> ## Content Index
> Fetch the complete content index at: https://longtalkshort.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Fifty Creators, No Two Alike: Wispr Flow's Hector Serrano on Paying for Ads That Perform
- URL: https://longtalkshort.com/hector-serrano-fifty-creators-no-two-alike-wispr-flow-ads/
- Published: 2026-11-02T16:13:47.000Z
- Updated: 2026-11-02T16:13:47.000Z
- Description: Wispr Flow's Hector Serrano explains how about 50 creators on commission make the AI app's ads: a pay floor, handpicked and deliberately varied recruits, and a push to experiment. The playbook, with its limits.
- Author: Truman
- Tags: Economics, Hector Serrano, Lenny Rachitsky, Lenny's Podcast

**If you sell software and your ads have started to look like everyone else's, Hector Serrano has an unusual suggestion: stop writing them yourself.** Serrano, head of content at Wispr Flow, the AI dictation app, explains to Lenny Rachitsky how the company built a roster of about 50 outside creators who make its ads and get paid when those ads perform. He starts with a caveat: none of this works unless the product is already growing on its own.

*In short: Treat paid ads as a fast way to learn which messages work, not just as a growth tap. Pay creators a share of what their ads earn, with a minimum to keep them going, and recruit a small, deliberately varied group by hand. Push them to experiment instead of copying the top ad. The numbers are the company's own, and the approach fits best with self-serve products.*

## 1\. Product first, then pour fuel

Founders often ask Serrano for "the hack". His answer is that there isn't one: Wispr Flow found product-market fit about six months in, and every channel after that has been "more fuel to the fire". **If your product isn't growing by itself, he says, go back to its positioning before you spend on ads.**

Once it is growing, he argues for turning on paid ads early, borrowing an e-commerce habit that many software companies avoid. **Running ads to strangers quickly shows whether your message lands** and whether cold traffic turns into users, which is far easier to measure than organic posts. The worry about exhausting your audience, he says, mostly goes away if you keep finding fresh angles and types of customer.

## 2\. Pay for results, not views

Wispr Flow first tried "organic UGC" (user-generated content): creators shared videos about the app on side accounts and were paid per thousand views. An intern's programme reportedly reached 500 million views in about ten weeks, but few viewers became paying customers, so the company shut it down. **The lesson Serrano took is to stop chasing vanity metrics.**

Its model runs on Trybe, a platform that lets brands run creators' videos as ads on Meta and pay them according to performance. For a shop, creators get a percentage of sales. **For software, where a free user might convert weeks later, Wispr pays a percentage of the ad spend behind each video.**

Creators starting out also get a "smart retainer": make, say, 25 videos in a month and you're guaranteed $1,500\. If your commissions come to $500, Wispr adds $1,000; if they reach $3,000, you just keep the commission. Payouts are weekly, earnings are uncapped, and briefs are kept minimal. "There's nothing more demotivating," says Serrano, himself a creator. So how do you find the right creators?

![A chart of creator pay against commissions earned, showing a flat floor at $1,500 that turns into a rising diagonal line once commissions pass $1,500, with two marked examples: $500 in commission topped up to $1,500, and $3,000 in commission paid as is.](https://storage.ghost.io/c/d7/93/d793216c-cf13-441f-bdae-4f3aef998c0e/content/images/2026/11/fig1-11.png)

How the "smart retainer" described in the episode works, using its example figures. Diagram by Long Talk Short.

## 3\. Recruit few, and recruit different

An unknown brand on a creator marketplace has no track record, so the best creators ignore it. Serrano started off-platform instead. He searched Instagram for niche keywords such as tech, lawyer or nine-to-five, then went through the followers of creators he liked, sending 150 to 250 handpicked messages over a week. That produced five or six good creators. He then grew the group five at a time, and accepts only one to three of the 1,000 or more weekly applications he says it receives.

**Variety matters more than numbers.** Each creator should speak to a different type of customer: parents, lawyers, engineers, sales reps. Technical audiences, he finds, respond to "boring" and "ugly" ads, sometimes just a long screen recording. Nobody will buy a $3,000 sleep product from an 18-year-old. To keep the group engaged, he holds weekly calls, uses creator coaches and organises a retreat.

