0 to 1.5M Installs: The ByteDance Playbook

When I joined ByteDance to work on SnapSolve, the product brief was straightforward: grow the app. The constraint that made it interesting was the category — doubt-solving for students, where the quality of the answer determines everything.

You cannot run acquisition campaigns for a product that doesn’t work yet. And for an edtech app, “working” means having educators who can answer every subject, at every difficulty level, within minutes. That is a supply problem, not a demand problem.

What followed was one of the clearest examples I have seen of why the order of operations in app growth matters more than the channels you choose.

Most app growth fails not because of the wrong channels — but because acquisition ran before the product was ready to keep anyone.

---

The standard playbook — and why it fails for trust products

Most app growth frameworks assume the product is ready and the job is to drive installs. Run Meta campaigns, optimise your App Store listing, set up referral loops, measure CPI and D7 retention. Rinse. Repeat.

This works for categories where the product works on first use: games, utility apps, content platforms. A user downloads a casual game and immediately knows if it’s fun. The trust threshold is low. The feedback loop is fast.

It fails for categories where value is conditional:

• An edtech app is only valuable if it has the right educators
• A fintech app is only useful if the credit or insurance product fits your profile
• A marketplace is only worth opening if supply is deep enough

SnapSolve fell squarely in the first category. A student opens the app, photographs a chemistry problem, and submits it. If no educator is online to answer that subject, the student closes the app and it never opens again. The product’s value was entirely dependent on supply depth.

Running paid acquisition before supply depth was built would have produced installs that churned immediately — destroying unit economics, killing our App Store rating, and burning the audience we needed most.

---

The decision: build supply before demand

The strategic choice was to flip the standard order. Build educator supply first. Then turn on acquisition.

This sounds obvious in retrospect. It was not obvious at the time — the pressure to show install numbers is real, and supply-building is slow, invisible work that produces nothing you can put in a weekly report.

The supply-building phase looked like this:

Educator recruitment at scale. We needed educators across subjects — mathematics, physics, chemistry, biology, history, economics — and across exam tracks: JEE, NEET, CBSE, ICSE, state boards. We built a dedicated educator acquisition team and treated educators like a user segment with their own onboarding funnel, quality tiers, and retention mechanics.

Quality control architecture. Not all supply is equal. A 13-year-old student submitting a physics problem at 11pm needs a correct answer from a credentialled educator, not a peer crowdsourcing guess. We built a verification layer — credential checks, answer quality scoring, subject-specific testing — that took time but produced a supply base users could actually trust.

Educator incentive design. Educators needed to be available, not just registered. This required building incentive structures that rewarded response time and answer quality, not just volume. Getting this right took multiple iterations.

By the time we turned on demand acquisition, we had 3,000+ educators across subjects and geographies. The product worked. When a student opened the app and submitted a problem, it got answered.

---

The distribution insight: supply as acquisition channel

Here is the part most app growth frameworks miss entirely: the supply base became the most effective distribution channel.

Each educator had their own students — tutorial classes, YouTube channels, WhatsApp groups, Telegram communities. When an educator joined SnapSolve, they did not just provide supply. They brought their audience.

We made this explicit. Educator referral mechanics, co-marketing content (educators sharing the app with their students), and feature-flagging educators as named contributors to their students (so students could follow specific educators they trusted) all turned the supply network into a distribution network.

The result: a significant portion of early installs came through educator-referred channels — effectively zero-CAC installs from users who arrived with pre-existing trust in the educator they’d followed.

This is not a unique insight to edtech. It applies to any category with a supply side:

• Marketplace apps: seller network drives buyer acquisition
• Fintech apps with agents: agent network drives customer acquisition
• Health apps with practitioners: practitioner network drives patient acquisition

The supply side, if treated as a distribution partner rather than just a cost centre, dramatically compresses acquisition cost.

---

The performance layer: turning on demand

Once supply depth was established, we ran a multi-channel demand strategy.

Performance acquisition: Meta and Google App campaigns with creative optimised for the student audience. Pain-point creative (exam stress, last-minute doubt clearing) performed significantly better than product-feature creative. The audience was 14–22 year olds in exam preparation cycles — targeting around exam dates (JEE, NEET, board exam calendars) produced significantly lower CPIs than evergreen campaigns.

SEO: Subject-specific content targeting high-volume student queries. “How to solve [type of problem]”, “NCERT solution for [chapter]”, “[exam] preparation tips”. This produced organic installs with high intent and materially better retention than paid channels.

WhatsApp for nurture: Post-install activation and retention. Students who opted in to WhatsApp received personalised content tied to their subject preferences and exam tracks. Reply rates on these messages were high — the product was solving a real, recurring problem (exam preparation), and students were in active need.

OEM partnerships: For Tier 2/3 reach, OEM deals with Android manufacturers provided distribution at device activation. Students in smaller cities getting their first smartphone had SnapSolve available on first boot — a channel that performance marketers often overlook because the contracts require relationship management rather than self-serve ad buying.

---

The 0 to 1M timeline

• Months 1–4: Supply building phase. Educator recruitment, quality architecture, incentive design. Zero paid acquisition. This is the phase most founders are afraid to stay in long enough.
• Months 5–6: Soft launch with performance acquisition at low budget. Watching activation and Day 7 retention carefully to confirm the product was working before scaling spend.
• Months 7–8: Acquisition scale-up. Meta, Google, educator-referred channels all running simultaneously. Organic and SEO starting to contribute meaningfully. 1 million installs.
• Months 8–24: Continued scaling to 1.5M installs. Retention optimisation, new subject verticals, regional language expansion.

---

The transferable framework

The SnapSolve growth model has a name I’ve used since: supply-first, demand-second.

It applies whenever:

1. The product’s core value depends on supply depth (content, educators, lenders, sellers, practitioners)
2. The target audience has high trust requirements before downloading (education, finance, health)
3. Retention is a function of product quality, not just acquisition channel

The framework:

Phase 1 — Supply architecture: Build the supply base before running demand campaigns. Instrument quality, not just volume. Design incentives for the supply side that align with user outcomes.

Phase 2 — Supply as distribution: Identify which supply-side participants have their own audiences. Build mechanics that turn them into distribution partners: referral programs, co-marketing content, attribution for educator/seller/practitioner-driven installs.

Phase 3 — Demand activation: Start performance acquisition only after supply depth is verified. This is when standard app growth playbooks apply — but the CAC, conversion, and retention numbers will be materially better because the product works.

Phase 4 — Organic compounding: SEO, word-of-mouth, and app store optimisation compound over time. Invest in these from Phase 1, but do not expect returns until Phase 3–4.

---

What I would do differently

One thing: start the SEO infrastructure earlier. We treated organic search as a secondary channel and prioritised paid acquisition. In retrospect, the subject-specific content we eventually built for SEO purposes would have been more valuable starting in Phase 1 — it would have driven educator discovery as well as student acquisition.

The second thing: measure supply-side retention, not just demand-side. We tracked educator churn but didn’t have the same rigour on supply-side cohort analysis that we applied to user cohorts. Supply churn is a leading indicator of product quality degradation — knowing that earlier would have allowed faster responses to specific subject or geography supply gaps.

---

Chandan Kumar is a full-stack growth marketer and founder of Grovio Labs. He writes the India Lens series on growth strategy for the Indian market. Related: Growth Marketing in India: What Western Playbooks Get Wrong · What Breaks at ₹10 Crore.