Changelog

You can now bring a custom website URL into Anagram when importing locations by CSV. If you provide one, Anagram will use that link anywhere the location’s website appears—so customers and teammates land on the page you intended, not just the default website pulled from Google Places.

What's new

• CSV website mapping: Location imports now support an optional website URL column, so you can set a custom site for each location during import.

• Custom links take priority: When a location has an imported website URL, Anagram shows that link instead of the Google Places website.

• Consistent location details: Opening a location now takes people to your custom website URL when one is available.

Why we made these changes

Google Places is helpful for filling in location data, but it doesn’t always point to the best destination for every store or branch. Some teams need location-specific landing pages, campaign URLs, or updated local sites that differ from Google’s listing. This change gives you more control over where each location link goes, while keeping imports fast and easy to manage.

Shoppers can now add Shopify product recommendations directly to their cart from the Anagram Experience, instead of leaving the conversation to open a product page first.

What's new

• Direct add to cart for Shopify recommendations: When a recommended product is backed by Shopify, shoppers can add it to their cart right from the recommendation card.

• Better measurement of recommendation performance: Add-to-cart clicks, successes, and failures are now tracked, giving teams a clearer view of which recommendations are driving purchase intent.

• Improved attribution for downstream conversion reporting: Successful add-to-cart actions can now be tied back to recommendation activity, making it easier to understand how recommendations influence cart and checkout behavior.

Why we made these changes

When a shopper sees something they want, sometimes sending them away to a separate product page adds friction at exactly the wrong moment. We made this update to shorten that path: see a recommendation, add it to cart, keep going. At the same time, we improved reporting so teams can better measure which recommendations are not just getting clicks, but actually moving shoppers closer to checkout.

Recommendation filters are now much easier to set up. Instead of typing tags, collections, and attributes by hand, you can now choose from values already in your catalog, which helps you avoid typos and reduces the chances of seeing misleading product counts.

What's new

• Searchable tag and collection pickers: When filtering product recommendations by tag or collection, you can now search and select from values already in your catalog instead of entering everything manually.

• Custom values still supported: If you need to filter on something that isn’t in the list yet, you can still type your own value and use it right away.

• Attribute filters guided by real catalog data: For attribute-based filters, you can now pick the attribute and then choose from the values that actually exist for that attribute in your catalog.

Why we made these changes

Recommendation filters should help you narrow to the right products quickly, not make you remember exact field names or troubleshoot why a valid filter returns no matches. These changes make the setup flow faster and more dependable by grounding filter choices in your actual catalog data while still giving you flexibility when you need it.

You can now download the full metric list behind any Insights metric as a CSV. This allows you to go do further analysis in Excel, Sheets, or your BI workflow.

What's new

• Export from metric details: Each metric’s conversation view now includes an Export CSV action, so you can download the related conversations without leaving Insights.

• Full result export: Exports include all conversations that match the metric and date range, not just the current page of results.

• Useful spreadsheet-ready columns: The CSV includes key fields like date, page, first message, and response count to make recurring themes and high-traffic pages easier to spot.

Why we made these changes

When a metric spikes, teams often want to go deeper: which pages are driving it, what customers are asking, and whether the same issues or intents keep coming up. CSV export makes that handoff faster by giving you the full conversation set in one step.

You can now start chat experiences from a centered prompt bar at the bottom of the page. Instead of choosing between a fully embedded chat or a floating button, you have a new option that feels more like a natural starting point for asking a question, then expands into a full chat panel when someone engages. A good use case of this is to make the Experience “Page-Aware” and then apply this Experience on each of your pages and be able to answer questions from any page, about any page.

What's new

• New centered prompt bar launcher: Chat experiences can now begin as a bottom-centered input bar, giving visitors a clear place to start typing right away.

• Expands into a full chat panel: When someone clicks, focuses, submits a question, or selects a suggestion, the experience opens into a larger chat panel for the full conversation. On mobile, a IOS-like drawer is used.

Why we made these changes

Some teams want something more inviting than a floating button, but lighter-weight than showing the full chat upfront. This new launcher gives visitors an immediate, familiar prompt to interact with, while keeping the page clean until they’re ready to continue the conversation.

Product recommendations now give shoppers a clearer sense of value by showing the original price alongside the current price when that information is available. That means recommendation cards can better reflect discounts and price drops without requiring shoppers to click through to the product page first.

