Almost every job-search tool wants to be the next great app. Polished onboarding. A dashboard with twelve tabs. Push notifications you'll mute within a week. The pitch is always the same: "the modern job hunt deserves a modern home."
We went the other way. JobSwitch lives inside WhatsApp. No app store download. No dashboard. No onboarding wizard. You message a number and a conversation starts.
On its face it sounds like the lazy option. It isn't. It's the harder decision technically, the easier decision for the user, and it changes how the product feels in a way you can't undo by adding a "WhatsApp integration" later.
The app tax is a real tax
Pick any job-search product with a great web app. Watch a candidate try to use it on a Tuesday evening. They sign up, they upload a resume, they answer twenty preference questions, and then they go to bed and never log in again. The work to start outweighed the benefit of continuing.
Every user has a finite amount of patience for installing yet another app they didn't ask for. The market for new apps in 2026 is not great. The market for a chat thread you already check forty times a day is, well, the best market that has ever existed.
WhatsApp is the only interface most of our users already trust
In India, WhatsApp isn't a chat app. It's the operating system for conversations with humans you actually care about — your family, your recruiter cousin, your college group. People treat their WhatsApp inbox the way Americans treat their primary email: with a baseline of trust that unknown apps don't get.
When we show up there, we inherit some of that trust. We also inherit some of the responsibility — a wrong move on WhatsApp feels far more personal than the same wrong move inside Yet Another Job App. That's a fair trade.
The interface that fits "say yes to a job" is a sentence
The core JobSwitch interaction — "here's a role I think fits you. Apply? Say yes or no." — is a one-sentence interaction. Building a dashboard for a one-sentence interaction is over-engineering. Building it as a message is obvious.
We have richer flows (asking for preference updates, surfacing application status, queueing follow-ups). They all decompose into messages too. The dashboard isn't load-bearing for any of them.
What we gave up
This wasn't free. WhatsApp's API has constraints:
- Meta's Cloud API rate-limits and template-message rules force us to be careful about how proactive the agent can be.
- We can't show a beautiful application timeline view. We render it as messages with the occasional rich card.
- File uploads (your resume) and downloads (the tailored copy we generated) are clumsier than they'd be in a custom interface.
The web app on the roadmap exists to give power users a calmer overview when they want one. But WhatsApp is and will remain the primary surface. That's a product decision, not a transitional state.
The unfair advantage of being where people are
The most interesting thing about the WhatsApp choice has been retention. Users who haven't messaged us in a week aren't churned — they're just waiting for the next role we surface. There's no "should I open the app today" decision, because there's no app to open. The thread is right there, next to the one with their mom and the one with their best friend.
That's hard to compete with. We aren't trying to win the attention war. We decided not to enter it.