Once upon a time in Hollywood in South Africa, my first startup had a problem.

We were building a job marketplace/online casting platform, matching people with roles in commercials and TV shows. Our biggest category that customers like Airbnb used us for was “real everyday people” for branded content. The shift in the industry we played into was authentic casting…people wanted real couples, real chefs, real surfers.

The problem was their profiles were notoriously real bad. They didn’t have the right types of photos or videos to get them through the door with clients. Customers were looking for people, seeing bad profiles, and scrolling on.

Back in 2016, I was also a notorious crook of Airbnb. I idolized their marketplace and brand and story. And one of the things I stole was an early idea of theirs, since they once had the same problem—grainy phone shots, or pics so bad that any sane traveler evaluating a couch would have been terrified.

We spun up our own version of their photography service and ran monthly photo/video sessions for people. We rented a studio with professional gear, had hair & makeup and stylists, and invited our users to come and enjoy being on a proper set, get high quality assets to make their profiles better, and even hired a casting director to give people tips while auditioning and being on camera.

The users who came to these sessions obviously got better profiles, they got audition and booking requests at much higher rates than the people who didn’t show up (and for higher paying roles), and also, they were the people who ended up loving WECAST the most and being the most engaged.

And like Airbnb who saw their revenue double within just a week post launching their photo product, we kept this initiative 100% free. It was so effective, that when we expanded into a marketplace for producers to find/book film students for roles like production assistants and assistant camera operators, we even started making our own brand videos entirely with our own users. Top to bottom. Rented locations, gear, sets of 30+ people sometimes. Paying them and giving them credits for their profiles to show they had actual work experience, with proof.

Our users loved this, had so much fun, and always messaged asking if they could be part of the next one.

We built a system that helped our users get better media for their profiles and get experience on sets—which helped them get more jobs and make us more money.

This was never part of the product itself. It was a free sidecar bolted next to it, built to kill the one thing standing between a profile or place and a booking: I don't have good media.

So, why write about this now?

Because in the most useful product research I've done with Little Moments—watching my wife give up on my own app in front of me—I’m seeing a similar problem come up and I think there’s a broader idea that relates to a universal problem across almost any product category based on one question: How do you save a behavior stall?

🥷 Build a free complement/sidecar product whose only job is to make your existing customers use the paid product more. A separate free thing that grows their appetite and makes your core thing more valuable.

In 2006, Nike did something a shoe company has no business doing: it shipped software. Up to then it had a strange blind spot. The moment you bought a pair of shoes, it lost you. You ran in them on a treadmill, a trail, wherever, and Nike had no idea. No relationship, no reason for you to think about it again until the soles gave out.

That software is now Nike Run Club: free, and it'll never charge you a cent. I say you because I hate running. So it’s not a we. It tracks your runs and hands you streaks, guided miles, and a hit of dopamine for the runners who need more of it every time they lace up.

And every one of those runs does the same thing underneath—it wears down a pair of shoes. Running shoes are toast at roughly 300–500 miles, and the app's whole job is to get you there faster. Strip the gamified runs away, and what's left is a shoe-consumption engine in running shorts.

I think of it as manufacturing/stimulating demand. Appetite for something isn't a fixed quantity you sit around and wait for. It's something you can build on purpose.

And what I like about this move now in particular, is the cost of doing it has gone way down. Those sessions I ran cost real money and it was definitely marked as a loss-leader. Nike Run Club wasn’t free by any means to make. A second product often meant a second team, a second roadmap, a second year of runway. Obviously with AI now that isn’t true anymore and the move has come downmarket for anyone with some creativity and a Claude subscription—which is everyone reading this.

Honestly, this is a great example of the idea in action. RevenueCat clearly asked that question and realized the way to drive more consumption and usage of their actual product—the subscription operating system of paid apps—was to get the people making the apps to keep building and growing their app.

A lot can get in the way of that. Drop off for new apps is high. Thousands are launched a month and never get one update. RevenueCat is 100% free to all these apps making under $2,500 MRR, so their goal is getting new ones past that point so they can earn.

This is why they launched StartApp School — completely free, bite-sized lessons on the actual job of building an app people will pay for. Paywalls, trials, pricing, retention: the stuff people fumble the first time.

