July 31

Express Apple Pay Checkout Across Cart, Product Pages, and Funnels

Brands on the native cart can now offer one-tap Apple Pay checkout directly on cart pages, product detail pages, and landing pages, letting shoppers skip the multi-step checkout flow entirely. The integration runs on the platform's own payment controllers and works across Safari and iOS surfaces. On the first brand where it went live, pairing the native cart with express Apple Pay took checkout-start-to-purchase from 19% to 44%, with approximately 50% projected as Apple Pay coverage completes. The native cart is now the default onboarding path for all new enterprise brands.

July 31

Native Cart Enabled Site-Wide

The Fermat-powered native cart is now active across an entire retail site for a major sports nutrition brand, extending the faster checkout experience to every shopper regardless of entry point. Previously available on select surfaces, the org-wide rollout means no shopper encounters the legacy cart flow. This milestone also reflects the platform's broader shift to the native cart as the standard starting point for new enterprise onboarding.

July 31

Ad-Platform ROAS Anomaly Detection

Behavioral analytics anomaly detection now monitors Meta and Google ROAS alongside on-site metrics, giving brands a single place to catch revenue degradation whether it originates in the ad account or on the storefront. The system applies per-metric volume floors to hold back immature data and avoid false alarms, and teams can mark expected changes, such as a planned promotion winding down, to mute alerts that would otherwise create noise. Rolling out brand by brand.

July 31

On-Page Shopping Assistant with Conversation Analytics

The on-page shopping assistant now surfaces pre-populated questions tuned to the specific page a shopper is viewing, including upsell and cross-sell prompts, so interactions are relevant from the first tap. A refreshed widget design adds thumbs-up and thumbs-down feedback buttons, giving brands a direct signal on response quality. A new dashboard page shows the full usage funnel alongside attributed conversions, making it straightforward to measure whether the assistant is contributing to revenue. Rolling out brand by brand.

July 31

Merchandising Controls for Search and Grid Results

Editors can now pin or exclude specific products from search results and intelligent grid layouts using the storefront filter system, giving merchandising teams direct control over which products appear and in what priority. Alongside this, product search modules now surface real trending terms and live autocomplete suggestions rather than static placeholder text, making search feel responsive to actual shopper behavior. Together these controls close the gap between automated ranking and editorial intent.

July 31

Curated Product Sets in Intelligent Grids

Merchandisers can now source an intelligent grid from a hand-curated product set rather than relying solely on automated ranking signals. This gives editorial teams direct control in situations where algorithmic ordering is not the right fit, such as seasonal pushes, brand partnerships, or clearance events. The option sits alongside existing automated modes so teams can choose the approach that fits each placement.

July 31

Experiment History Preserved with Reusable URL Slugs

Ended external-split experiments are now retained in the platform with their full results intact rather than being removed after conclusion, giving teams a clean historical record of what has been tested and what it produced. URL slugs from completed experiments can be reassigned to follow-up tests, eliminating the need to create new paths and preserving any existing traffic or link equity. This makes iterative experimentation tidier to manage over time.

July 31

Usage-Based Pricing for August

August, the digital twins product, has moved to fully usage-based pricing with no base subscription fee. New organizations start with 1,500 free credits and pay only for what they use beyond that. Enterprise plans are unchanged. The change lowers the barrier to getting started and aligns cost directly with usage.

July 31

Fixes and Improvements

Storefront carousel editors now show product counts in category pickers, checkboxes respond correctly to clicks, and on-model photography extends to price-threshold carousel tiles, resolving a set of usability gaps that slowed merchandising work. Funnels now correctly signal consent initialization so third-party consent managers boot in the proper sequence, addressing a compliance gap for brands using tag-manager-delivered consent tools. Session semantic search received reliability and speed improvements, including elimination of a search hang and faster result retrieval, with results now linking directly to the relevant moment in a session replay.

July 24

Pierre Analytics Answers Rebuilt and Faster

Pierre's analytics engine has been fully rebuilt and is now the default for all data questions across every brand with dashboard access. Answers arrive roughly 3 times faster at about a third of the previous cost. Pierre now computes only what it can verify, asks clarifying questions when a query is ambiguous, and declines to answer rather than guess when data is insufficient. This is a general availability release affecting all brands immediately.

