Onfleet Revenue in 2020: About $800K a Month, Billed Per Task, 0% of the Order

Onfleet’s CEO never offered a revenue figure himself, but he confirmed one. On the August 25, 2020 recording, Nathan Latka multiplied Onfleet’s roughly 900 customers by their average bill of about $900 a month and asked Khaled Naim: “You guys are somewhere around $800,000 a month in revenue?” Naim said: “Yeah. That’s right.” Over twelve months that comes to about $9.6 million. That multiplication is ours; Latka’s own sign-off called it “almost a $9 to $10,000,000 run rate.” What Naim added himself was a forecast: “This year, we’re definitely on track to double... Probably more than double revenue this year.”

That money comes from software fees, not from a cut of what gets delivered. In the month before the interview, Onfleet’s customers ran more than 3 million delivery tasks through its software, using their own drivers or couriers they had hired. Onfleet charged them by feature set and task volume. “So we don’t take a margin,” Naim said.

Onfleet revenue in 2020: whose number is $800K?

The $800,000 is Latka’s multiplication and Naim’s confirmation, and the tape goes no further than that. Both inputs came from Naim. He said “we’re close to 900” customers, and that the average bill was “closer to 900 now” a month, up from the roughly $800 Latka remembered from their September 2019 conversation. Strictly, 900 times $900 is $810,000, so Latka rounded down.

The year-earlier comparison is mostly Latka’s too. He asked whether $800,000 was “double where you were twelve months ago about.” Naim said “Yeah. Yeah” and went straight to his forecast for the year. In his sign-off, Latka put the year-earlier figure at “400,000 just a year ago.” Naim never said that number. It does match Latka’s own notes, though: the “about 500 ish customers” and roughly $800 average bill he cited from September 2019 multiply to $400,000.

We aren’t adding anything from Onfleet’s GetLatka profile. An earlier version of this article quoted a 2020 run-rate row from it. That row had no recorded source and has since been removed from the profile. The profile’s remaining rows for 2017 and 2019 still have no recorded source, and its latest figure is a GetLatka estimate, so none of them is used here.

What the 3 million actually is

An earlier version of this article said Onfleet “moves” three million deliveries a month. That makes it sound as if Onfleet does the delivering, and it doesn’t. Latka asked how many tasks had run “across the entire base” the month before, and Naim said: “We are north of 3,000,000 now.” Those tasks belong to Onfleet’s roughly 900 customers. The drivers doing them are, in his words, on the “customer’s payroll or they are contracted as a… third party courier.”

What Onfleet sells is the software that runs those drivers: “routing, dispatching, real time tracking, analytics, proof of delivery, customer facing communications.” When a retailer hands work to a courier company, Onfleet acts as “a collaboration tool or orchestration layer,” and the retailer still gets “all of the data, the visibility as if it were their own fleet.” The billing unit is simple: “when a driver starts a delivery, completes it, we say, okay, that’s one task.” Latka worked the monthly total down to “100,000 plus a day,” which is 3 million divided by 30.

Why Onfleet takes 0% of the order

So we don’t take a margin. So that’s the nice thing about Onfleet is we charge you based on the feature set that you need within the software. It’s a SaaS model.

Khaled Naim, CEO and co-founder, Onfleet, August 2020

Onfleet does charge by delivery volume, Naim said, “but we don’t care if you’re delivering something for $10 or $100.” Latka suggested Onfleet’s investors might want a percentage of each sale. Naim’s reason for not taking one was practical: “We’re not driving the sales, so we can’t really command a margin. And we don’t even know the value of the items that are going through.”

Each plan includes a set number of tasks in its base rate, and customers pay per task above that threshold. Across all customers the average was “about 20¢ a task.” The biggest accounts paid about 10¢ and the smallest a little over 20¢, and below the first tier customers pay a lot per task. Latka turned the average bill into volume: $900 at 20¢ is 4,500 tasks a month, “about a 150 deliveries a day.” Naim agreed, but warned that the spread was wide: “we have customers doing hundreds of thousands of deliveries a month and customers doing hundreds a month.”

