How Substack Really Works: Core Audience, Metrics, and Silent Readers Explained

Over the past weeks, something unexpected happened.

I reduced my activity.

I stopped actively tagging other writers.
I stopped fueling cross-dialogue for reach.
I stopped feeding the Notes stream for visibility.

And yet:

My open rate increased.
Reading depth increased.
Unsubscribes did not spike.
Engagement stabilized.

This is not a theory.
These are my numbers.

So no — I am not claiming that “the algorithm has radically changed.”

What I am observing is a redistribution of weight.

What Used to Work

Mentions worked.

Cross-citations worked.

Dialogue loops worked.

They created momentum.
They generated visibility spikes.
They signaled social relevance.

But those effects were short-term.

They amplified the surface.

What I Observe Now

The system appears to weigh sequence more heavily than signal.

Not revolution.
Not conspiracy.
Not dramatic transformation.

Just weight.

More attention seems to go toward:

– whether readers continue reading
– whether they return
– whether texts form a trajectory
– whether a series exists
– whether attention sustains

A mention creates motion.
A series creates gravity.

Gravity lasts longer.

The Trust Problem

This is where the real choice appears.

An author can chase expansion.
Or an author can build trust.

Trust is not built through tagging.
Trust is built through consistency.

When a reader returns not because they were mentioned,
but because they know the line will be held.

The temptation is always reach.

Reach is visible.
Trust is invisible.

But trust compounds.

And this is the difficult part:
if you write for reach, you will gradually soften your edge.
If you write for trust, you may lose noise — but you strengthen your core.

That is not romantic.
It is structural.

Update

Substack has clarified that they recently changed how delivery and open rates are calculated.
Views in the app were previously undercounted, and are now included.

This means part of the increase many of us are seeing may reflect more accurate measurement — not necessarily a behavioral shift.

That distinction matters.

The strategic question, however, remains unchanged:
whether we optimize for surface signals or build long-term reader return.