---
title: Where Your Next 1,000 Subscribers Actually Come From
description: "Newsletter signups, registrations and subscriptions all draw from one small population: readers who go three, five, ten articles deep. Most publishers have never measured that pool. Here is what it looks like, and how it moves."
url: https://preview.artificialpoets.com/blog/deep-readers-subscriber-funnel/
site: Artificial Poets
type: post
date: 2026-08-10T03:32:09+00:00
modified: 2026-08-14T18:30:48+00:00
categories:
  - Audience
---
# Where Your Next 1,000 Subscribers Actually Come From

## Why direct audience became the whole strategy

The industry consensus hardened through 2025–26: with search referrals down a third and falling,[[1]](#ref-1) the only defensible audience is the one that comes to you on purpose. Publishers responded by investing in newsletters, apps, registration walls and subscriptions — the "own the relationship" playbook every strategy deck now contains.

What the decks skip is the arithmetic underneath. Direct relationships are not conjured by a newsletter module; they are *converted* from a specific reader state — the state of being deep in your content, voluntarily, right now. The Globe and Mail's Sophi system demonstrated the scale of what sits behind that conversion when the targeting is right: a **222% increase in registrations and 51% in subscription conversions**.[[2]](#ref-2) The raw material for those numbers was not traffic. It was engaged sessions.

Which raises the question almost nobody can answer about their own site: **how many of your sessions ever reach the state where conversion is plausible?**

## What the deep-reader pool actually looks like

We measure a multi-title special-interest publisher network continuously. Here is the session-depth distribution on one title before any intervention — and it is the ordinary shape, not a bad one:

| Articles reached in a session | Share of sessions |
| --- | --- |
| 2 or more | 6.2% |
| 3 or more | 1.9% |
| 5 or more | 0.32% |
| 10 or more | 0.05% |

Read the bottom rows again. The population that reads five articles in a sitting — the readers who plausibly register, subscribe, or become daily habits — was **about one session in three hundred**. The ten-article readers were one in two thousand.

This is the funnel above the funnel. Your newsletter module's conversion rate is measured against the sessions that see it in a receptive state, and that population is two orders of magnitude smaller than your traffic number. Growing conversions has two levers: convert the pool better, or **make the pool bigger**. Nearly all publisher effort goes to the first lever. The second is where the compounding is.

## The pool was shrinking on its own

Across the comparison titles in the network — properties where nothing changed for nine months — the deep pool eroded at every depth we measured: sessions reaching three or more articles fell **22%**, five or more fell **17%**, ten or more fell **13%**.

No strategy change, no redesign. Just the ambient drift of shallower visits that the whole industry is reporting.[[3]](#ref-3) If your subscription plan for next year assumes this year's pool size, it is assuming something that is measurably not holding.

## What happened when the pool got bigger

Two titles in the same network enabled our Platform — contextual recirculation that serves a next article as the reader scrolls. On the measured title above, the distribution moved like this over the first months:

| Articles reached | Before | After | Change |
| --- | --- | --- | --- |
| 3 or more | 1.9% | 3.7% | +91% |
| 5 or more | 0.32% | 1.04% | +221% |
| 10 or more | 0.05% | 0.29% | +443% |

Two things to say plainly, because the honest version is the useful one.

**The bases are small.** Even after tripling, fewer than 2% of sessions reach five articles. This is a small population getting much larger, not a majority behaviour — the 3+ figure, on the biggest base, is the one to plan against. But small bases are exactly what loyalty funnels are made of: a subscription programme does not need 40% of sessions to go deep. It needs the deep pool to be twice as large as it was.

**The lift compounds with depth** — +91% at three articles, +221% at five, +443% at ten — and the mechanism shows in the telemetry: a reader who accepts a second served article continues to a third **50%** of the time, and past the fourth the continuation rate runs **70–82%**. Attention per article settles around 19 seconds and stays there. The hard part is the second article; after that, you are serving a reader who has already decided to keep going.

## What to do with this

1. **Measure your pool.** One query: share of sessions reaching 2, 3, 5 and 10 distinct articles. If your analytics can't produce it, that is finding number one.
2. **Trend it before you target it.** The network data says the pool erodes by default now. Know your slope before setting next year's subscriber goal on top of it.
3. **Point conversion at the pool, not the traffic.** Registration prompts and newsletter units shown to five-article readers address the 0.3–1% who are most likely to say yes — and interrupt the 90%+ who were leaving anyway. Depth is the targeting signal most stacks already have and don't use.
4. **Grow the pool deliberately.** Whatever the mechanism — editorial packaging, contextual recirculation, product work — hold comparable properties back as a comparison, because the default trend is down and a flat pool is already a win.

- Deep-reader pool on an ordinary title: 5+ articles = **~1 session in 300**; 10+ = ~1 in 2,000
- Left alone for nine months, the pool shrank at every depth: **−22% / −17% / −13%**
- With contextual recirculation: **+91% / +221% / +443%** at 3+/5+/10+ — small bases, stated
- The mechanism: continuation past the fourth article runs **70–82%**
- The proven downstream: Sophi's targeting produced **+222% registrations, +51% subscription conversions**

## FAQ

Conversion rate divides by an audience whose size you are not tracking. If the deep pool halves and your conversion rate holds, absolute signups still fall — and the dashboard says nothing changed. Measure the pool and the rate.

It is normal. Published depth data is rare, which is partly why every publisher assumes theirs is uniquely poor. The distribution above sat inside a healthy, professionally run network. The actionable fact is not the level; it is that the level responds — in both directions.

Directionally, yes — habit research consistently ties repeat behaviour to early depth (one more page on a first visit nearly triples week-one return probability[[4]](#ref-4)), and registration systems like Sophi monetise exactly that state. Your own funnel data can confirm it in an afternoon: segment conversions by session depth at the moment of conversion.

## How big is your deep-reader pool?

The Session Depth Assessment estimates your 3+ and 5+ populations from two analytics inputs, benchmarked against the measured panel in this article.

## References

1. Chartbeat data reported in Press Gazette, “Global publisher Google traffic dropped by a third in 2025.” [pressgazette.co.uk](https://pressgazette.co.uk/media-audience-and-business-data/google-traffic-down-2025-trends-report-2026/)
2. A. Hermida et al., *Sophi at The Globe and Mail* (2024).
3. Nitropay, “Protecting Ad Revenue Per Session in the Age of AI Search.” [blog.nitropay.com](https://blog.nitropay.com/protecting-ad-revenue-per-session-in-the-age-of-ai-search/)
4. Chartbeat, “Encourage multiple page visits, use social video for successful 2026 publishing strategy.” [chartbeat.com](https://chartbeat.com/resources/general/successful-2026-publishing-strategy/)

## See what your session depth looks like

We read your analytics with you for twenty minutes and tell you what share of your sessions stop at the first page. You leave with the annotated read, whether or not we ever talk again. 20 minutes, your analytics, no deck.

[Book my session-depth read](/request-a-demo/)