July 24

Shopper-Facing Grid Filters in Product Funnels

Product grids inside funnels can now offer shoppers filters for brand, product tags, price range, and on-sale status, all configured per funnel from the dashboard without engineering work. Promo badges also render on products, respect their active date windows automatically, and function as live grid filters, so expired deal labels retire themselves and shoppers never see a stale promotion. Both capabilities are generally available to brands using the product grid in their funnels. Large-catalog and outlet funnels in particular benefit from giving shoppers a way to narrow results without leaving the funnel.

July 24

Shopping Assistant Impact Measurement

Brands can now see exactly what the on-page shopping assistant is driving through a new impact tile in the dashboard. The tile surfaces the assistant's full usage funnel alongside directly attributed conversions, giving a clear view of its contribution to purchase activity. This is a generally available enhancement available to all brands running the shopping assistant.

July 24

Recently Viewed Products in Cart

Carts can now display a shopper's recently viewed products, creating an additional path back to items they considered before adding to cart. The capability is configured at the storefront level and is generally available. It is designed to reduce drop-off for shoppers who browsed multiple products before committing.

July 24

Cross-Brand Attribution Accuracy for Multi-Brand Organizations

A fix to journey-scoped attribution eliminates cross-brand attribution leakage for organizations operating multiple sibling brands under one parent account. Each brand's results now reflect only its own traffic, so performance reporting is cleanly separated. This is a generally available enhancement affecting multi-brand accounts.

July 24

Live Inventory Gating in Funnels

Funnels now sync live stock levels and hide out-of-stock products, preventing shoppers from landing on items they cannot purchase. Inventory status is checked in real time and the product grid updates accordingly. This is a generally available enhancement rolling out to brands on supported inventory integrations.

July 24

Split-Shipment Inventory Fallback at Checkout

Orders can now fall back to a secondary inventory location when a primary location is out of stock, keeping checkout available rather than blocking the purchase. This reduces friction for shoppers at the point of conversion when inventory is unevenly distributed across fulfillment locations. The capability is generally available.

July 24

Behavioral Cohort Analytics Now Available to Brand Users

The User Cohorts dashboard automatically groups shoppers by behavioral patterns, such as high-intent non-buyers or repeat browsers across categories, and surfaces metrics, side-by-side comparisons, and linked session recordings for each group. Brand users can now access these views directly, without requiring internal support. The capability is rolling out brand by brand, with custom cohort definitions configurable per account.

July 24

Research Surveys and Summaries Available in Slack

Brands can now run surveys, receive summarized result cards, and ask follow-up questions directly inside Slack by mentioning the research assistant in any channel. Research artifacts including completed reports can be shared read-only across a brand's workspace, making findings accessible to broader teams without requiring a separate login. This capability is rolling out brand by brand.

July 24

Fixes and Improvements

Cart resilience was improved so shoppers with an expired stored cart now receive a fresh cart automatically rather than an error, and cart IDs persist correctly across browser tabs. Session recording on Shopify storefronts now stops immediately when a shopper explicitly rejects tracking consent, bringing behavior in line with consent signals. The module editor grid layout no longer breaks when category names are long. Multi-platform support was added to the email and SMS capture modal, so brands are no longer limited to a single retention provider for list growth.

July 17

AI-Powered Shopper Surveys, Generally Available

Brands can now run end-to-end shopper research directly from the platform. A co-pilot helps draft survey questions based on what a brand wants to learn, the survey is launched to shoppers, and results are returned as a clean, print-ready report. Every answer in the report includes a transparent evidence trail showing exactly what it is grounded in. The full workflow, from question drafting to finished report, is available to all brands today.

July 17

AI Search Now Ranks by Generated Product Categories by Default

The platform's search ranking now incorporates AI-generated product categories, so shoppers using broad terms get better-organized and more relevant results. A search for a general category like rain gear, for example, surfaces results grouped by intent rather than returning a flat, unstructured list. This strategy is now the default for all search-powered storefronts and required no action from brands to activate.

