247leads over the period
÷3cost per lead, from launch to the final stage
−68%cost per click
Client and goal
An online service for crypto holders that pulls data from crypto wallets, calculates tax and prepares the reporting documents. At the start there was no product yet, only a landing page and a hypothesis. The goal: measure real audience interest and collect sign-ups to validate the business model before investing in an MVP.
What I did
- Channel choice. An international audience meant Google Ads, with search ads on the most relevant queries.
- Stage 1. Launch on a narrow core of the most commercial searches to get clean first data on demand.
- Stage 2. Expand to adjacent searches to grow reach.
- Stages 3–4. Weekly negative keyword work and bid adjustments. This is where the cost per lead dropped sharply.
Progress by stage
The account was billed in Russian roubles (RUB); what matters here is the relative change.
| Stage | CPC, RUB | Leads | CPL, RUB |
|---|---|---|---|
| Launch (week 1) | 17.46 | 4 | 406 |
| Keyword expansion | 14.99 | 52 | 432 |
| Fine-tuning | 7.30 | 59 | 208 |
| Further optimisation | 6.29–6.77 | 101 | 196–210 |
| Final stage (17 days) | 5.65 | 31 | 188 |
Totals for the period: 30,250 impressions, 7,451 clicks and 247 leads.
Takeaway
This case isn't about leads for the sake of leads; it's about validating a business model before spending on development. 247 sign-ups proved the market needs the product. The threefold drop in cost per lead came from discipline, not a one-off setup: the main lever was weekly negative keyword work. Search ads are an ideal way to test demand for a new product, because they measure interest in money rather than guesses.