## 4\. Fight sameness

The danger, he says, is the leaderboard. When creators see which ads get the most spend, they copy them, and **Meta's ad system groups near-identical ads together, so the copies stall.** At Wispr's spending level, he says, an ad can burn out within weeks once it reaches about $100,000 in spend, so the company launches 500 to 600 fresh ads a month.

His answer is to tell creators that the top earners got there by experimenting. One hit showed raindrops "typing" on a keyboard left out on a roof. He also keeps AI out of anything customers see, because "AI slop" invites backlash.

![A five-step flow: choose a pay model, recruit off-platform, build community, scale five at a time, and push for different ads. Below it, three cards: a creative lead, a media buyer, and organic channels such as founder-led content and YouTube.](https://storage.ghost.io/c/d7/93/d793216c-cf13-441f-bdae-4f3aef998c0e/content/images/2026/11/fig2-11.png)

The playbook as described in the episode, in our grouping. Diagram by Long Talk Short.

## 5\. Who runs it, and the wider flywheel

His hiring advice is to look past résumés. **Find someone who has built a following, even 10,000 people**, which shows they understand the consistency it takes, then give them room to fail, the way a football academy develops players. Pair them with a media buyer who manages the ad accounts. For founder-led content, he suggests treating the founder's channel like a Netflix series tied to the company mission, investing most in YouTube packaging, and remembering that every organic touchpoint makes paid ads cheaper.

## Weighing it: what holds up

**Framing:** The episode's promise of "paid-growth alpha" overstates it. It includes sponsor reads, and much of it is about hiring, creator careers and founder content rather than ads. "Alpha" also implies an advantage that will shrink as more brands copy it.

**Holds up:** In August 2026, Wispr raised $280 million at a $2 billion valuation in a round led by Menlo Ventures, and it says revenue grew more than 30 times year on year, a company figure that isn't audited. **Meta says its Andromeda system groups ads with similar visuals and themes and treats them as one creative**, and it recommends genuinely different concepts. Meta publishes no similarity threshold, and the percentage in the captions is garbled. Lamine Yamal did win the World Cup with Spain at 19.

**Partly holds:** This is one company with unusually strong word of mouth. Rachitsky suggests it fits bottom-up products better than enterprise suites; Serrano thinks it "works for everyone". The idea that buyers need about seven exposures is marketing folklore. **Our own observation: paying creators a share of ad spend rewards whatever the media buyer chooses to scale, not conversions directly.** If you copy the model, U.S. FTC rules require paid endorsements to be disclosed.

**Not checked:** the internal figures: $13,000-a-month creators, $300,000 paid out, 50 creators, 5,000 lifetime ads, 120 employees. Treat them as the company's claims.

**Disclosures:** Serrano is a Wispr employee, so this is also promotion and recruiting for his company. He calls Trybe the best platform, and no affiliation with it was mentioned either way. Rachitsky has said he is a fan of Wispr Flow, not an investor. The episode has ads for WorkOS and DX. We worked from YouTube's automatic captions ("Tribe" for Trybe, for example), so quotes may contain small mistakes.

**Bottom line:** the pay-for-performance structure and the push for variety are sensible and easy to test. The results are one fast-growing company's, self-reported.

**Related reading:** [The Seed Investor Who Won't Fight the Giants: David Tisch](https://longtalkshort.com/david-tisch-boxgroup-20vc-seed-investor-wont-fight-the-giants/) · [Patient on the Goal, Impatient on the Day: how Jeff Bezos ran Amazon](https://longtalkshort.com/ethan-evans-jeff-bezos-strategic-patience-amazon/)

## Long Talk Short

- Get the product growing first; ads add fuel, they don't create the fire.
- Use paid ads early, as a fast way to test your messaging on strangers.
- Pay creators a share of what their ads earn, plus a floor, and don't hand them scripts.
- Recruit a small, deliberately varied group by hand, and push them to experiment.
- Copies of a winning ad stall; fresh concepts keep the system working.

**The one thing to remember:** on platforms that punish sameness, a varied group of motivated creators may beat a bigger budget, as long as the product is already working.

*This summary is general information, not business, legal or investment advice.*

Source: Lenny Rachitsky with Hector Serrano, *Lenny's Podcast*, on YouTube. This is an independent summary and is not affiliated with Lenny's Podcast, Wispr Flow or Trybe.