July 17

Memory and Question Reframing in the Research Assistant

The AI research assistant now surfaces remembered brand goals as chips above the composer, giving users a quick view of the context it is working from before they ask a question. The question reframer, which turns vague or broad asks into precisely researchable questions, is now on for all users. Together these changes make it faster to ask well-formed questions and easier to pick up where a previous session left off.

July 17

Pierre Gives More Reliable Funnel Performance Answers

Pierre's responses to funnel performance questions now draw from a cleaned data source, reducing the chance of misleading or inconsistent numbers. When a question falls outside what it can answer reliably, Pierre will decline rather than guess, and it now asks for clarification when a question's scope is ambiguous. These changes apply to all brands with dashboard access.

July 17

Behavioral Cohorts Dashboard Rolling Out

A new dashboard section automatically groups site visitors into behavioral cohorts, such as high-intent shoppers who have not yet purchased or repeat browsers who span multiple categories. Each cohort surfaces metrics, side-by-side comparisons, and linked session recordings so teams can move from page-level data to people-level understanding. The dashboard includes list, detail, and comparison views. This capability is rolling out brand by brand.

July 17

Better Marketing Attribution and More Complete Purchase Data

Session attribution now covers all pixel sources, with improved referrer handling and a new referrer-host classification tier that reduces sessions incorrectly bucketed as direct traffic. Campaigns arriving via email or SMS links are more consistently credited to the right channel. Separately, purchase records that previously had missing totals or line items are now enriched from order data rather than dropped, improving revenue accuracy across reports.

July 17

Cleaner Dashboard Labels for Session Data

Two navigation labels in the analytics dashboard have been updated for clarity. The section formerly called Commerce Graph Logs is now Session Logs, and the section formerly called Behavioral Analytics is now Session Replays. A recent-replays table has also been added to make it faster for brand users to find recordings without additional filtering.

July 17

Per-Variant Offer Targeting and Campaign Link Attribution Fix

Offers can now target specific product variants rather than only whole products, enabling size- or color-specific promotions without creating separate offer rules for each product. In a separate fix, redirect links now carry the full destination query string, so UTM-tagged campaign links preserve their attribution parameters through the redirect rather than losing them.

July 17

Fixes and Improvements

Experiment results on storefronts now aggregate at the experiment level and display the statistical-significance indicator on the correct arm, making win and loss readouts cleaner and easier to interpret. A cart-drawer variant that outperformed in testing has been promoted to the default experience. AI Search funnel traffic is now excluded from overall dashboard performance totals, keeping core funnel metrics and search-page traffic separate. A cart-handoff bug on Salesforce Commerce Cloud storefronts that could empty a shopper's bag when editing it at checkout has been fixed, and bridged checkout sessions are now single-use before delivery to the shopper, closing that class of edge cases. Custom modules now render correctly in the storefront editor preview pane.

July 9

Friction Points

Every brand on FERMÀT's behavioral analytics now gets Friction Points, a ranked feed of the specific page elements frustrating shoppers, including rage clicks, dead clicks, and error clicks. Each item links to session replays, so your team can watch real shoppers hit the problem. It is on by default, with no setup required.

July 9

Site Error Logs with Slack Alerts

Session Logs give you a live feed of storefront errors, including JavaScript errors and broken interactions, with device and country breakdowns so you can see who is affected. You can create alert rules and route notifications to your own Slack channel, so your team hears about a broken experience as it happens. Connect your workspace from settings and set per-rule alerts.

July 9

Dedicated Search Engine for Storefronts

Storefront search, product listings, and category pages now run on a purpose-built search service, so relevance improvements land across every surface at once. Product-list filters are now real checkboxes with working multi-select, search suggestions support arrow-key navigation, and pressing Enter in grid search works as expected. Every search result click is also tracked with its position, giving the data to tune relevance per brand.

July 9

AI Shopping Assistant

The on-page shopping assistant now answers shipping, returns, and warranty questions from a brand's own help-center content, in addition to helping shoppers find the right product. It can sell consultatively, asking clarifying questions and guiding shoppers to a recommendation, and it presents as a proper bottom sheet on mobile. Whether and how it appears on a site is controlled entirely from the dashboard, with no brand-side code changes.

