The 2026 selling season is about to look very different. We’re at peak adoption of consumers using LLMs, like ChatGPT, Copilot, Gemini, and Perplexity, to shop. Because of this, the consumer journey and path to conversion are now more complex. The trust consumers place in LLM responses to their shopping prompts is surprisingly high, with most shoppers relying entirely on the AI to shape the decision all the way to purchase. 

95% of shoppers who get a product recommendation from an LLM won’t visit a brand or retailer website to verify it. Among 18–27 year olds, 64% buy on the recommendation alone.

Twelve months ago, most brands were still focused on ranking in traditional search results. Consumer shopping behavior has changed more in the last 12 months than in the previous five years combined. That shift is going to define how every peak event plays out in 2026. 

Amazon and Walmart are blocking LLMs from crawling their product data. So, keeping consumers within the Amazon ecosystem to maximize Prime Day returns and overall conversions is critical. 52% of all U.S. consumers still begin their online product searches directly on Amazon,^¹ which is reassuring for Prime Day activity. But when we factor in the LLM component, our strategy to maximize Prime returns gets more nuanced. 

Previously, we wouldn’t have as heavily considered our clients’ .com websites as part of the Prime Day journey, but if LLM recommendations are relying on Google and other general product data feeds, then ensuring product data hygiene from those outside sources is critical. 

How should brands reach shoppers on Amazon and LLMs at the same time?

• For consumers already starting their journey in Amazon search, keep them engaged, targeted with appropriate content and ads, and slam dunk the conversion. 
• For consumers starting their search in LLMs, make sure client .com product data feeds are accurate, updated, and organized so the right recommendations entice consumers to seek out the shopping moment outside the LLM. 

There’s more to this year’s approach. Prime Day lands June 23–26 this year, the earliest window since 2021 and the second year at four full days.¹

Back-to-school demand is already overlapping with it. BFCM is less than six months out. And the LLMs shoppers are using to find and compare products have improved significantly since last peak season.

LLMs are pulling richer product details, making more accurate comparisons, and starting to support transactions directly. The brands and retailers who will perform best during peak 2026 are the ones whose catalogs are accurate, structured, and available everywhere the AI is looking. 

At Rithum, we work across 900+ channels with brands and retailers preparing for these events right now. We’re deep in the planning alongside them, from media strategy to product data to channel prioritization, and everything below reflects what we’re seeing and recommending heading into peak selling in 2026. 

What are the major 2026 peak sales events, and when do they happen? 

• Amazon Prime Day: June 23–26¹ 
• Back-to-school: May through September (peaks during and after Prime Day) 
• Labor Day: September 1 
• Amazon Prime Big Deal Days: October (dates TBD; October 7–8 in 2025) 
• Singles’ Day: November 11 
• Black Friday: November 27 
• Cyber Monday: November 30 

Beyond the timeline, two things connect every event on this list: AI is reshaping how shoppers discover products, and the brands with accurate, structured data across channels will capture more demand than those without it. 

How should brands prepare for Amazon Prime Day in June 2026? 

Prime Day runs June 23–26 across 35 categories, with deals dropping three times daily and new inventory going live every five minutes during peak hours.¹ Amazon is putting particular emphasis on fresh groceries and household essentials, building on 4 billion grocery and essentials items delivered via same-day delivery over the past year.² 

Last year’s Prime Day (July 8–11) generated $24.1 billion in U.S. online spend, up 30.3% year over year.³ Four of the top five items sold were household or grocery products.⁴ School supplies rose 175% and dorm essentials 84%.³

Across our own client base, the four-day format shifted buying patterns. Shoppers browsed Days 1 and 2, then converted late, with Day 4 GMV jumping 38% year over year. 

One Rithum client pivoted mid-event to back-to-school bundles and optimized product titles on Day 2. That shift drove a 15% sales lift over the prior year. The teams that adjusted in real time outperformed the ones that set a plan and left it. And that was before AI was deeply embedded in the Prime Day experience. 

What role does Alexa AI play in Prime Day 2026? 

