Appbroda's $2M Run Rate Is a Take Rate on $20M of Someone Else's Ad Revenue
On tape in January 2023, Appbroda's Ashish Aggarwal declined to give a revenue figure. The $2M run rate got assembled anyway, out of three numbers he did give — and it is a slice of his customers' money, not a subscription line of his own.
On this page
Nathan Latka starts every interview with the same question, and Ashish Aggarwal declined to answer it.
We can’t disclose exact figures, but it’s we’re just about twenty months old, and we are between the 0 and $5,000,000 revenue range already.
A $0–5M band is not a revenue figure; it is the polite version of no. Eleven minutes later the number was on the table anyway, and Aggarwal had signed off on it — not because he gave it up, but because he answered three narrower questions accurately and the arithmetic finished the job. That is the interesting thing about Appbroda, recorded in late January 2023: its revenue line is not something the company reports. It is something you calculate from a number the company does not own.
Whose money is being counted. Appbroda charges a percentage of the advertising revenue its customers earn. In the twelve months to January 2023 those customers earned somewhere around $20–30 million from mobile ads, and Appbroda kept 8–12% of it. The eight-figure number belongs to 320 game developers. The seven-figure number is the business.
The number he wouldn’t give, assembled from parts
Latka never asked for the revenue figure again. He asked for volume, then for price, then for the cut — and Aggarwal answered all three.
- Ten billion ad impressions. Asked how you measure whether a developer is pushing real volume through the platform, Aggarwal reached for the ad industry’s version of GMV: “for example, last year, 2022, January, December, we touched about 10,000,000,000 ad impressions, which is a very large number, but, you know, like, it’s it’s not that large in the ad world.”
- A blended $3–4 CPM. The value of an impression depends on where the phone is. “CPCs vary from US to India. So India will be on the lower end and US will be high. Blended, maybe you can look at, like, maybe 3 to $4 is what that means.”
- His own multiplication. Latka first heard $3 as a cost per click, and Aggarwal corrected him: “CPM is simply 10,000,000,000 divided by thousand into $3.” Ten billion impressions divided by a thousand, times three to four dollars — which is exactly where the top of his range comes from. “So if you’re doing, like, about 10,000,000,000, you might be making a GMV of 20 to $30,000,000. Ballpark. Just ballpark.”
- The take rate. “So we work on an entirely revenue share based model. Anywhere between eight to 12% is what we charge of the entire ad revenue that the developer makes.”
Then Latka read the chain back to him: “you’ve got 1,400 apps that have used you across 320 unique developers over the past twelve months. Basically, last year, you tested 10,000,000,000 ad impressions generating somewhere around $20,000,000 of GMV for these 320 unique developers of which you keep eight to 10% of that 20,000,000.” Aggarwal’s reply was two words: “Bang on.”
Appbroda revenue = developer ad revenue × a take rate of 8–12%Only the second term belongs to Appbroda. The first belongs to 320 game developers.
Two things are worth noticing about that summary. The take rate slipped from the 8–12% Aggarwal stated to the 8–10% Latka repeated, and the developer revenue was taken at the bottom of the ballpark rather than the top. Run the band yourself and the spread is wide: 8% of $20 million is $1.6 million, 12% is $2.4 million. When Latka rounded it to “which is about a 2,000,000 run rate” and asked where the company had been a year earlier, Aggarwal did not argue with the frame. He answered it: “I think one year ago, we were at half of this.”
The $30 million is not Appbroda’s revenue
This is the trap the episode title sets, and it is worth being blunt about: the $30 million figure is ad revenue earned by game developers, most of it paid out by Google, Meta, Iron Source and AppLovin, and almost all of it stays with the developers. It is also not a measured number — it is Aggarwal’s own back-of-envelope CPM calculation, offered twice as a ballpark. What Appbroda invoices is the slice.
GetLatka’s own rows for Appbroda track the same shape and give the doubling a spine: $500K for 2021, flagged as an estimate; $1M recorded in June 2022 and again in November 2022; $2M recorded in January 2023, which the database does not flag as an estimate. Aggarwal’s memory of the first year matched: asked what total revenue was in 2021, he said “whirlwind, I think it was, like, less than 500,000.” The company had launched that June, in the middle of the pandemic, and was — in his words — “just about twenty months old” when this conversation was recorded.
The 320 customers behind that line are developers, not games. “We’ve had 1,400 apps and games make money more money than they were before through our platform. But that’s split across about 320 unique developers,” he said, “because one developer makes multiple games and multiple apps.” GetLatka records the same 320 as of January 2023.
Why two million developers need an ad-ops middleman
A pure revenue-share model only works if there is a large pool of people whose ad revenue is being left on the table. Aggarwal, who spent five years at Google across Play and AdMob before starting the company, argued that the pool is enormous and structurally ignored.