July 9

Fixes and Improvements

Several improvements are now live. Build-your-own-bundle now lets shoppers add the same product more than once, with cart quantities displaying correctly for bundle lines. Sticky add-to-cart buttons, badges, and media galleries no longer overlap incorrectly on product pages. A prepared-cart checkout race condition is resolved for brands on the Rithum platform. Pierre no longer declines questions it can answer, the session recordings list no longer flashes during infinite scroll, and reports can now be exported as PDF.

July 3

More Accurate Order Attribution

Order attribution accuracy has jumped from roughly 25% to nearly 100% on recent orders. A new persistent visitor identity layer stitches pixel events and session recordings together, so channel and campaign reporting reflects almost all orders instead of a quarter of them. Sessions missing UTM tags are now classified using ad-platform click IDs, so paid traffic is credited correctly instead of being grouped into other.

July 3

Native Cart Handoff

Carts built on FERMÀT experiences can now sync directly into a brand's own cart and checkout, including subscription products, so shoppers no longer hit a separate checkout when they buy from a FERMÀT page. This removes one of the biggest points of friction in main-site funnels. Rolling out brand by brand.

July 3

Fixes and Improvements

Several fixes are now live. Popularity signals now flow from constituent products to virtual products for better ranking, and virtual products no longer disappear from brand-filtered search. Web-pixel capture now retains full order totals on Shopify, improving revenue accuracy. And page-type detection now handles locale-prefixed URLs correctly, so analytics no longer mark large portions of a site as unknown.

July 1

Introducing august: Digital Twins of Your Real Customers

august is live: a human prediction platform powered by Digital Twins of real customers. Each twin maps to a single real shopper and is built from real behavioral session data, so brands can run interviews and focus groups with the people who actually buy from them and get answers in seconds instead of weeks. Twins recall memories from every real interaction and form higher-level reflections over time, and they push back on overconfident claims rather than agreeing. Live now at getaugust.ai.

Product Documentation

June 25

Lead-Capture Forms in Custom Modules

Custom modules can now capture and route form submissions, so brands can launch lead-capture, contact, and waitlist forms directly inside a funnel with no custom development. Submissions are delivered by email, turning a storefront into a lead-generation surface.

June 25

Richer Storefront Content

Storefronts now pull in far more product and brand content automatically, including reviews, size charts, product attributes, and policy details like returns, price guarantee, and free shipping. Shoppers see richer, more accurate detail on the page without manual setup per product.

June 25

Fixes and Improvements

Several fixes are now live. Funnel navigation now loads the selected product instead of occasionally falling back to the template hero. A Shopify catalog sync bug that could silently drop an entire product batch on a single parse error is resolved, preventing catalog gaps. And the funnel editor no longer crashes when a module references a deleted product.

June 18

Trending Search Suggestions

Storefront search bars now surface top and trending queries as shoppers type, so they reach products in fewer keystrokes. Suggestions are drawn from real engaged-session demand, so they reflect what shoppers are actually looking for rather than a static list. Live for all brands.

June 18

Put FERMÀT-Built Modules on Your Own Site

Module Studio can now generate a copy-paste snippet for any FERMÀT-built module, so you can drop it directly onto your main storefront, not just inside a FERMÀT funnel. The comparison tables, finders, and quizzes built for your funnels can now live on a PDP or category page on your own site, with no engineering lift. Open Module Studio, select any saved module, click Generate Snippet, then copy and paste the snippet into the page where you want it to appear. Live for all brands.

June 18

Storefront Fixes and Improvements

Several storefront fixes are now live. Promotions targeted to a specific customer group are no longer shown to all shoppers. A disabled upsell drawer no longer opens on add-to-cart. And body tiles no longer collapse when a reviews anchor sits inside the hero.

June 12

Self-Serve AI Search Onboarding

Brands can now set up AI Search through a guided onboarding flow at /onboarding/ai-search. The flow walks teams through the required setup steps, including ecommerce connection, knowledge base context, suggested edits, and handoff details, so they can get started without waiting for engineer-led configuration. This makes AI Search easier to launch for new brands and self-serve teams.