Those round numbers don’t fully add up, and it’s worth saying so. Three million tasks at 20¢ is $600,000 a month, not $800,000. Latka’s per-customer math goes the other way: 900 customers at 4,500 tasks each is about 4 million tasks. The gap fits inside the rounding, since “north of 3,000,000” has no upper limit and the 20¢ is an average across tiered prices with base fees on top. Still, the tape never reconciles it.

Who uses it

Latka assumed Onfleet’s customers looked like Drizly, the alcohol delivery service Onfleet had recently announced as a customer. Naim corrected him. Many customers are traditional retailers with physical stores, such as Total Wine & More, which he called “the largest independent retailer of fine wines and spirits in the country.” Gap runs same-day delivery fleets on Onfleet in some international markets. Sweetgreen also works with Uber Eats, but Naim said many restaurants now use both. “Because of the margin profile of having their own drivers, especially if it’s a larger order, like a catering order,” it makes sense for them to run their own drivers through Onfleet.

According to Latka’s notes from earlier interviews, Onfleet had 300 customers in 2017 and about 500 in September 2019. Naim put the count at about 900 in August 2020.

From a Stanford address app to break-even

Onfleet launched in 2015, but it started in 2011 as a location-services project at Stanford aimed at emerging markets. The idea was to help people “that lack functional street addressing systems in their countries to communicate locations with delivery services.” The product, called Addy, let anyone create a URL for a physical location. The delivery companies the team met had no back-end technology (“pen and paper, chat apps, phone calls, text messages”), while Silicon Valley companies around them were managing drivers with smartphones. Onfleet was built to close that gap.

By August 2020 Onfleet had been at break-even for “a couple of years” and was “still breakeven,” Naim said. Revenue was growing faster than the company could hire. It had also turned down buyers: “we have been approached by a number of folks for M and A, but have decided to continue pushing forward.” Latka, citing an earlier conversation, put the equity raised at $5 million. Naim didn’t dispute it and added close to $1 million in debt from Lighter Capital. That debt came in tranches, starting as revenue-based financing and later becoming a fixed-term loan, “late twenty eighteen maybe.”

We wanted to get the business to a point where it could stand on its own two legs and we don’t need outside funding, but we are at the point now where it looks like there’s huge exciting market opportunity to go after, and we would raise some capital.

Khaled Naim, August 2020

He said Onfleet would be “announcing something in the near future,” and when Latka asked about size he said “probably in the 10 to 15,000,000 range.” He also downplayed how much the company needed: “We don’t need a ton of capital. We’re a super lean operation.” The tape doesn’t say whether or when a round closed. GetLatka’s funding rows for Onfleet name no source, so we don’t report one here.

Team, retention, and a correction on air

Onfleet had 40 people, of whom “about 15. 15 to 17” were engineers. Sales had roughly three SDRs and five AEs, “eight to 10” in all. Naim expected to reach 65 or 70 people by the end of 2021, many of them in sales, and was hiring for revenue operations for the first time: “a new position we’ve never hired before.” He admitted the sales structure was still a work in progress: “there’s still a lot to learn. We haven’t quite figured it out yet.”

The retention numbers needed a correction on air. Latka read out figures from his notes on their previous interview. Naim said the net number sounded right, that it had since gone up, and corrected the rest. Net revenue retention was “about 115, 116%” a year, and gross annual revenue retention was typically “around 85%.” Latka did the math: 15% gross churn plus 30% expansion gives 115% net retention, and that adds up. Naim warned against treating 2020 as a normal year. May and June brought “just crazy expansion” as people ordered groceries from home, churn was higher than usual, and “this year has been just across the board anomalous.”

Asked why he shares numbers at all, Naim said: “we don’t really have anything to hide. Our investors aren’t asking us to be super quiet about our numbers.” He added that Onfleet had been on the Inc. 5000, “so we’ve had to disclose our numbers.” He had just started reading Meditations by Marcus Aurelius, the CEO he follows is Bill Gates, and his favorite tool lately is to “have a good lawyer.” His advice to his 20-year-old self: “Invest in Amazon.”