This year, Alexa for Shopping (formerly Rufus) will sit directly in the Amazon search bar, search results, and product detail pages for every U.S. customer.¹ Amazon has told sellers that Alexa pulls from listing content to answer shopper questions and make recommendations. If your product attributes are incomplete, Alexa will recommend someone else’s.

During Prime Day 2025, shoppers who engaged Rufus were 60% more likely to complete a purchase. Alexa now helps shoppers build personalized deal guides, set price alerts, trigger auto-buy at target prices, and review 365-day price history. Every one of those features pulls from your product data. If your titles, descriptions, and images aren’t accurate, you’re not showing up in those conversations. 

Rithum’s retail media advertising connects spend to product performance across Amazon, Walmart, and Target so you can shift budgets while the event is live. Marketplace listings keep data accurate across 900+ channels so you can coordinate pricing, inventory, and promotions across Amazon and every channel competing for the same shopper during Prime Day. And inventory management prevents oversells and stockouts during the exact hours demand spikes. 

How does AI shopping change peak season strategy in 2026? 

When AI gets product information wrong, 58% of shoppers lose trust in the brand itself, not the AI tool that served it. And shoppers aren’t loyal to a single AI platform either: ChatGPT’s daily active user share has fallen steadily since early 2025, Gemini has more than doubled its share in the same period, and 20% of AI users now regularly use two or more chatbot apps. Brands that only optimize for one platform are leaving reach on the table. 

The shopping capabilities on these platforms have also matured. Perplexity offers one-click checkout through Buy with Pro and PayPal integration across thousands of merchants. Gemini supports agentic checkout through Google Pay, including auto-buy when a price target is met.

ChatGPT launched native shopping features in late 2025. These are live commerce channels, and every one of them relies on structured product data to generate accurate recommendations. 

We built ChatGPT and Perplexity Feeds for this: product data compiled, optimized for LLMs, and delivered for ingestion so your brand controls how it appears in AI recommendations. Our upcoming GEO (generative engine optimization) pilot provides a way to track how products appear, move, and compare within AI-driven results. Rithum and Stripe are also building the commerce infrastructure for transactions that happen inside the AI conversation itself. 

The path from catalog accuracy to LLM delivery to in-conversation purchase is what determines whether your products get recommended during peak season. Every step depends on the one before it, and most brands are still missing at least one. 

How early does back-to-school shopping start in 2026, and what should sellers prioritize? 

Back-to-school retail is projected at $85.42 billion in 2026, up 3.3% year over year.⁵ 67% of families started buying by early July 2025, up from 55% the year before. With Prime Day now in June and school supplies already among its top-performing categories (up 175% during Prime Day 2025³), back-to-school effectively starts during Prime Day. 

84% of parents say prices are too high. 82% are prioritizing value over brand. But 80% still say buying the specific products their child wants is important.⁵ Parents are cutting back on clothing and switching to store brands in some categories while spending full price in others. Brands that understand which of their products fall into which bucket can plan promotions accordingly. 

We recommend spreading promotions across May through September rather than concentrating spend in a single burst. Rithum’s product catalog feeds push near-real-time updates to price, stock, and product data across every channel where shoppers and AI agents are looking. For brands selling across multiple regions, automatic localization for language, currency, and regional specs keeps listings accurate without rebuilding feeds from scratch. 

What should brands plan for across Q4 2026 peak events? 

September through November packs five major sales events into 90 days, and performance in each one can help shape the strategy for the next. 

Labor Day (September 1) — The last major sales moment before Q4 takes over.

Brands use it to move seasonal stock in a smart way. They set pricing rules based on demand signals, not flat markdowns. This helps them carry less dead stock into the holiday rush.

Rithum automates promotional updates across marketplaces. Order management routes to the best fulfillment location based on your rules. 

Prime Big Deal Days (October) — Amazon’s separate October event, distinct from summer Prime Day.

The 2025 event (October 7–8) outperformed the prior year globally.³ Shoppers complete holiday purchases here, and the performance data directly informs Q4 decisions. Brands that connected product feeds to LLM platforms earlier in the year will have compounding advantages by October.