June 12

Direct Checkout on Add-to-Cart

Product Funnels can now send shoppers directly to checkout after they add an item to cart. The new template setting skips the cart step for high-intent buying moments, while leaving normal "Open PDP" behavior intact. Brands can use it to test a faster path to purchase on conversion-focused funnels.

June 12

Smarter Product Search

Product search now handles misspelled brand names, boosts exact attribute matches, and returns the specific matching variant when shoppers search by size, color, or other product details. Search also uses stock status and newer-product logic to keep results relevant across large catalogs. This reduces dead-end searches and helps shoppers get to the right product faster.

June 4

AI Search Performance in the Dashboard

AI Search customers can now view AI visibility and performance directly in the FERMÀT dashboard, without waiting on a manual report from the FERMÀT team. This makes it easier to track performance over time and stay current on where AI Search is driving visibility and funnel impact.

Faster Storefront Product Search

Storefront product search now reads directly from FERMÀT's search index, reducing latency and helping shoppers get results faster, especially on larger catalogs. This improves the browsing experience anywhere shoppers use search across FERMÀT-powered storefronts.

May 28

Smarter Search Ranking

Two improvements to how search results are ranked. Products whose most popular variants are out of stock now down-rank and are filtered out of result hydration, so shoppers spend less time clicking into dead-end pages. Ranking also factors purchase signal weighted by product margin, surfacing results that both convert and contribute meaningfully to revenue.

May 28

Session-Aware Product Recommendations

Recommendation carousels now factor in the product a shopper is currently viewing and recent in-session signals when sourcing recommendations, instead of treating each surface independently. Recs feel more relevant for brands with broad catalogs, where what to show next changes meaningfully across categories.

May 28

Per-Tile Direct Checkout on Shopify

Direct-checkout add-to-cart (skip-cart, straight to checkout) can now be turned on for a single tile through the dashboard, instead of requiring an account-wide setting. Useful for high-intent campaigns, paid social funnels, and bundle pages where the cart step is friction.

May 22

Custom Modules: Drag-and-Drop Image Generation + AI Style Adjustments

Custom Modules now support image-based design generation and AI-powered style adjustments, letting brands create and iterate on custom module designs directly in the dashboard without engineering help. Drag and drop an image onto the funnel preview within a Custom Module image drop area, or upload an image via the paperclip icon in the Pierre chat to use it as context when iterating on your design. Style adjustments are made through Edit with AI, where Pierre modifies existing configuration rather than hardcoding values. Available to all customers.

May 21

AI Search Onboarding Improvements

AI Search onboarding is now a single conversational flow instead of a multi-step wizard. The new flow automatically extracts brand voice, styling, prompts, and ecommerce connection from a brand's website, replacing what used to be a 16-step setup. New AI Search customers get to a working article materially faster, and existing customers adding sibling brands or sub-lines can do so without the previous setup lift.

May 21

AI Search Analytics Dashboard

A new dashboard for AI Search performance with three tabs: By Opportunity, By Funnel, and Category Breakdown. Brands get a visual read on where AI Search is driving impact across their catalog, alongside multi-headline weekly wins surfaced directly in the view. Loading and empty states have been tightened so the dashboard reads cleanly from first session.

May 21

Configurable Recommendations on Category Grids

Category Grid modules now pull from a dedicated recommendations endpoint with configurable strategies, including curated, co-purchased, and co-add-to-cart. Brands can choose which logic surfaces in each category module instead of relying on a single global default, giving merchandising teams direct control over what shoppers see in category browsing surfaces.

May 14

Multi-step A/B Tests

Brands can now run A/B tests across full multi-page sequences, not just single pages. Variant attribution carries cleanly through to checkout, so teams can compare end-to-end journey performance without losing signal at handoffs.

May 8

Desktop Layout Modes: Full Width or Contained per Funnel

Teams can now choose between two desktop layout modes for any funnel, Full Width (edge-to-edge) or Contained (centered with a max-width preset from 1024px to 1920px). New funnels default to Contained / Standard (1440px); existing funnels stay on Full Width. Each module also has a new Layout tab so brands can override the funnel-level setting on a per-module basis, for example a full-bleed banner inside a contained body.

