Pay ₹3 up front, then ₹499 every month: the auto-debit business behind India's "voice lock screen" app

Hey Buddy is an Android tool that locks the screen when you say a phrase. It launched in 2026 October 6 and got 264k installs in a month. The product has almost no technical substance. What's worth looking at is how it takes payment: after install you pay ₹3 for a two-day trial and sign a UPI autopay mandate, which then charges ₹499 every month. This model didn't originate here. The leader, Vokey, already has 5.54M installs. Hey Buddy is a follower that arrived a few weeks late in this wave, and behind it is a Bangalore app-portfolio company that buys user acquisition.

A review from Oct 10

Four days after launch, a user named Sujay Banerjee left a one-star review on Google Play:

"It's a scam!! Don't install… 1st they will charge you 3 rupees during setup for premium access and after the payment you won't be able to record your voice command for lock/unlock of phone!!! RETURN OUR MONEY YOU FRAUDS"

82 people marked it as helpful (Play Store India page, 11-06), and it's still the most prominent review under the app. Other five-star reviews on the same page read like store blurbs: "Overall, Hey Buddy is a fantastic and innovative app… A must-have for every Android user".

The ₹3 this one-star review mentions is the entry point to the whole business.

What it is: a floating overlay on your screen

The store listing is packed: voice-open apps, make and answer calls, play music, lock the screen, "hands-free while cooking, driving, or working out," plus a long string of keywords like voice launcher, voice dialer, and call assistant. But opening apps, calling, and playing music are already done by Google Assistant, Gemini, and Android's built-in Voice Access. The one thing the system doesn't do is "lock the phone when you say a phrase."

Its website heybuddy.pro says only this: "Say the word. Your phone locks." You record a phrase only you know (two or three words), and it puts a lock overlay over the current app; say another phrase, or enter a four-digit backup PIN, and the lock is removed. The phrase can be spoken in 9 languages including English, Hindi, Telugu, and Tamil, using the phone's built-in speech recognition.

The terms of service are more candid: it "is a convenience feature, not a security product, and does not replace your phone's own lock screen"; the notification shade still pulls down; and "after a phone restart, the lock won't listen until you open the app once." If you forget the backup PIN, a restart clears the lock. In short, it's a floating window plus system speech recognition, which could be built in a few days.

Where the money comes from: ₹3, two days, ₹499

The store page labels it "free, no in-app purchases, no ads," and the Data safety section says "no data collected, not shared with third parties." Money isn't collected through Play. The bottom line of the website homepage reads:

₹3 for a 2-day trial, then ₹499/month by UPI Autopay. Cancel any time.

The terms of service say verbatim: ₹3 "authorises a UPI Autopay mandate and opens a 2-day trial; after the trial ₹499 is auto-debited every month until you cancel. Payments are processed by Cashfree Payments."

This design serves several purposes:

• Bypassing the Play cut: money flows through Cashfree, an Indian local payment company, so Google's 15–30% fee is avoided.
• High price: ₹499 is roughly $6 a month, which is very expensive for a utility app in India.
• Room to scale paid acquisition: Meta installs in India are cheap, so if only 1–2% of new users get debited for a month or two, the ad spend can be recovered (inference; actual conversion rates aren't public).

Same funnel everyone in this category uses. Vokey runs a ₹1 trial that converts to ₹499/month. A Play review says: "The app initially asked me to pay ₹1, but ₹499 was subsequently deducted from my account without my clear knowledge or consent" (10-28). Another says ₹499 was charged and they cancelled the auto-debit within an hour or two: "my 499 not back why?" (10-31). For TalkKey, someone has already made a YouTube tutorial on "how to turn off TalkKey auto-debit."

Where the users come from: Hindi-language ads

Hey Buddy isn't on X, Reddit, or in the media. Its volume comes from paid ads, and there's evidence in several places:

• Its privacy policy says it sends install, signup, and subscription events plus the ad ID back to Meta "so we can measure our advertisements," and records which campaign, ad set, and ad each user came from. It also pulls install sources from Play. That contradicts the store listing's "not shared with third parties" line.
• The developer, SPACEWIRE TECH, is hiring on LinkedIn for someone to "manage app growth across multiple apps with Meta Ads and Google Ads," on Jooble for a Hindi ad-script writer, and on Indeed for an editor who makes AI videos for digital ads.
• Its Facebook page, heybuddy.live, posts videos titled in Hindi: "बस बोलकर करें Phone Lock & Unlock!" (Lock and unlock your phone just by speaking!).

