activation rate for driver apps, and why it means a first order

what activation rate measures in a driver app

signups are easy to count and easy to celebrate. in a partner app they tell you even less than usual, because signing up costs a driver nothing. they type a phone number, enter an OTP, pick a city. it says nothing about whether they'll ever pick up a parcel.

activation is the step where the product does what it exists for. for a project tool that's creating a first project. for a driver app it's the first order, picked up and delivered, with money owed to the driver at the end of it. before that moment the driver is still deciding whether your app is worth their time. after it, they've earned something, and they know they can do it again.

activation rate is the share of new drivers who reach that moment. it's the closest thing supply onboarding has to a scoreboard, because everything upstream, the document screens, the approval wait, the training videos, exists to get people there.

why the event should be a first completed order

the formula is trivial. the definition is where teams go wrong, and in driver apps they almost always go wrong in the same direction. they pick something too shallow.

"documents uploaded" and "approved" are the usual suspects. they're easy to track, they live in your own system, and they go up when the onboarding team works harder. but an approved driver who never goes online is a cost with no revenue. if your activation rate rises because you sped up approvals, you've learned something about approvals. you haven't learned whether more people are delivering.

Shadowfax draws the line in the right place. its UDRHP only counts a partner once they've "completed at least one delivery". that's the definition we'd use too. it's the earliest event that means the whole chain worked, from documents to a customer getting their order. it's also not too deep. a motivated new driver can reach it in their first few days, so it arrives early enough to act on, unlike thirty day retention.

if you want a stricter version later, add a count or a time limit, like "three completed orders in the first fourteen days". start with one, though. one completed order is the moment the driver stops being a signup.

how to calculate activation rate

once the event is defined, the arithmetic is one line:

the window matters as much as the event. pick one, seven or fourteen days after signup are common, and keep it fixed. calculate it by signup cohort, meaning everyone who signed up in a given week, rather than dividing this week's first orders by this week's signups. a driver who signed up a month ago and delivered today belongs to last month's cohort.

to check the arithmetic: if 1,000 drivers sign up in a week and 300 of them complete a first order within fourteen days, the rate is 30%. that number on its own isn't very useful. the stages in between are where the information is.

split it by city and by the driver's language and the leak usually becomes obvious. a stage that loses most of your Hindi speaking drivers but few of the English ones is telling you something specific.

what is a good activation rate for a driver app?

people want a number. we don't have an honest one for driver apps, and we'd be suspicious of anyone who does. the companies with the data don't publish a signup to first order rate, and every company draws the funnel differently.

the SaaS benchmarks you'll find don't transfer. Userpilot, for example, puts average SaaS activation around 37.5% across its customers. that's for software trials, where a user creates a document. a driver has to get verified, collect a bag and ride across town. it's a different kind of event.

what you can say is that the stakes are high. one vendor, Caller Digital, puts onboarding a rider at ₹1,600 to 2,400 by mid 2026, and the same vendors say half a new cohort stops logging in by day 88 to 110. every driver who gets through approval and never delivers is most of that cost with nothing to show for it. so the benchmark that matters is your own number from last month. keep the definition fixed, and if the rate goes up, you're doing something right.

how to raise it

once you can see the stages between signup and first order, improving the rate mostly comes down to three kinds of work.

the first is shortening the path. Caller Digital counts 12 to 17 steps from download to first delivery, over 72 to 96 hours in a typical case. question each one. does the driver need a referral code to start? can training happen while documents are being checked instead of after? every step you remove is one fewer place to lose someone, and the early ones cost the most, because that's when a driver is most likely to try a competitor's app instead.

the second is fixing the walls that no explanation can fix. a verification queue that takes days, a joining fee, a deposit, a slot that needs a team lead's approval, a zone with no orders. if your funnel leaks there, no amount of guidance will help, and it would be dishonest to pretend otherwise. fix the operation, or at least tell the driver plainly what's happening and how long it will take.

the third is helping at the moment of need. most drivers who stall aren't unwilling. they're on a specific screen, often the first live order, and don't know the next tap. a Borzo survey from 2023 found 82% of couriers wanted a local language inside the app, while a VVGNLI study and IDinsight found most drivers are schooled and already owned smartphones before joining. so the problem is language and clarity. this is what heytappr is for. it's a voice guide inside your Android partner app that highlights where to tap and talks the driver through it in Hindi first, then regional languages, until they finish a first order.

it also measures the thing this whole post is about. heytappr gives your ops team a funnel from signup to first job and shows the exact step where drivers quit. guides are edited from a dashboard and go live without an app release, so when you find the leaking step you can fix it the same day. it ships as an Android SDK. for the wider picture of where drivers drop off, see our guide to supply-side marketplace onboarding.

whichever you start with, change one thing at a time and watch the stage it was meant to affect as well as the final rate. it's the only way to know which change did the work.