They landed in the wrong country, currency, or language.

Inferred

IP country, browser language, current store, and currency behavior.

Example eligibility rule

Offer a locale switch when IP country and browser language agree and both differ from the current store, or when the visitor changes currency twice. If location and language disagree, keep the current store and show only a quiet selector.

IP, browser language, or a .com default puts them in the wrong catalog. Prices in the wrong currency. Copy they cannot read. They have to hunt a flag icon before they can shop or sign up.

Avoid this assumption

Auto-redirecting without a way back is how you hide the catalog they typed the URL for.

This is manners, not targeting. Detect, offer the match, and let them keep the default. Easy to hold out against the untranslated homepage.

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: Experience

Offer the matching store first. One tap to stay on the current one.

Compare lighter and deeper responses

Choose the lightest response that solves the need.

Assist: Offer useful help

A quiet bar: "Shop in £ / English?" with keep-this-site.

Adapt: Change a meaningful part

Offer the matching store first. One tap to stay on the current one.

Transform: Tailor the journey

Route new sessions to the matching store, always with an undo.

What to watch out for

• Hard-redirecting VPN or travel traffic with no undo.
• Translating the nav and leaving checkout in the default language.
• Treating language as a personalization "win" instead of a prerequisite.

Control

Default store for everyone on that URL.

Treatment

Locale offer or route, with a holdout.

Success measure

First-session task start after the switch (search, PDP, signup).

Guardrail

Undo rate and bounce. If people flee the "helpful" store, you guessed wrong.

Among visitors who met the locale-mismatch rule, compare the switch offer with the unchanged store. Measure successful task starts after switching and track undo, bounce, and manual store changes.

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

• Adobe Target Uses machine learning (a Random Forest model plus a multi-armed bandit) to automatically select the best-performing, marketer-defined experience for each visitor, rebuilding its model roughly every 24 hours.
• Sitecore Personalize Real-time web personalization and A/B/n testing inside Sitecore, driven by live behavioral and contextual signals.
• 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.
• 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.
• Uniform Edge-delivered, component-level personalization and testing for composable front ends.
• Webflow Optimize Runs three experience types on Webflow sites: standard A/B tests, AI-optimized tests that keep shifting traffic to whatever's currently working instead of calling one winner, and manual rules-based personalization for defined audiences.

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

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