If you've got a subscription app, or you're even toying with the idea, go work through it. It's free, it's fast, and it's made by the team a huge number of app builders trust to run their subscriptions.

1. The enemy is usually under-use. Most products don't churn because someone else shipped a better one. They churn because the customer never used them enough to feel why they mattered. The highest-leverage lever is often getting the people already paying you to lean on you twice as hard. No CAC, no ad auction.

2. The demand is already there and doesn’t need to be reinvented. You're just unlocking it, and the reason this works without feeling like a trick is because the want is already in the room. In Airbnb’s case people were there looking for a place to stay; in mine they were trying to find the right person. The gap wasn’t actually the right place or right person, it was something else blocking it. Your customers generally would happily use you more. They pay you after all and signed up— they're just missing the next idea, the nudge, or the confidence. The runner wants to run; perhaps they don't know where? You're simply using another thing to clear the thing holding back the desire that already exists.

3. A free complement compounds where a free version cannibalizes. The freemium thing of your main product hands over a slice of what you charge for, and people like to pitch a tent there forever. A free complement that is a totally different thing to your main product pulls the other way: the more they use it the faster they reach for the part you sell. Razor-and-blade logic turned inside out—you give away the thing that makes the blades run out faster.

4. Loss-leaders keep people coming back. My severely overpriced grocery store sells what they call 5-Buck Clucks! It’s a whole cooked rotisserie chicken for just $5. I go get it sometimes. But I also get other stuff. The chicken pulls me inside over and over again, which is the same reason Costco hotdogs are so damn cheap. $1.5 forever is on purpose. The sidecar is that same trade moved from the checkout aisle into the product: you eat the cost of something free so the thing you charge for gets used more.

5. A sidecar doesn’t have to be forever. Giving away something for free forever can sound scary. But Airbnb's free photography existed to kill one friction: hosts couldn't show what they were renting. It worked, scaled to a global photographer network, and then once phone cameras got good enough that Airbnb itself now tells hosts "a good-quality smartphone is often sufficient" they retired the free version and turned it into an optional paid add-on, billed out of your next payout. The friction died so the sidecar's job was done. Build yours to solve a real gap now, not to carry forever.

6. You can run this play for new users too. This idea so far has been on existing customers. Pushing along the demand you already have and supporting your main product. But you could also do something similar called engineering as marketing, where the free thing pulls in new users instead of deepening the ones you have. For example JWT.io (Auth0), HubSpot's Website Grader, Algolia's DocSearch. Same sidecar philosophy but pointed at acquisition instead of consumption. I say this to say that sidecar thinking can be pointed at different problems you might have—acquisition or usage/retention.

You have the idea, the context, the why. Then it gets flattened into a vague Jira ticket, handed off, and rebuilt by someone or some agent, missing half of what you meant.

Rezonant is built to stop that leak. It's a product workspace that takes the messy idea, pressure-tests it like a sharp teammate would, and shapes it into engineering-ready work: tickets with full context, synced to your real Jira, Linear, docs, and repo, ready to push or hand straight to a coding agent.

It exists for the gap we know about—product intent getting diluted on its way to shipped code. Rezonant carries it the whole way through—find out how below.

Besides Airbnb and Nike, there are some other examples of this move. Notice the format…How do we get people [doing your consumable thing] more by doing or providing something different for free?

You might be thinking, what is a tire company doing here?

Well, in 1900 there were about 3,000 cars in all of France, which is a problem when you sell tires. So Michelin printed a guide to restaurants worth driving to. People drove farther, wore out tires faster, bought more Michelins. A 124-year-old free product so good it became the world's authority on fine dining built to move rubber.

Takeaway: the best sidecar can become its own thing beyond the thing it was built to sell.

I used to have one of their machines. And I can confirm that as a coffee lover and person who likes to create things, their free recipe app was a very smart free add-on. Inside of it, a good share of its best drinks call for two or three pods instead of one. I fell for it, but it didn’t feel like falling for it because I was making delicious coffees. In turn, my machine was being fed more pods more often.

Takeaway: build the helper around the consumable unit you actually sell.