Singles’ Day (November 11) — $202 billion across major Chinese platforms in 2024, making it the largest single shopping event globally.³

The event continues to expand into Southeast Asia, where Chinese ecommerce players now account for up to 50% of B2C GMV in key markets. U.S. and European retailers saw an 11% year-over-year lift in influenced revenue last November. Rithum supports 600+ global marketplaces with built-in localization and dedicated account strategists across EU, US, and APAC. 

How should brands prepare for Black Friday and Cyber Monday 2026? 

Black Friday (November 27) and Cyber Monday (November 30) — BFCM 2024 delivered $10.8 billion (Black Friday) and a record $14.3 billion (Cyber Monday 2025).³

Promotions started climbing November 16 last year, nearly two weeks early. Mobile drove 56% of online holiday spend last season.³ AI-assisted shopping volume, which surged during Prime Day 2025, will be even higher by November as more consumers adopt LLM-based product discovery. 

During BFCM 2025, we advised clients to spread budgets from October through December. Clients increased spending by only 3% on average and still hit efficiency goals with a 1.7 advertising cost of sales. Here’s how those Prime Day lessons apply to holiday planning. 

Across all of these events, the operational requirements are the same: accurate product data everywhere you sell, budgets that can move in real time, and listings that don’t break under pressure. Rithum’s product catalog feeds deliver data to every channel and AI platform in the right format. Paid search and shopping ads optimize across Google, Microsoft, and Meta with feed-level precision. And marketplace listings catches broken or non-compliant listings before they cost you revenue during the highest-traffic weeks of the year. 

How do you prepare for peak season across 900+ channels at once? 

Our 2026 Commerce Readiness Index found that 43% of retailers and 37% of brands are prioritizing AI enablement across operations. The best teams have one thing in common at their busiest times. Product data, ad spend, inventory, and fulfillment all run in a single connected system. When something changes mid-event, they can act right away. 

Rithum brings all of that into one platform across 900+ channels and AI platforms. If you’re planning for peak 2026 and want to talk through your approach, our team is here. 

Footnotes
¹ Amazon Press Release, “Mark Your Calendars: Amazon Announces Prime Day Event from June 23–26,” June 2026. press.aboutamazon.com
² Forbes, “Amazon Prime Day 2026 Moves to June—This Time With Alexa AI Powering the Cart,” June 2026. forbes.com
³ Adobe Analytics, various reports including “Prime Day Event Drove $24 Billion in Online Spend Across U.S. Retailers” (July 2025), Black Friday 2024, Cyber Monday 2025, holiday season 2025, and Singles’ Day data. business.adobe.com
⁴ Numerator, “Amazon Prime Day 2025 First Half Results,” July 2025
⁵ MRI-Simmons, “Back-to-School Shopping Trends of the 2026 Season,” May 2026. mrisimmons.com | Accio, “Back-to-School Trends 2026.” accio.com

The post How to prepare for peak season 2026: Prime Day to BFCM appeared first on Rithum.

Second in a series on building stronger AI-driven commerce with Rithum

At a glance  

• AI shopping agents evaluate your product data before they evaluate your brand. Incomplete or poorly structured catalogs get excluded from recommendations before a shopper sees them.  
• Most brands still rely on AI platforms scraping their product pages. Direct, structured feeds give you control over how products are represented inside AI environments.  
• Reaching an AI response and earning the recommendation are different problems. Without visibility into AI platform performance, improving placement becomes guesswork.  
• Payment infrastructure inside AI shopping environments is being built now. Brands that address the earlier stages will be positioned to capture those transactions as they scale.  

AI agents are already shaping what consumers see and buy. 70% of consumers have used an LLM to shop in the last three months, and 19% are now purchasing from brands they’d never encountered before those recommendations. But when it comes to making sure your products appear in those results, you’re operating against a black box. Are consumers seeing your products? Are they clicking? How will they purchase in an AI-driven environment? 

Demystifying that black box comes down to four stages: AI-ready product data, LLM connection, monitoring and optimization for AI engines, and in-LLM payments. Each one feeds the next. A gap at any stage means a leaky bucket for everything that follows. 

With Rithum, you can address each stage of the agentic commerce funnel today and prepare for where the space is heading next. 

