---
title: Newsletter & Paywall Targeting by Reader Depth
description: "Newsletter signup targeting by reader depth: prompt the 3+ article readers most likely to say yes, suppress the leavers, and grow the pool that gets asked."
url: https://preview.artificialpoets.com/use-cases/depth-triggered-conversion/
site: Artificial Poets
type: page
date: 2026-08-09T19:46:31+00:00
modified: 2026-08-18T02:10:14+00:00
---
# Ask at the Right Article

**Measured on live deployments**

"I'm showing newsletter popups to readers who were leaving anyway — how do I ask the ones who are ready?" Stop asking at the door. Only 6.2% of sessions reach a second article, so an arrival popup interrupts the 94% who were leaving anyway: the Platform fires your prompt at the depth threshold you set, and grows the pool that reaches it, with 3+ article sessions up 91%.

[See it on my numbers](/request-a-demo/)

Horse & Rider · Ride TV · Equine Network · Equus Magazine · The Score · Equine Network Lockup

## The problem, in your numbers

The prompt was never the problem, the audience was. Registered readers convert to paid at roughly 45 times the anonymous rate (Piano), and one extra page moves week-one return probability from 8% to 22% (Chartbeat). A better popup asks the same wrong audience; a depth trigger changes the audience.

- **0.22%** Anonymous readers converting to paid, vs ~10% once registered (Piano benchmark, via UniSignIn)
- **0.62%** Median free-to-paid newsletter conversion, platform-wide (beehiiv)
- **6.2%** Sessions reaching a second article before enablement, on the network we measure
- **−22%** Nine-month erosion of the 3+ article pool on titles that change nothing (our panel)

## Where the ask lands

Share of sessions reaching each depth state before enablement. The default popup fires at state one; the readers most likely to say yes are two states further in. On enabled titles the 3+ pool grew 91% and the 5+ pool 221%, still under 2% of sessions.

- **100%** 1 page — everyone gets the popup, the default
- **6.2%** 2+ articles
- **1.9%** 3+ articles — ask here
- **0.32%** 5+ articles — ask here

## The evidence, at its honest tier

- **The pool this use case targets grew where the Platform ran** — 3+ article sessions +91%, 5+ sessions +221% on enabled titles — from a small base: under 2% of sessions reach five articles even after the pool more than tripled
- **Context** — multi-page sessions +46% (10.3% → 15.1%) on enabled titles vs −16% (12.0% → 10.1%) on comparison titles over the same nine months
- **On unchanged titles the target pool eroded** — 3+ −22%, 5+ −17% — waiting shrinks the audience you would prompt
- **We measure the depth states and the pool** — Signups and stop rate stay measured in your ESP and wall — the conversion number is yours, not ours
- **Traffic did not grow.** — Sessions and users were flat — this is a depth result, and we say so

The pool this targets, measured: [Equus Magazine tripled the sessions that reach five articles](/customers/equus-magazine-deep-readers/).

**Introducing**

## Artificial Poets Platform

One engine behind every solution on this site. It learns your archive and your readers, then acts inside your CMS, your templates and your ad stack.

- It learns your archive: **Every story you have published, current again** — The engine understands each piece by what it is about, not when it ran or where it was filed. A feature from 2019 competes for the next slot on merit with one from this morning.
- It reads the visit: **What a reader wants, without asking** — Interest builds from what someone actually does in the session. No login, no third-party cookies, nothing leaving your domain. Useful on the second pageview, not the tenth visit.
- It chooses: **The right next read, not the popular one** — Someone comparing products and someone following a running story want different things. A most-read list gives both the same five links and serves neither.
- It serves: **There before the reader leaves** — The feed, the recommendations, the search answer and the signup ask all run on the same engine, in your templates and your ad stack. Any slot that arrives with them is yours to sell.
- It proves: **A lift you can defend, or we say so** — Every deployment runs beside titles that did not get it, plus a serving pause. That is how a result becomes a number you can take to a board instead of a vendor claim.

## FAQ

### Does this replace Piano, Zephr, or our ESP?

No. Your wall vendor and ESP keep the ask; the Platform supplies what they don't have — a depth state per reader, and a growing pool of readers who reach it. In Salem Reporter's 30-day head-to-head, a registration wall out-signed a newsletter form 16× (Leaky Paywall); that is an argument about where you ask. Depth state is the argument about when, and it feeds the instrument you already run.

### What depth threshold should we set?

Start from where your own conversions already cluster rather than from our number. The pattern in the measured data is that the readers most likely to say yes are two states further in than the default arrival popup, and the baseline month exists so that threshold is chosen from your funnel, not ours.

