How a two-person team built a $2M ARR AI job application app

Executive overview

Most consumer apps stall because they chase views rather than conversions. Dan built Massive — an AI that auto-applies to jobs — to $2M ARR with a tiny team by obsessing over audience targeting and marketing distribution.

Waitlists are a trap: they signal interest but almost never convert. Get people to pay from day one.

The core insight: hitting the right audience matters more than volume — a video with 11M views can under-convert while the same format targeting a tighter demographic drives exponential revenue.

From zero to first revenue

• Viral LinkedIn post (8M impressions, 5K comments) validated the idea before building
• Two months of building + waitlist generated 40–50K signups — but under 0.5% converted
• First paying users came from manual outreach: scraped Hacker News posts, sent personal messages
• First consumer subscription felt like proof; the waitlist itself felt like wasted time
• Lesson: skip waitlists; use influencers to test paid conversion immediately

Influencer and UGC strategy

• Start with a handful of influencers (spend $1–2K max) to test whether a message converts before scaling
• Proven distribution beats clever creative — use channels with consistent views across videos
• 80% of content should replicate formats already proven to go viral; 20% for experiments
• Iterate relentlessly on one format before moving on — weeks of refinement separate 382 views from 1.7M
• UGC scales poorly for niche products; paid ads give faster conversion data and clearer signals

Audience targeting is everything

• The hook determines the audience — "stop using LinkedIn" converts white-collar; "stop using Indeed" attracts blue-collar (who won't pay)
• Wrong audience produces high views and near-zero revenue; right audience produces the inverse
• Organic platforms optimise for views, not conversions — you must build your own signal early, at low volume
• Skits underperform for Massive because they don't call out the specific problem clearly enough
• Study comments to diagnose audience mismatch fast

What actually drives conversion

• Lead with the problem, not the product — surface pain the viewer didn't know they had
• Great creative overcomes bad UX: a product with a 2010-era checkout still converts if the ad is compelling
• Retargeting sequences work: multiple consistent ads that progressively identify a pain point outperform single-touch
• "Rip" proven onboarding and ad structures — marginal optimisation rarely beats copying a format that already converts
• Paid ads outperform UGC for products with niche audiences because you can optimise explicitly for conversion events

Market and audience selection

• Targeting a slightly older or higher-income demographic can triple revenue at one-fifth the download volume (e.g. Turbo Learn vs. younger note-taking apps)
• Headway ($200–300M ARR) built a simple book-summary app through pure paid-ads arbitrage — no novel product insight
• Small demographic or positioning shifts can produce step-change revenue jumps (Masterschool example: 3x revenue in one quarter)
• Always ask: who will pay the most for this? Build toward that audience, not just the largest one

Shady tactics to avoid

• Fake UGC (fabricated job offers, stolen creator videos, comment-farming): definitely illegal, FTC risk, destroys brand trust
• Copying viral videos word-for-word and claiming real outcomes violates FTC rules and poisons the broader ecosystem
• Grey-hat tactics tend to become embedded in company culture — avoid them regardless of short-term upside