Why we removed Demo uploads
We ran a demo feature for 6 months. It let users upload a PDF and see our conversion in action — no signup required. Basic startup advice. Show value, remove friction.
However, we just removed it. Here's why.
The Idea
The logic was simple: reduce friction. Let people experience the product before committing their email and money. They upload a document, see how it works, and hopefully think "I need this."
The Numbers
Over 6 months, we had 10,667 demo uploads. Not bad for organic traffic. But here's where it falls apart:
| What we measured | Result |
|---|---|
| Demo completion rate | 36.7% |
| Users who subscribed | ~147 |
Two out of three users never saw the output. They uploaded a file and left before processing finished.
Slightly more than 1% converted to paid subscribers.
What Went Wrong
1. Costs. Obviously
Every demo cost us money. To control costs, we only processed the first page of the file. But with so many demo uploads, we started to bleed money. We are a bootstrapped company without crazy VC money. So we have to be mindful of our expenses to survive.
2. The experience was too slow
Document processing takes time. Our OCR needs to analyze the PDF, we need to parse the response, build the tables. For a logged-in user who just paid, waiting 10-15 seconds feels acceptable.
For an anonymous visitor with zero investment? They're gone in 5 seconds.
A 36.7% completion rate tells you the demo didn't demo well.
3. The anomaly
Our final month saw an increase of 32% of our entire 6-month total. We never figured out why. Marketing push? Bots? A viral share?
We considered adding stronger bot defenses, but the ROI didn’t justify diverting time from core product work. When you can't explain a 3x spike in your core acquisition metric, your instrumentation has failed.
The Takeaway
"Let users try before they buy" is good product thinking. But it should not be at the cost of your runway.
We believe that if your value proposition is strong enough, people will sign up. If they aren't willing to type in an email address to solve their problem, they were never going to pull out their credit card anyway. We want to focus on the 1% who stay, instead of optimizing for the 99% who leave.