They read deeply on the first visit, then start to leave.

Inferred

First session, two substantially completed articles on the same topic, no subscription or follow action, and exit starting.

Example eligibility rule

Count an article as substantially completed after at least 70% scroll depth and 60 seconds of active reading. Trigger only after two qualifying articles in the same topic during the first session. Do not trigger from background tabs or rapid scrolling.

A new reader finishes most of a long article, opens another story on the same topic, and then moves to leave. That pattern suggests medium intent, but the reader has not declared a preference. The default response is an email-capture popup unrelated to what they read.

Avoid this assumption

Inferred interest is not consent. Reading two climate-policy articles does not justify subscribing the reader to a newsletter or declaring that climate is a permanent preference.

Test a quiet next step against doing nothing. The signal is inferred, so the treatment should remain easy to ignore and should not write a lasting preference without permission.

Reach

How many eligible people can you reach in the chosen channel and time period?

Cost

What will content, delivery, QA and ongoing maintenance cost?

Return

What incremental outcome would cover that cost, and what is it worth?

Planning inputs, not measured results. Estimate reach × incremental outcome rate × value per outcome, then subtract cost. Use a range and a consistent time period.

Example format: Content

Offer "Follow climate policy" or "See this week's climate briefing" in a dismissible inline card.

Compare lighter and deeper responses

Choose the lightest response that solves the need.

Assist: Offer useful help

Show the next climate-policy article in the normal related-content area.

Adapt: Change a meaningful part

Offer "Follow climate policy" or "See this week's climate briefing" in a dismissible inline card.

Transform: Tailor the journey

If the reader chooses to follow the topic, use that declared preference to shape the next visit and subscription offer.

What to watch out for

• An email modal before the reader demonstrates interest.
• Treating one article view as medium intent.
• Saving an inferred topic as a permanent preference without asking.

Control

Standard related-content and subscription experience.

Treatment

Topic-relevant next step for qualifying first-session readers, with a holdout.

Success measure

Topic follow, subscription exploration, or another substantially read article.

Guardrail

Immediate exits, dismissals, and lower article completion.

Among first-session readers who met the two-article rule, compare the topic-relevant treatment with standard related content. Measure topic follows, subscription exploration, and another substantial read without combining them with low-intent first-page visitors.

Win or lose, record what worked, for whom and why. What you learn becomes the next signal.

• Convert Experiences Website A/B tests and 1:1/segment personalizations built on the same testing infrastructure, with a visual editor and dev-level code hooks.
• Insider One Cross-channel engagement platform (web, app, messaging) that builds a unified profile from attributes, real-time behavior, and precomputed predictive scores to drive personalization and campaign targeting.
• Kameleoon Audiences Experimentation platform with an AI Predictive Targeting feature that scores each visitor's likelihood to convert (0-100) and lets you target or exclude by that score.
• VWO Testing Testing platform with audience/behavior targeting and a visual editor; personalization runs as a separate module from Experimentation. As of September 2026, VWO and AB Tasty merged their brands under the new parent brand Wingify, with each platform still running independently during the transition.
• AB Tasty content personalization Audience- and intent-based content targeting for anonymous visitors, built on live signals (referral source, device, location, scroll depth) rather than login or cookie data. As of September 2026, AB Tasty and VWO merged their brands under the new parent brand Wingify, with each platform still running independently during the transition.
• Conductrics Rule-based and algorithmic (contextual bandit) targeting that reacts to in-session behavior, not just page loads.

The ad made a promise the page doesn't keep. Keep that promise on the page they land on.

A first-timer and a regular see the same homepage. Help new people get oriented. Let regulars skip ahead.