May 5

Storefront Iconography: Brand-Switchable Icon Sets + Custom Uploads

Brands can now pick the icon set that matches their visual identity across the entire storefront. Choose between two new built-in sets, Soft (rounded, friendly) and Sharp (angular, bold), or upload your own SVGs for full creative control. 43 canonical icons per set cover every interactive surface, cart, navigation, search, filters, variant selectors, media controls. Set the iconography under Settings → Branding → Iconography. Custom uploads include a graceful fallback so storefronts never render broken icons.

May 1

Rich Text Inputs across Banner, HeroHeader & Cart Summary

Storefront copy fields in Banner, HeroHeader (header, sub-header, review, review sub-text), and Cart Summary's free-shipping success text now support inline rich text. Use bold, italic, underline, strikethrough, and inline links from a stripped-down Medium-style toolbar. Links get an optional display-text override and a per-link "Open in new tab" toggle. Existing plain-text content renders unchanged, no migration required. Hero Header copy fields also pick up native line breaks via shift+enter, replacing the old <br> workaround.

April 24

Pierre Chat as the Dashboard Home Page

Upon arriving to app.fermatcommerce.com, brands land in a chat with Pierre, FERMÀT's AI copilot, instead of a data-dense home page. Ask Pierre about mobile conversion drops, AI search visibility, or test performance and get the answer with supporting session recordings, heatmaps, and dashboards inline. Action items for new anomalies, opportunities, and weekly reports are easily accessible via the home dashboard.

April 24

Full-Page PDPs on Desktop

Desktop product pages now open as full-width, two-column layouts instead of side drawers. The new design anchors media on the left with a flowing details panel on the right, bringing the desktop experience closer to what shoppers expect from modern brand sites in 2026. Rolling out one brand at a time, Glossier first.

April 23

Custom Quizzes Powered by Product and Performance Data

Brands can now generate quiz and finder modules grounded in real product data and recent performance signals, instead of treating the catalog as flat. Quizzes pull product attributes, categories, and 30-day performance to favor stronger products in their recommendations. Generate, iterate, and place them in your storefront layouts.

April 22

Prefilled Carts

Brands can now generate special links that preload specific products into a shopper's cart when shoppers click through from an ad or influencer post, turning a multi-step browse-and-add flow into an instant checkout-ready experience. A notification banner confirms the items were added. URL-parameter based, so any existing forever link works by appending the products and quantities.

April 22

Lifestyle Images Hover Effect

Product Grid cards can reveal a per-product lifestyle image when shoppers hover over them, configured per product in the dashboard. The default behavior remains the existing carousel arrows on hover.

April 20

Faster Order Tracking Setup for Pixel-Based Brands

Pixel-based purchase tracking on a new ecommerce platform is now a one-step config toggle, replacing what used to require around 5 days of integration work per platform. Speeds up onboarding for brands on platforms outside Shopify.

April 15

Static IP Routing for IP-Whitelisted Integrations

Brands that require IP whitelisting to grant access to their APIs can now be served from a static IP, removing a recurring blocker in setup.

April 13

Storefront Motion Effects

Marquee banners, animated badges on product grids, and discount price animations on product pages are now native module options. Use them sparingly to draw attention to what matters; one well-placed animation lands harder than several.

April 13

Compact Add to Cart Button

Product grids now support a compact icon-overlay Add to Cart button that sits on the product image instead of taking a full-width slot below it. This is the new default and keeps grids cleaner, especially on mobile.

April 13

AI Media Generation in the Carousel

Brands can now generate on-brand product imagery directly from the Media Carousel editor by selecting a variant or style, choosing a theme, and writing an image prompt. Generated images go through an approve or reject review flow before being saved as product media, helping brands fill in lifestyle and hero imagery for SKUs without vendor or studio shoots.

Product Documentation

April 7

Free Shipping Success Text Configurable

The free shipping success message in Cart Summary is now fully customizable per-brand. Previously hardcoded, you can now enter your own message in the Layout Editor when "Show Free Shipping Progress" is toggled on. Leave it blank to keep the default.