### Does it tell a subscriber from a first-time arrival?

It tracks reading-depth state per reader, so your rules can distinguish someone on their fourth article from someone who just landed. What the prompt then says, and whether a subscriber sees anything at all, stays in the tool you already run.

### Are these before/after dashboard numbers, or is there a control?

The depth figures run against a comparison group of titles in the same network that changed nothing — where the 3+ pool fell 22% and the 5+ pool −17% over the same nine months. One enabled title ran +12.2%, then a four-week pause in serving during which the Platform did not serve; the metric fell to −2.7% and recovered to +6.9% after serving resumed. Conversion incrementality on your prompts is measured in your own stack, with a design agreed before enablement.

### "The model decided" won't survive my board — and is reader-level targeting a GDPR problem?

The targeting rule is a sentence, not a score: ask at the third article, gate at the fifth, never prompt a reader the intent model reads as leaving. Your team sets the thresholds and can read them back. The state is behavioural, anonymous and first-party — no PII stored, GDPR/CCPA-compatible — closer to a meter rule than to propensity-to-subscribe modeling on personal data.

### Where does the depth data live?

On your properties. The profile is anonymous and first-party, with no PII, no third-party cookies and nothing leaving your domain, and serving events land in your own analytics stack so the record is yours to query.

### What does integration take?

Authorization, CMS access and infrastructure access, then a JavaScript tag carried out by our team. Your prompts, offers and rules stay where they are; what changes is when they fire.

## Related

- [Equus Magazine Tripled the Sessions That Reach Five Articles](https://preview.artificialpoets.com/customers/equus-magazine-deep-readers/) — Most publishers measure whether a reader gets to a second article. This is what happens after that: sessions reaching three articles nearly doubled, and sessions reaching five tripled.
- [Where Your Next 1,000 Subscribers Actually Come From](https://preview.artificialpoets.com/blog/deep-readers-subscriber-funnel/) — Every conversion a publisher cares about — newsletter signup, registration, subscription — draws from the same small pool: readers who went well past the article they landed on. That pool is the top of every loyalty funnel you run, most publishers have never measured its size, and on the network we measure it was quietly shrinking at every depth until something changed it.
- [Session Depth Assessment](https://preview.artificialpoets.com/resources/session-depth-assessment/) — Two numbers from your analytics place your single-page rate and depth distribution against a continuously measured publisher panel.
- [Build the Audience That Comes Back](https://preview.artificialpoets.com/solutions/direct-audience/) — "My subscriber targets keep rising while my funnel shrinks — where does direct audience actually come from?"

## See how this works on your titles

Thirty minutes on your analytics. We tell you what share of your sessions stop at the first page, and what this would realistically move first for a setup like yours. You leave with the annotated read. No deck.

[See it on my numbers](/request-a-demo/)

How these numbers are made: [the measurement method](/solutions/measurement/). Figures from a multi-title publisher network measured continuously, August 2025 to May 2026. Last updated August 15, 2026.

## Questions this page answers

### What is depth-triggered conversion?

It is firing your newsletter, registration or paywall prompt at a reading-depth threshold instead of on arrival or after a timer. The reader who has read three articles is a different prospect from the one who just landed. On the network we measure, only 6.2% of sessions reached a second article before enablement.

### How does depth targeting work with our existing prompts?

Your prompts, rules and offers stay in the tools you already run. The Platform tracks each reader's depth state in real time and fires your prompt at the threshold you set, while growing the pool of readers who reach it: 3+ article sessions rose 91% and 5+ sessions 221% on enabled titles.

### How much does it cost?

There is no public price list; pricing follows network size and formats. It does not replace your subscription stack, so nothing already paid for is displaced. The four-week baseline records your current prompt performance first, which is what any depth-triggered result gets compared against.

### Who is it for?

Audience and consumer-revenue teams whose prompt interrupts everyone and converts almost nobody. If your popup fires on arrival, it is asking the roughly 94% of readers who were leaving anyway and burying the offer for the few who were not.

### What are the alternatives?

Time-on-page and scroll-depth triggers approximate interest with proxies that a slow scroller and a distracted tab both satisfy. Exit-intent fires when the decision is already made. A better-designed popup asks the same wrong audience; changing the trigger changes who is asked.

### How do I get started?

Pick the depth threshold and keep your existing prompt and offer. Authorization, CMS access and infrastructure access are the requirements, four weeks of measurement-only baseline record the current conversion behaviour, then enablement on a named day, with your funnel numbers staying in your own stack throughout.