Roughly 2,000,000 developers globally fund content creation through advertising, and, in his telling, “there’s a very large knowledge gap on how best to utilize the tools in this industry to maximize your revenue from your app.”
Google, Meta, Iron Source and AppLovin “only have sales or service teams that cater to the top two or 3% of developers.” Everyone else self-serves a system nobody explained to them.
Founder, on tape“It’s not that 80% of the revenue is controlled by 20% of the developers. It’s more of forty forty twenty. 40% of the revenue is controlled by the top, but the remaining 60% is split between small and medium sized developers.”
Appbroda does not try to replace the ad networks. “We are building on top of the marketplaces. We are not disrupting the marketplace itself,” Aggarwal said. The developer pays, not the network — though the networks approve: Appbroda had a partnership with Google and was trying to build ones with Iron Source and AppLovin, “because the more we help their end customer optimize, they make more money.”
The five numbers Appbroda gets paid to move
Asked to name the metrics his team optimizes, Aggarwal listed them in the order money flows through an ad stack.
- Ad requests — “that’s totally dependent on your traffic.”
- Match rate — “how many requests you send, but how much did you get filled in your inventory?”
- Show rate — how many ads actually got displayed. “Most people don’t realize there’s a big drop, like 40% between ads you show and ads you fill.”
- CPC — “how much is the advertiser actually paying you for a click on the ad?” Blended, he put it at roughly $1.15 to $1.16 against that $3–4 CPM.
- Clicks — the volume the other four feed into.
Every one of those is a line on someone else’s dashboard. Appbroda’s revenue is a derivative of all five.
The ten-day clock
A rev-share deal has no seat price to defend and no annual contract to renew, which makes the first fortnight the entire commercial relationship. Aggarwal described a customer with no patience for onboarding.
Our customer is a very impatient customer. You need to increase their revenue in the first fourteen days. If you don’t do it, the customer is gone. He or she is gone with someone else.
Ashish Aggarwal, Appbroda
He then gave the deadline twice, tighter the second time, and attached a number to it: integration had to take under thirty minutes and was already down to five, and “the developer sees at least 10% revenue increment in the first ten days of doing work with us. Otherwise, we’re out of the business. We have no chance to say.” That is an unusually falsifiable activation bar. It also explains why churn in this model is quiet: nobody cancels a subscription, the ad revenue simply stops flowing through you.
Service revenue turning into product revenue
The other reason to keep the two pots of money apart is that Appbroda’s own pot was changing character. Asked for the split between service and SaaS, Aggarwal gave two different answers for two different windows.
“I think the last twelve months, it’s due towards service” — the period the $2M run rate was calculated over was majority services work.
“But in the last six months, it’s 80% product and 20% service.” The tech investment, he said, had absorbed the previous half-year.
None of it was funded. “We’re bootstrapped. We haven’t placed any round of funding,” he said early on, and later framed it as a hiring filter: “it’s a natural question when you join a startup, are you guys raising funding? Have you raised funding? Do you plan to raise funding? And the answer is no. We think that, you know, as a SaaS company, we can be profitable. We’ve done the math.” The company was profitable from day one, he said, headquartered in Dubai with him working from Delhi and the team fully remote — about 52 full-time people, roughly 20% of them in tech, three or four in product, another 15% in business development.
Fifty-two people against a roughly $2 million run rate is the tension in his stated goal, which Latka spotted in his bio: $1 million of revenue per employee by 2026.
20–50xhow far from $1M revenue per employee Aggarwal put the company, in his own estimate
A bad year for ad spend is the pitch
Recorded in the opening weeks of 2023, with ad budgets contracting, Aggarwal made the case that a percentage-of-upside model sells better in a downturn than in a boom.
This year is gonna be bad for ad spends in general, macroeconomic factors. But we’re particularly excited about this year because publishers or app developers will struggle with making more revenue, which is why companies like ours come in and kind of say, okay, even if we can help you make 10% more, we’re extremely valuable in recessionary climate.
The customer acquisition behind the doubling was not sophisticated. “We’re being very hacky. We’ve used a lot of email marketing. We’ve used a lot of LinkedIn. We’ve used a lot of tools for free on the net,” he said — Chrome extensions that surface a decision-maker’s email, SignalHire, LGM, then a direct pitch. LinkedIn was also his answer when Latka asked for his favourite tool for building the company.
Latka closes the same way every time: something you wish you knew when you were 20.
Fall forward. Great quote from Denzel Washington. Right? I think persistence is is important.
Ashish Aggarwal, Appbroda
Sources Nathan Latka’s interview with Ashish Aggarwal of Appbroda, recorded late January 2023; GetLatka company data for Appbroda — revenue rows for 2021 (estimate), June and November 2022 and January 2023, and 320 customers recorded January 2023.
