Case study: how I use vibe-coding for marketing as a designer

I'm building a tool that helps new businesses create their first basic brand. I wanted to run an experiment for an underused customer acquisition channel – physical mail.

I assumed companies are most likely to need branding right after incorporation. So I wanted to reach out to freshly incorporated tech companies with personalised logo options.

1. Get a list of companies — Go to UK Companies House (they also have an API).

2. Filter for the right companies — Search for tech companies (SIC code: 62012) incorporated in the last 14 days and export their addresses.

3. Create logos — Vibe-code a logo generator using their company name (this already existed in my product), and export as PDF.

4. Send letters — Upload the PDFs to a service like Pingen to print and ship them (~$2 per letter in the UK).

The point of this post is to show how I used vibe-coding not only for building my core product but also in marketing.

But if you’re curious about the results of the experiment itself, it failed on the numbers. We offer branding for $19, so customer acquisition cost can’t be higher than that.

• I sent 30 letters; 4 were undeliverable.

• I could only track the QR code—it got 1 scan.

• At least one recipient completed the full funnel but abandoned at payment.

However, if I were an agency looking for projects and could afford a higher CAC, I would continue experimenting with this channel. For example:

• Geography — I targeted UK businesses. As an agency, I’d try to go local. Other countries may also have better open rates (I’d expect close to 100% in Germany), and printing and shipping costs are lower there ($1.6).

• Business types—I targeted only tech companies, I’d test other industries.