How do you tell whether low activation is caused by targeting or onboarding?

I recently ran my first small paid acquisition experiment for a workout-tracking mobile app.

The campaign targeted Android users in Poland and produced:

• 27 installs

• 3.47 PLN cost per install

• 8,705 impressions

• 4,481 accounts reached

• 1.94 average frequency

The campaign was planned to run for seven days, but I stopped it after four because the number of daily installs appeared to be declining.

I also adjusted the campaign almost every day. Looking back, that made the results difficult to interpret and probably didn’t give the platform enough stable data to optimize.

However, the more interesting problem appeared after the install.

Users were willing to install the app and create an account, but only a small percentage reached the product’s core action: selecting a plan, starting a workout and recording a session.

I currently see two possible explanations.

Acquisition problem:
The campaign reached people interested in fitness, but not necessarily people interested in planning and tracking workouts.

Activation problem:
The app was originally designed around my own habits. It may require too much setup, present too many options or fail to communicate a clear first step.

The hypotheses I’m considering include:

• Mandatory registration creates unnecessary friction

• The first screen doesn’t provide one obvious action

• Users should be able to explore before creating an account

• A ready-made plan should be available immediately

• The product shows too many features before demonstrating its core value

For founders who have dealt with a similar problem:

• How did you separate poor targeting from poor onboarding?

• Which event did you use as your activation metric?

• Did delaying registration improve activation?

• What qualitative or quantitative data would you collect before changing the onboarding?

I’m especially interested in frameworks or experiments that helped you identify the actual point of failure.