What's new

• Comparison pricing in recommendation cards: Recommended products can now show a crossed-out original price next to the current price.

• Better discount visibility: When a product is on sale, shoppers can see the before-and-after pricing at a glance directly in the recommendation experience.

Why we made these changes

When shoppers see a recommended product, price is one of the first things they use to decide whether to engage. Showing the original price next to the current one makes promotions easier to understand and helps recommendation blocks do a better job of driving discovery.

You can now connect Klaviyo to Anagram and turn conversations into list growth. Once connected, Anagram can capture a shopper’s email during chat, pull out the other details you care about, and send everything to the Klaviyo list you choose automatically after the conversation ends.

What's new

• Choose the exact list to send to: When you enable the Klaviyo skill in an experience, you can pick which Klaviyo list new contacts should be added to.

• Capture more than just email: Define extra shopper details to collect from conversations, like budget, product interest, or other attributes that matter to your team.

• More natural lead capture in chat: Anagram can guide the conversation to gather email and selected shopper details in a way that feels relevant to the interaction, instead of asking abruptly.

Why we made these changes

For many teams, one of the most valuable outcome of a great conversation is what happens next: following up, segmenting shoppers, and building stronger lifecycle marketing. This update closes the gap between chat and email by making it easier to capture contact details in the moment and send them straight into Klaviyo without the manual work.

We’ve reorganized the editor to make the most important setup tasks easier to find and faster to complete. Instead of jumping between multiple tabs, you can now configure your agent’s goal, conversation starter, skills, and guidelines in one streamlined workflow, with appearance and performance details separated into their own clear areas.

What's new

• A clearer top-level editor: The editor now centers on three main areas — Agent, Appearance, and Insights — replacing the previous, more fragmented tab structure.

• One place to set up your agent: The new Agent tab brings together your goal, conversation starter, active skills, and guidelines so you can shape how the experience behaves without bouncing between screens.

• More control over the first message: You can now edit the opening conversation starter and the text input placeholder directly from the Agent tab.

• A no-suggestions starting experience: You can choose to start with just the chat input and no suggestions, which is useful when you want visitors to begin with their own question instead of selecting a prompt.

• Faster management for suggestions and skills: Page-aware suggestions and location settings now have clearer management actions, and skills can be reviewed from a compact summary before opening their detailed settings.

Why we made these changes

As experiences become more sophisticated, the old editor flow made simple setup feel more spread out than it needed to be. We rebuilt the editor around the decisions teams make most often: what the agent should do, how it should look, and how it’s performing. The result is a more focused setup flow that helps you go from idea to a polished experience with less navigation overhead.

You can now choose a custom font color for your Anagram experience directly within the Design tab. Set it once and every text element in the experience updates to match.

What's new

• Font color picker: A new color control in the Design tab lets you set the base text color for your experience. The color is applied consistently across elements in the experience so nothing looks mismatched.

• Hex preview: The selected color value is displayed inline next to the picker so you always know exactly what's set without guessing from a swatch.

Why we made these changes

Since Anagram experiences are embedded directly in your site, the experience needs to feel like part of the page. Background and primary colors get you most of the way there, but if your site uses a non-black text color, the experience would look off. Font color closes that gap.

This is part of a broader push to give you enough design controls to make Anagram feel native to your brand without needing to touch any CSS.

Different pages need different kinds of experiences. You can now choose how your Anagram experience appears so it better fits the context of the page, the amount of guidance a visitor needs, and the role the experience should play in the journey.

What's new

• More flexible presentation options: You can now choose from multiple ways for an experience to appear, depending on how prominent or lightweight it should feel on a given page.

• A better fit for different page types: Keep an experience fully present when it should lead the journey, introduce it more gently when visitors need a lighter touch, or make it available in a way that supports the page without overwhelming it.

• Easier ways for visitors to begin: Some experience layouts can now greet visitors with starter copy, suggested prompts, or an input directly on the page, helping them begin more naturally.

• A clearer expanded experience: When a visitor opens the full experience, it appears in a dedicated panel with clearer framing and controls, making it easier to understand and close when they are done.

Why we made these changes

An experience should match the page it lives on. On some pages, it makes sense for the experience to be front and center. On others, it should support the journey in a more lightweight way. This update gives teams more flexibility to shape that experience intentionally, so each page can offer the right level of guidance without forcing the same presentation everywhere.