Skip to main content

Meet Retab

Learn how Retab brings shoppers back after they switch to another tab, by turning your background tab into a reminder of what they were shopping for.

Written by Vasil Dachev

Meet Retab, an AI Agent in Uxify’s Agent Catalog that refocuses shoppers on your store after they drift to another tab. While your store sits in the background, Retab adds an emoji next to your tab and alternates the title with a message from the page, such as the product they were viewing, the stock left in their size, or what’s in their cart. In a row of look-alike tabs, yours becomes the one that catches their eye.

Retab is powered by Uxify’s AI engine together with real-user data from Reality. It chooses which pages to cover, writes page-relevant messages for them, tests which ones bring shoppers back, and reports how many shoppers returned and how quickly.

Unlike a generic blinking tab or a notification tool, Retab works as an autonomous agent. Once activated, it decides where to run, writes and tests its own messages, keeps what works, reports its results, and can be customized through natural-language instructions.

How Retab works

Retab focuses on the moment a shopper moves your store out of sight.

A shopper comparing prices, checking a flight, or answering a message hasn’t left your store. They’ve just switched tabs, and your store slips into the background along with everything in their cart. With a dozen tabs open, it’s easy to forget it’s there.

Retab handles this in four steps:

  1. The visitor switches tabs. Your store drops into the background.

  2. Retab catches their eye. While they’re on another tab, yours shows an emoji and alternates the title between your own and a line from the page, such as “Only 2 left in M” or “2 items still in your cart”.

  3. The visitor comes back. Your tab is the one that moves and speaks to them, so it’s the one they click.

  4. Right where they left off. They’re back on your store with their cart intact, and your tab looks exactly as it did before.

Retab only works while your tab is hidden behind another one. The page the shopper sees is untouched, nothing is added to how your pages load, and the moment they return, your title is exactly as you set it. Your favicon never changes, so shoppers always recognize your store.

How Retab decides what to show

Retab doesn’t use generic reminders. Once a day, it uses Reality to find the pages where encouraging visitors to return makes sense, such as product pages, high-exit collection pages, and the cart. It then sets up truthful, page-relevant emoji and title messages for those pages.

Every message reflects what’s actually on the page:

  • ⌚ Only 3 left in stock for a shopper browsing watches

  • 👕 Only 2 left in M for a shopper looking at a shirt in their size

  • 👟 Free returns on all runners for a shopper comparing running shoes

  • 🛒 2 items still in your cart for a shopper who left items behind

Retab keeps learning. When it has enough data to judge a message, it tests a new title or emoji against it, keeps the one that brings more shoppers back, and retires the one that doesn’t. Every run builds on the previous one, so results compound over time.

Activating Retab

Retab is available as part of Uxify’s Agent Catalog.

After installing the Uxify snippet:

  • Open the Agents section in your Uxify dashboard.

  • Select Retab.

  • Review or customize the agent’s instructions.

  • Activate the agent.

Once enabled, Retab sets itself up with a relevant emoji and tab title and starts bringing shoppers back.

You don’t need to write code, design anything, or set up rules.

Dashboards in Retab

The Retab page shows the agent’s activity in three sections:

  • Results

  • Checkins

  • Instructions

Together they show how many shoppers Retab has brought back, what it has changed, and how it is configured.

Results

The Results tab summarizes Retab’s impact on your website.

At the top of the page you’ll find three key metrics:

Shoppers reactivated

The number of visitors who came back to your tab after Retab showed them a message during the selected period. A visitor counts as reactivated when they return to the tab they left.

Reactivation attempts

The number of times Retab showed a message on a background tab. Not every attempt brings a shopper back, but each one is a chance to win back a visitor who drifted away.

Dividing shoppers reactivated by reactivation attempts gives you Retab’s return rate, which the agent also tracks in its checkins.

Time to reactivate

The median time between a visitor leaving your tab and coming back to it. A shorter time means Retab is bringing shoppers back while they’re still in the buying mindset.

The chart below shows shoppers reactivated over time, along with the daily average. Use the date range selector to change the period.

The Who counts as reactivated panel next to the chart explains exactly how a return is counted.

Checkins

The Checkins tab is Retab’s activity timeline. Retab checks in once a day.

Each checkin records what the agent did during that run, so you can see how its messages and coverage have changed. Expand a checkin to see the steps it completed (Activate, Choose the daily signal, Learn from each run, and Measure and report).

Typical entries include:

  • The current return rate and whether it is holding steady

  • New messages added or tested

  • Coverage expanded to more pages, such as high-exit collection pages

  • Messages retired after underperforming

For example, Retab may report:

Retab reactivation steady at 35%, four new messages added.

Each checkin includes a timestamp, giving you a full history of the agent’s activity.

Instructions

Like every Uxify AI Agent, Retab can be customized.

The Instructions tab defines how Retab behaves. Instead of working through multiple settings, you adjust the agent with natural-language instructions. To change them, click Edit instructions.

Changes take effect automatically, with no extra implementation needed.

Retab’s instructions are organized into four steps:

Activate

Activation instructions define how Retab is set up when it starts running.

Typical instructions include:

  • Set up Retab after deployment with a relevant emoji and tab title.

Choose the daily signal

These instructions define where Retab runs and what it shows.

Typical instructions include:

  • Once a day, use Reality to find pages where encouraging visitors to return is relevant.

  • Set up truthful, page-relevant emoji and title messages for those pages.

  • Keep that setup in place until the next run.

Learn from each run

Learning instructions control how Retab tests and improves its messages.

Typical instructions include:

  • Review the current setup, comparable return data, sample sizes, and saved results.

  • Try a new title or emoji only when there is enough data to judge the previous one. Otherwise, keep it.

  • Record what was tried and what happened, for use in the next run.

  • Create rules only for URLs that appear in the data.

These instructions keep Retab honest and data-driven. It only changes a message when the evidence supports it, and every message reflects what’s really on the page.

Measure and report

Reporting instructions control what Retab tells you about its work and when.

Typical instructions include:

  • On the 1st day of every month, send me an email summarizing the results for the previous month.

As Uxify’s Agent platform evolves, more instructions and workflows can be added to further customize Retab.

Why Retab matters

Shoppers rarely buy in one sitting. They compare stores, check prices elsewhere, and get pulled away by messages, and many never find their way back to the tab they left, even with items in the cart.

A generic blinking tab is easy to ignore. Retab reminds shoppers of what they cared about, like the last item in their size or the cart they left, right when they’re deciding where to click. And because it only works while your tab is in the background, your pages, design, and speed stay exactly as they are.

The result is a lightweight, fully automatic agent that wins back distracted shoppers and shows you exactly how many come back and how fast.

Did this answer your question?