April 7

Upsell Cards Refresh

Cart page upsell cards have been visually overhauled with a cleaner layout, discount badges that overlay the product image, a column layout on desktop, and larger CTA buttons for improved tap targets and upsell conversion.

April 6

AIS Reddit Outreach Opportunities

Reddit is a major source of AI citations, and the same threads show up across weeks and months in LLM responses. AIS Reddit Outreach Opportunities identifies the specific Reddit threads that AI models consistently cite when answering your tracked prompts, so your team can engage strategically and influence how AI talks about your brand. Available for AIS Enterprise accounts.

Product Documentation

April 6

AIS Refresh Flows

LLMs have a recency bias, content that hasn't been updated in roughly 6-7 weeks starts losing AI crawler activity and citations. AIS Refresh Flows automatically detect when your AI-optimized shops are declining and regenerate the content to restore visibility, keeping your brand competitive in AI search without manual effort. Available for AIS Enterprise accounts.

Product Documentation

March 27

Search & Discovery Anomalies

Three new anomaly types automatically detect when on-site search fails shoppers: Search Abandon (users view search results but never click a product), High-Intent Underperforming (users click a result and view the product but don't add to cart), and Search Loops (users search repeatedly in rapid succession). Each anomaly generates a full report with per-query breakdowns, behavioral signals, session recordings, and recommended actions, surfaced alongside existing behavioral anomaly reports.

Product Documentation

March 17

AI Search Opportunities Dashboard

FERMÀT's AI Search engine generates AI-optimized pages and citations are rising across AI search platforms. The Opportunities Dashboard is the next layer, it analyzes prompt trends, competitor citations, and content gaps to deliver prioritized weekly recommendations with auto-generated articles. Go from opportunity to live content in one click. Available for all full-tier AI Search brands with ~4 weeks of prompt history.

Product Documentation

March 17

Heatmaps in Behavioral Anomaly Reports

Every Behavioral Anomaly Report now includes a Heatmap Evidence section, see exactly where users are clicking (or not clicking) directly within the report. Visually validate rage clicks, dead scrolls, and confusion loops without scrubbing through individual session recordings. Toggle between heatmaps and session recordings in the same view.

Product Documentation

March 14

Build Your Own Bundle Updates

Bundle Builder now supports description fields so shoppers can see what's included, direct product page integration so bundles can surface inline on PDPs without a separate funnel, and a series of experience improvements across the creation flow.

Product Documentation

March 13

AI Search FAQs & Knowledge Base Access

Updated AI Search documentation now covers external knowledge base access, team members without a full FERMÀT login can browse content, view metadata dates, and check funnel approval/rejection status. Plus new FAQs on prompt targeting, ROI measurement, and how AI Search fits alongside Funnels and DPP.

Product Documentation

March 6

CSS Customizations & Custom Module Builder

Build entirely new modules with Pierre by describing what you want in plain language, or restyle any existing module using Edit with AI, no code, no dev ticket. Custom modules are saved to your org's library and reusable across every funnel. Available in the funnel editor across all surfaces.

Product Documentation

February 27

Unified Analytics Dashboard

FERMÀT's analytics have been rebuilt into a single, unified dashboard, bringing together Experiments, Funnel Analytics, and Incremental Impact under one roof. No more jumping between tools: select your funnels, pick your metrics, and get the full picture across CVR, ROAS Lift, experiment comparisons, and ad-level performance, all with shared filters in one view.

February 18

FERMÀT MCP Server (Beta)

Connect Claude AI directly to your FERMÀT data with the new MCP server. Available for Claude Desktop and Claude Code users, you can query funnel performance, analyze campaign data, and get AI-powered insights about your store, no manual exports required. Connect via Claude's Connectors settings using your FERMÀT platform credentials.

February 17

Module Visibility Controls

Show or hide individual funnel modules by device type, mobile, desktop, or both. Control which modules appear on each screen without duplicating your funnel. Works across landers, PDPs, and BDPs.

