I wasted $3,500 on X ads and got zero actual signups

Running ads on X (formerly Twitter) for a B2B SaaS product feels like throwing money into a black hole if you don't account for the bot problem. I spent four months—from May to August—testing a paid acquisition strategy for my latest project, an AI-native ticketing system. The goal was to see if we could drive high-intent traffic to our freemium platform, where users sign up directly without a demo. Instead, I ended up with a massive gap between "platform metrics" and "real-world business value."

For context, the product is an LLM agent-based workflow. It aggregates support issues from GitHub, Discord, forums, and email, processes them via AI, and drafts replies back to the original source. It’s a technical tool meant for developers and support teams, which is why I thought X would be the perfect playground.

The Setup and Strategy

We didn't just blast the whole world with ads. After an initial period of broad targeting, we pivoted to focus specifically on Tier 1 markets. The logic was simple: users in these regions generally have a higher propensity to convert to paid subscriptions.

• Ad Format: Text-based posts and image-based creatives.
• Call to Action (CTA): Direct links to our landing page or deep links to our technical blog on Medium.
• Total Budget: Approximately $3,509 over the four-month period.

The Data Discrepancy

This is where things get weird. If you look at the dashboard X provides, the numbers look absolutely massive. They claim we hit 46 million impressions and over 26,000 link clicks. On paper, that looks like a successful campaign. However, when you cross-reference those numbers with our internal analytics (using Matomo), the reality is much more sobering.

The actual traffic reaching our site was a fraction of what the ad platform reported:

• Total Website Visits from X: 613
• Total Actions on Site: 885
• Actions per Visit: 1.4
• Average Session Duration: 50 seconds
• Bounce Rate: 74%
• Actual Signups: Effectively zero

The Bot Problem and My Takeaway

Why the massive gap between 26,000 clicks and 613 visits? The answer is the elephant in the room: bot engagement.

Every time we ran a post, we were immediately hit by a wave of crypto bots and random engagement accounts. These bots are programmed to interact with posts, which inflates the "engagement" metrics and can even trigger fake clicks that never actually resolve into a page load. This creates a false sense of momentum in your ad manager.

If you are building a technical AI workflow or a developer tool, a "high impression" campaign on X might actually be a waste of your budget. The engagement is often superficial or entirely automated. For a B2B SaaS looking for real users who will actually integrate an agent into their GitHub or Discord, the signal-to-noise ratio on X is currently too low to justify a heavy spend. My advice? If you're doing a deep dive into X ads, don't trust the platform's click counts. Always verify everything against your own server-side tracking or Matomo/Google Analytics to see the true human traffic.