The Hidden Cost of “Free” Users in SaaS (And Why 30% Weekly Active Usage Is Our North Star)

One of the most common pieces of advice in SaaS is also one of the most dangerous:

“Just add a free plan. You’ll figure out monetization later.”

We followed that advice.
Then we reversed it.
And now we are doing free again, but this time with a clear north star.

Here’s what actually happened while building SocialPost.ai, and why about 30% Weekly Active Usage (WAU) is the metric that now decides the fate of our free plan.

We Started Free (Because Everyone Does)

When we launched, we led with a free plan.

Not because it was perfectly thought through, but because:

The market expected it
Competitors had one
We wanted fast feedback and distribution

At first, it worked.
Users signed up. Feedback flowed. Learning accelerated.

But as usage grew, so did real costs. Cloud compute, AI usage, data, and support all added up. Like many founders, we started telling ourselves a familiar story:

“Free users are slowing us down.”

So We Took Free Away and Growth Stalled

Before the product was truly finished, we removed the free plan.

The logic sounded reasonable:
-Serious users will pay
-Free users don’t convert
-We need cleaner signal

What actually happened:

Signups dropped sharply
Feedback loops dried up
Organic growth stalled
Momentum disappeared

Nothing broke technically, but energy vanished.

That’s when it clicked. We hadn’t removed noise. We had removed oxygen.

The Real Cost of Free Users (The Honest Version)

Free users do cost money:
-Cloud compute and storage
-AI inference
-Analytics and data footprint
-Support and edge cases
-Feature requests that don’t always align

That part isn’t controversial.

What surprised me was this:

Free users weren’t the problem. Undisciplined free was.

We blamed free for issues that were actually about product maturity and lack of guardrails.

The Missing Piece: We Prune Free Users Aggressively

When we brought free back, we didn’t do it naively.

We built an automated pruning process:

-If a free user doesn’t log in for 30 days, their account is suspended
-Continued inactivity leads to automatic deletion
-Multiple warning emails go out along the way
-The moment they log back in, the clock resets

No manual reviews. No exceptions.

This keeps our user base:

-Clean
-Lower-cost
-High-signal
-Focused on people who actually need the product

Free isn’t set it and forget it.
It is actively managed.

What Happens When We Eventually Remove Free Again

We are also being intentional about how we remove free when the time comes.

When we decide to turn off the free plan:

-New users will no longer be able to sign up for free
-Existing free users will be allowed to remain on the free plan
-As long as they continue logging in at least once every 30 days, they keep access

This avoids rug pulls, preserves trust, and rewards users who are actively using the product.

Free is not a trap.
It is a contract with clear expectations on both sides.

Why 30% Weekly Active Usage Became Our North Star

When we reintroduced free, we needed a hard, defensible line that told us when free had done its job.

That line is about 30% Weekly Active Usage.

Here’s why:
-Below 20% WAU means curiosity and tire-kicking
-20 to 25% WAU means early value, but fragile
-Around 30% WAU means real habit formation and clear product value
-40% or higher means exceptional engagement, usually later-stage

At around 30% WAU:
-Users return without reminders
-Activation works without hand-holding
-Value is obvious and repeatable
-Monetization pressure is fair and earned

Below 30%, free is a learning tool.
At 30% and above, it becomes a business decision.

The Two Rules That Keep Free Honest

This time, free isn’t permanent.

We’ll keep the free plan until one of two things happens:
-Our cloud consumption credits expire
-Weekly active usage reaches about 30% of users

Free exists to get us there, not to live there forever.

How I Think About Free Users Now
I don’t see free users as freeloaders.

I see them as:
-Early signal
-Product QA at scale
-Distribution leverage
-Proof of relevance

But only if they are actually using the product.
Inactive free users don’t help anyone, including themselves, and it’s okay to let them go.

Final Thought
Free users are expensive.
But silence is more expensive.
If you’re early, still learning, and still shaping the product, free paired with discipline and a clear north star might be the most honest mirror you’ll ever get.

Curious how others here handle turning off free, grandfathering users, or defining product readiness with engagement metrics.