Product Documentation

January 7

Interactive Hotspots in DPP Heatmaps

DPP Heatmaps now show interactive hotspots highlighting the areas users click most on your product pages. Each hotspot displays a click count; select it in the side panel to jump to its exact location on the heatmap. Available under DPP → Behaviour Analytics → Heatmaps.

January 5

Scroll Depth in DPP Heatmaps

Track how far visitors scroll through your DPP templates, measured in 5% increments up to 100%. Identify exactly where users drop off so you can optimize layout and section placement. Available under DPP → Behaviour Analytics → Heatmaps → Scroll.

December 24

Heatmaps for DPP

Aggregated click heatmaps are now available for DPP product groups used as landing destinations for PLA and DPA campaigns. Filter by time, template, category, device, channel, and campaign to see which layouts drive the most engagement, turning product-group experimentation into data-driven decisions. Available under DPP → Behaviour Analytics → Heatmaps.

December 11

Session Recordings for DPP

Session Recordings are now available inside DPP. Navigate to DPP → Session Recordings to see sessions mapped to each product group, with full context including device, traffic source, destination, template, and behavioral tags like rage clicks. All standard funnel filters apply, plus a new template filter to narrow by DPP template.

November 14

Session Recordings

Watch real user behavior across your funnels with replayable session videos. Filter by device, traffic source, user actions (viewed, added to cart, checkout), and UX insights like rage clicks. Available for the last 60 days of sessions, accessible from your Homepage or Funnel Edit page.

Product Documentation

November 6

Pierre File Upload Support

Pierre now accepts file uploads to better understand your requests. Attach PDFs, images (JPEG, PNG, WebP) as reference materials when building funnels or bundles. Upload multiple files up to 50MB total to give Pierre the full context needed for more accurate and tailored results.

November 6

Device-Based Traffic Splitting

Split traffic based on device type (mobile vs desktop) to create optimized experiences for each platform. Route visitors to different FERMAT funnels or external destinations based on their device. This FFL routing feature works seamlessly alongside existing geo and weight-based routing options.

October 29

Pierre Conversation Management

Directly tag products, bundles, and funnels in your Pierre conversations for more precise AI assistance. Use the plus button to add specific products or funnels to your messages, eliminating guesswork when referencing items in your catalog.

October 15

Pierre Conversation Management

You can now manage multiple conversations with Pierre. Create new chats directly from the sidebar and easily toggle between conversations to keep different projects organized. Rename conversations for easy reference or delete ones you no longer need.

October 3

Pierre AI Updates

Bundle Builder is now integrated into Agent Mode for a seamless workflow. Funnel Builder has been enhanced with BI insights access and web scraping capabilities, allowing you to create funnels directly from Meta ad copy, blog posts, and other web content.

September 26

Funnel & Bundle Building with Pierre

Pierre can now automatically create high-converting funnels and intelligent product bundles tailored to your brand and catalog. From funnel layouts to bundle pricing, Pierre speeds up creation from hours to minutes, while ensuring brand consistency and conversion best practices.

September 22

GA4 + Google Ads Data Sync

You can now sync Google Ads performance data alongside GA4, giving you the full picture of your funnel, from ad spend to clicks to on-site behavior to conversions.

September 18

Heatmaps: Clicks & Scroll Depth

You can now track where users click and how far they scroll in your funnels with new Heatmap insights. Easily spot hotspots of interaction and see which sections drive the most engagement, or where users drop off.

September 15

Navigation Bar Update

We've reorganized the left-hand navigation bar to make it easier to explore FERMÀT's growing suite of apps, including Funnels, AI Search, and Pierre. Standard tabs like Media Library and Product Catalog remain consistent across apps, while app-specific tabs (Offers, Links, PDPs) are now grouped within their respective apps.

September 5

GA4 Flow Migration

We've upgraded GA4 syncs to Airbyte for smoother performance, greater reliability, and full visibility into data status, no disruption to your existing funnels.

September 3

Custom Font Styling

Brands can now upload custom fonts and fine-tune typography across all FERMÀT experiences. Configure up to 6 text styles with granular controls for font size, weight, line height, and spacing to ensure perfect brand alignment on every page.

Product Documentation