A few Hindi "Hey Buddy kaise use kare" tutorials on YouTube have only tens to about a thousand views each. They aren't a major source.

The install curve also looks paid (AppstoreSpy estimates; don't trust single-day numbers): cumulative installs were 27k on Oct 20, adding one to two thousand a day the first week; around Oct 21 it jumped by about 25k in one day, then ran 16k to 28k a day for five days, then 10-26 to 157k; after that it came in bursts, 11-02 to 246k, then 11-06 to 264k (Google Play, 11-06). It only charted in India's Productivity Top Free, peaking at #50 (10-29) and falling to #69 (11-06). It doesn't look like word of mouth building slowly; it looks like the ad budget switching on and off.

It could scale this fast because the company has done this before. SPACEWIRE TECH LLP was incorporated in Bangalore in October 2025. Its website lists five apps aimed at mass users in India (astrology, religion, UPI, music, location). One of them, the ringtone and wallpaper app Bhaktilok, has 5.67M installs. The media buyers, Hindi script writers, and AI video team were already in place. Once they saw Vokey work, they shipped their own version within weeks.

Who it's competing with: fourth in a crowded category

"Voice lock" in India isn't a blue ocean; it's a group of apps copying each other. Here are others using the same UPI auto-debit model (Google Play installs, 11-06):

Several more Voice Lock apps differ by only a name. Their "Payment & Cancellation" pages are easy to find online.

The other group is long-running global ad-and-in-app-purchase apps: VoiceOS (15M), o16i's Voice Screen Lock (listed 2022 years ago, 13.2M), and One byte Studio's AI Voice Lock Screen (11.2M). New ones keep arriving every October.

Compared with Hey Buddy and similar apps, the features and prices are nearly identical (AwaazPro is even cheaper). Multilingual commands, showing what was recognized before saving, and a long list of voice-assistant keywords mostly exist to capture more store search terms. It didn't break out because it's better. It broke out because it charges and buys ads.

How long this funnel can last

"₹1 trial turns into a monthly charge" is already an open secret in India and is being reported as a scam. Business Standard (July), Mint's Hindi edition (August), and Moneylife (September) have all covered it, and the Telangana cyber police issued a warning in May. Hey Buddy's most prominent reviews call it a "scam."

This model makes money because users don't notice they've signed up for auto-debit. If Google Play tightens its rules on misleading subscriptions, or NPCI tightens the rules on Autopay mandates, the whole category gets hit together. Its revenue, paid conversion, and refund rate aren't public. (Inference: with this many copycats, each one's profit is shrinking.)

If it were you

Grade C. You can learn from it, but I wouldn't build the Nth voice lock app. This category in India already has seven or eight apps running the same model, and the leader has 5.54M installs. The model itself has been named by media and cyber police, and the window could close on a policy change at any time.

Who it fits: Agencies already running mass-market apps in India that have Meta/Google media buyers and a Hindi creative team. What you'd be learning isn't "voice lock," it's the monetization.

Who it doesn't fit: Indie devs who haven't run ads (there's no technical moat here, it's all about ad spend). Anyone who doesn't want their name next to "scam" in the reviews. Anyone planning to copy this outside India (UPI auto-debit is India-only, and the equivalents in Europe and the US are already taken by large, long-running ad spenders).

How to do it: If you want to borrow this playbook, pick a lightweight tool nobody has built yet that can be shown in one action (one gesture, one voice command, one visual that explains it). Run Hindi short videos on Meta, and collect payment outside the app through Cashfree or Razorpay with UPI Autopay. Send subscription events back to Meta so it optimizes for paying users. Show the trial price and monthly fee on the first screen before payment, and put the cancel option somewhere obvious. Conversion will drop somewhat, but you'll avoid the "scam" reviews and be better protected if policy tightens (inference).

How to acquire users: Meta and Google app ads, vertical Hindi and regional-language creatives, with the effect shown in the first second (say a command, the screen locks). Wait until a category leader has proven the model and search terms are heating up, then enter, when ad buys are cheapest.

When to move: As early as possible, and treat it as a short-term business. This time, the category went from one leader working to crowded in just a few weeks.

What would show this read is wrong: (1) If Vokey, TalkKey, and the others were still growing weekly in early 2027, and neither Play nor NPCI took action, then this funnel is more durable than it looks and the window is wider than estimated. (2) Conversely, if Play starts delisting these apps, or NPCI requires Autopay to be reconfirmed before the trial ends, then even the "borrow the playbook" route has to be recalculated.