You own a growth channel if customers keep arriving from it after you stop paying. You rent one if the flow stops the month the invoices do. Most companies past product market fit run almost entirely on rented channels, and the rent only moves in one direction.
The test for sorting your own budget takes 20 minutes and needs nothing but last month's numbers.
The budget conversation runs the same way in most growth companies. Growth slowed, so the plan on the table is to raise the ad budget. It worked last year, in the sense that spend went up and revenue went up behind it.
Then somebody asks what happens if the spending stops. The room goes quiet, because everyone already knows: the new business stops with it.
There is a word for paying every month for something that disappears the month you stop. Nobody in these meetings uses it about their growth engine.
Do you own your growth or rent it?
The sort is simpler than any attribution model. Go down every line of the marketing budget and ask one question: if this payment stopped tomorrow, what would still arrive next month?
Whatever still arrives after the last invoice clears is what you own.
Paid social, search ads, retargeting, sponsored placements: the flow stops within days. That spend can still be worth every dollar. It is rent, and rent buys real shelter.
An email list, an affiliate bench, pages that rank, customers who refer: those keep producing next month whether or not you paid this month. They are assets. They took longer to build, and they are the reason some companies get cheaper to grow while others get more expensive.
The founder who ran the experiment
In August 2026 a two-time founder named Ankit Agarwal shut down Klydo, his quick-commerce fashion company in Bangalore, and published the numbers.
By the standards of the playbook, the team had done everything right. More than 50,000 orders. Conversion in the quick-delivery cluster ran 35-40% above the slower deliveries, and buyer retention hit 1.8x. The business reached contribution-margin breakeven on a fraction of what peers raised.
The playbook says all of that is supposed to make acquisition cheaper. Loyal customers come back on their own, word spreads, and paid spend becomes a smaller share of each new cohort.
His acquisition cost never moved. ₹900 to ₹1,200 per customer, first month to last.
His explanation deserves to be read in his own words: "When someone else owns the discovery surface, your loyal customer still comes back through a retargeting ad. You rent the relationship every time."
Retention lived in his product. The door his customers walked through lived on a platform. Every improvement he made accrued to one side of that arrangement, and the platform priced the door at auction anyway.
Why the rent always rises
An ad auction has a property that founders feel long before they can name it: your price is set by the next bidder rather than by your own performance.
You can improve your creative, your landing page, your offer. So can everyone bidding against you, and new bidders keep arriving. Yesterday's winning bid becomes the floor under tomorrow's.
The platform's incentives point the same direction. Its revenue is the sum of everyone's CAC. A platform that made your loyal customers free to reach would be donating its own margin, which is why the retargeting ad exists at all.
Layers on top of the platform inherit the arrangement: an agency retainer on top of auction spend is rent for managing rent. And when the blend of channels shifts underneath a fixed budget, the average cost moves on its own, a mechanism worth understanding separately.
The auction converts every competitor's funding round into your acquisition cost.
What owning a channel actually looks like
At FX Replay, a company that makes trading simulation software, I ran growth for 18 months. The ad account was the visible engine. The compounding happened elsewhere.
The email program grew into 5.9 million monthly sends opening at 26%. The affiliate bench grew past 4,000 partners returning roughly 5:1. Both were built once and then produced monthly, on terms no auction could reprice.
When a Black Friday came, the sequencing ran owned first, partners second, paid last. Paid amplified a machine that already worked without it.
YouMail, a call-protection app I worked with for years, is the same lesson from the other side. Its front door belonged to the app stores, and the move that changed its economics was building a web funnel it owned, where conversion went from 9% to 22% on pages it controlled end to end.
Different companies, same shape: the growth that compounded came from assets, and the assets had to be deliberately built while the rented channels were still doing the heavy lifting.
Rent has a use
None of this argues for turning the ads off. Rented reach is how a company without an audience gets its first one. Paid is still the fastest way to test a message, reach a cold market on a schedule, or fill the top of a funnel while slower assets mature.
The failure mode is renting forever: years of spend in which every single customer arrives with a same-day payment attached, and none of the traffic leaves behind a list, a bench, a page, or a referral loop.
The companies that escape do it while the rent is still flowing. Every paid visit is a chance to capture an email. Every paid cohort can seed a referral. Every winning ad message can become a page that ranks for the same intent.
The 20-minute audit
Three questions, one sitting, last month's numbers in front of you.
The first question is the one that matters most, and it is the one most companies cannot answer, because their reporting was never asked to separate arrivals that were paid for that day from arrivals that were not.
That share, tracked monthly, is the closest thing to a deed your growth engine has. If it rises while revenue grows, you are buying assets with your marketing budget. If it sits at zero year after year, you are a tenant with good revenue.
Tenancy is survivable right up until the landlord reprices, and the auction repricing is not a risk. It is a schedule.
Sorting a company's channels into owned and rented is one of the first things I do inside a growth audit, because the ratio decides what the next dollar should buy. Let's talk.