A monthly magazine of the community's best images, pulled from the month's 10,000 top-rated generations. Getting featured is a status prize and the only way to play for it is to generate more. Meaning burning the credits that are the product.

OpenAI is doing the same thing with token awards for hitting milestones like YouTube views.

Takeaway: turn the behavior that costs you money into the one that wins people status.

Free printables obviously. Coloring pages, planners, party kits, you name it. They all exist because every page drains the ink cartridge that's HP's real margin.

Takeaway: if you sell the consumable, give away things that get consumed.

I touched on it earlier, but it was in today’s ad and you might have skipped the ad, so I’ll mention it again because it’s a great example.

RevenueCat built a free high quality e-learning platform for app growth. Their freemium product is simply a free version of everything until you cross $2.5k MRR. That makes a clear priority with aligned incentives helping their customer be as successful as possible getting paid users for their products.

Their StartApp School gets experts teaching people how to do it. And this example does two things…it helps the existing base do better, and it attracts new people who might start with the free material and then naturally flow into applying it with RevenueCat.

Takeaway: your sidecar product can do double time work—acquisition and engagement/retention

It comes down looking for where and why people stop using your thing. What’s happening the moment they close the product?

There's almost always a stall. Some moment users run out of ideas, lose the thread, or hit a wall and drift off.

You don't have to guess where it is either. The people who churned will often tell you the exact moment they quit if you ask them.

And your sidecar doesn't have to be a whole app or new service like like taking photos in a studio with a crew. It can be some standalone tool, a calculator, a template that kills the blank page, a tracker, a community. Pick the format that fits the stall in the behavior.

It can even be a feature within the main app, vs a sidecar thing…back to what I said about my app Little Moments.

We were lying on the couch. I made Julia feel guilty for not using my app and said “open it”. She did and I leaned in to watch. She was on the Capture page to pick a moment, and had no photos. She picked the day before. No photos. Then she closed the app.

The app asks you to capture one small moment a day—a photo (which can be a 2s live photo or 2s of a video clip) and a couple of lines or more if you have more to say. This one 30s session on the couch showed the problem, which seemed more true after watching more replays. People had intention to capture, but some days there was just nothing in front of them worth capturing. And if you took no photos on Tuesday and Tuesday is over, well, you can't go back and grab a moment for a day you missed. It's gone.

So in my next build, I’m shipping something called Smart Moment Nudges. Opt in, turn on location, and the app gives you a gentle nudge when you're out somewhere (e.g park, a restaurant, a new neighborhood) to grab a shot while you're in it. Works on a geofence. The thing capping usage wasn't motivation, it was "I've got nothing to capture today” which in a new app quickly becomes “close, goodbye see you never”. So I’m trying to build a mechanic that puts something in front of you before the day slips by. And this pairs with the “Throwback” feature, showing you photos from This Day In The Past.

In short, before you spend another dollar getting a stranger to notice you, look hard at the people already paying you and ask the cheaper question: what's the one thing that, if they did far more of it, would make you impossible to cancel. And then, what could you give them for free that gets them doing it?

Demand isn't only something you go out and hunt for. Among the people who already chose you it's something you can build and convert. And if you run the numbers, you’ll see the cheapest growth on the table is the customer you already have using you twice as much.

So, where are your users stalling?

Happy tinkering, testing, and trying.

› “My favorite productivity hack is to stop trying to be productive all of the time.” — Justin Welsh

› beehiiv is the all-in-one platform that brings together newsletters, websites, and every tool you need to grow and earn. Start a newsletter today, it might be the best decision you ever make. Use my code THEDIFF30 for 30% off your first 3 months (Start a newsletter).

› While Claude or Codex is grinding away in your terminal, you're mostly just watching a spinner. I built something to fill that dead air: one line to install, and it feeds you a fresh idea or a bit of news from the topics you care about, right in your status line, while the agent finishes. Learn something in the gaps instead of refreshing X. (Install bits)

› My free builder tools do the math for you — an app revenue calculator, an app health calculator, and a competition calculator (See how your app compares).

› Check out my personal app store of tools I actually use — these are the vetted products I lean on to build things and run my business (Explore my stack).