This week, I worked with a client who was ready to scale their app. They had a decent landing page and a good offer.
They had spent $2,500 on a small ad campaign. The conversion rate was terrible.
Their immediate thought: “The ad copy is bad. We need better creative.”
My response: “Stop the campaign. Now.”
The problem wasn’t the ad copy, or the offer, or the landing page. It was one missing, non-AI feature: basic event logging.
You cannot buy data if you can’t capture it. Running ads before you have event logging (who used the app and how) is just lighting money on fire to keep yourself warm.
It means the $2,500 campaign was not a learning expense — it was a vanity cost.
Here is the simple framework we developed (The Pre-Spend Check):
The “Lights On” Check. Before spending $1 on ads, you must install an analytics tool. Install PostHog or Mixpanel. Don't use Google Analytics for product learning.
The “Heartbeat” Event. Identify the ONE action that proves a user is seriously engaged (e.g., “Clicked Generate Report” or “Submitted First Task”). This is your product’s heartbeat.
The Spend Rule. If you can’t reliably see the Heartbeat event tied to a specific traffic source, you pause the spend. No exceptions.
Our inability to identify product usage was the biggest failure. Don’t repeat it.Speak soon,— Dmitry
P.S. If you’re wasting money on ads or engineering, it means your funner is broken. You're not alone. Subscribe to summaries of my sessions with real founders here. Learn from others' mistakes — it's much cheaper.