More importantly, you can see which feature, told which way, to which audience, is winning, and feed that back into the next round of briefs.
Test the funnel before you spend a dollar
Swipe up on your own ad on your phone, inside Instagram's built-in browser, not Safari. Does the page load? Do your conversion events fire? Does the landing page match what the creator just said?
Then read the funnel step by step:
- Good hook and hold rate, no clicks: the ad got attention but didn't give a reason to act.
- Clicks, no conversions: the landing page probably needs work.
- Conversions, but the wrong customers: look at your campaign setup and bidding. Cost caps are worth testing.
Give ads time, and scale in steps
Their Meta lead says you probably won't learn anything for about five days. An ad launched on a Friday reaches a different crowd than one launched on a Monday.
Winners also don't always survive the jump. A small winner in testing can die when it moves to a scaling campaign with ten times the budget. Scale in steps you can afford.
Keep a company brain
Since April, Wispr has kept an internal "marketing OS" in a GitHub repo. Lessons from paid, landing pages, newsletters and creator partnerships all go in, and they feed the next round of briefs. AI coding tools like Codex and Claude Code help keep it running.
Without something like this, 50 new creator ads land and nobody has a system for sorting what worked, launching fast and moving on.
Respect the viewer's time
Google catches people searching for something. Meta reaches people scrolling after a long day. So an ad has to entertain, make them feel good, or at least not waste the 15 seconds.
His last point: creators are your sales team now. They're in front of more people than your sales reps ever will be. Pay the good ones what they're worth.
The Hard Part Isn't the Ads
Read the program back and most of it isn't filming. It's everything around the filming:
- deciding what the ads should say, and which angles are missing
- finding the gaps in the account and assigning the right creator to each one
- recruiting, vetting and onboarding creators
- weekly coaching, feedback and reviewing every video
- keeping hundreds of ads a month different enough that Meta treats them as different
That's a full-time job, and a hard one to hire for. You need someone who understands creative strategy, the ad account and how to manage creators, all at once. Most brands don't have that person, and it takes a long time to find one. Wispr's own Meta lead calls it "a very difficult job that AI cannot do for you yet."
High volume works, but the volume comes out of the system, not the other way around. Across our own accounts, the top 10 ads take a median 41% of spend, and two out of three of September's top-10 creatives were new that month. You make a lot of different ads because you can't predict which few will carry the account.
Should You Copy It?
Some of it, yes. Some of it only works at Wispr's scale.
Worth copying at almost any budget:
- Fewer, better creators. A few dozen good people making lots of different ads beats a roster of hundreds.
- Give every creator a gap to fill. A segment, format or market nobody else is covering.
- Make "different" the brief. Don't let creators clone the winner. Ask for a new hook, setting or angle.
- Show creators the results. A weekly look at what's working beats any written brief.
- Brief one person, one problem, one proof. Let creators who already like the product say it their way.
- Test the funnel on your phone first. Make sure the page loads and the events fire before you spend.
Think harder about:
- The pay model. A percentage of spend is great when budgets are flexible. If yours are set ahead of time, a cost that grows with every dollar you scale can become a problem.
- The volume. 500+ ads a month needs the budget to give each one a real read. Smaller accounts should run fewer ads, each more different from the last. See how many ads to run on Meta for what accounts at different spend levels actually run.
For more on ad volume, hook rates and lifespans across accounts, see the Meta Ads Statistics hub.
Want This Without Building It?
That's what we do. We handle the creative strategy, find the gaps, brief and manage the creators, and deliver the ads: UGC, plus motion, static, GIF and hybrid ads. Flat fee, so the cost doesn't climb when an ad takes off. See our UGC ad agency page, or book an intro call to talk through your account.