Bad catalog data keeps products out of AI recommendations

AI shopping agents match product attributes against a shopper’s query. Missing specifications, inconsistent formatting, or outdated inventory signals knock products out of results entirely. 

When an AI assistant returns incorrect product information, shoppers blame your brand. Catalog quality becomes a brand trust issue, not just an operational one. Find out exactly what it’s costing you. 

The work on catalog structure, attribute coverage, and category alignment usually happens earlier in this process. Tools like Catalog Assist and Magic Mapper focus on those areas, handling attribute gaps and cross-channel categorization so product data is usable across AI-driven environments. With your catalog complete, structured, and current, you can tackle the remaining stages of the funnel with reliable inputs. 

Scraped data adds risk you cannot control  

AI platforms still rely heavily on crawling websites, marketplace listings, and third-party sources to assemble product information. Inconsistencies follow. Pricing, availability, and product descriptions can all drift away from the current state of your catalog.  

When that happens, the AI response reflects whatever information it was able to gather, not the current reality of your inventory. Very few shoppers click through to verify those details elsewhere, which turns the AI output into the primary version of the product. 

Rithum replaces that with direct, structured feeds into AI platforms. Rithum’s ChatGPT and Perplexity Feeds get your product data live and accurate on LLMs in three steps: your data is compiled into a feed, that feed is optimized for LLMs, then delivered directly for ingestion. Your brand owns its presence on LLMs instead of leaving it to crawlers. 

Rithum’s Stripe partnership extends this further by allowing brands to connect once and distribute product data across multiple AI platforms as they come online. Instead of building new integrations for each new surface, you can test across an assortment of LLMs and understand the ROI, all while keeping your product data updated and aligned. 

Getting into the system is not the same as getting selected  

AI-generated responses return a limited set of recommendations. Products compete for inclusion in that shortlist, and small differences in product data, relevance, or confidence signals can determine which products appear. 

Your feed is not a set-it-and-forget-it deliverable. You need to understand how your products are ranking across AI platforms and how those rankings shift over time.  

Our upcoming GEO (generative engine optimization) capabilities provide a way to track how products appear, move, and compare within AI-driven results. 

But monitoring only addresses half of the equation. Once you understand how you’re ranking, you need a way to improve those rankings. 

Rithum’s upcoming Performance Lab translates those signals into specific optimizations to improve how products appear in LLMs. Between GEO and Performance Lab, brands can move from “live but invisible in recommendations” to earning placement where it actually drives revenue. 

Because Rithum connects monitoring and catalog management in one place, you can act on performance signals directly. No exporting data, no cross-referencing tools, no guessing what to fix. There are plenty of myths about how agentic AI actually works. One of the most costly is assuming that presence alone drives results. 

Agentic checkout infrastructure is taking shape  

Payment is the final stage of the funnel: a shopper completes a purchase inside the same environment where they found the product. 

The current state of in-LLM checkout is uneven. Some platforms are testing in-conversation transactions, others are still building toward it, and some LLMs, including ChatGPT, have moved away from native in-chat checkout toward third-party app integrations instead. 

Rithum’s Stripe partnership provides the infrastructure for this layer. Product data flows from Rithum into AI platforms. When a transaction occurs, Stripe processes the payment while Rithum handles the necessary inventory updates and order orchestration. The brand stays the merchant of record and retains control of the post-purchase experience. 

Google’s UCP, available through Rithum’s Google Shopping feeds, opens another route, allowing brands to opt products into agentic checkout through AI-enabled search and Google Shopping, with support for loyalty programs and order management.  

Checkout only functions when the upstream stages are already working. Catalog data, platform access, and product-level performance all shape whether a shopper reaches the point of transaction. Getting those right now is the most direct path to capturing agentic commerce revenue as it grows. 

Rithum connects the full agentic commerce funnel  

Agentic commerce does not reward partial readiness. Every stage you leave unaddressed leaks value from the funnel. The teams gaining ground are the ones connecting these stages rather than running them as separate workstreams. 

Rithum and its upcoming agentic commerce capabilities connect these stages inside a single platform. Catalog improvements flow into AI feeds. Feed performance is measured. Optimization updates feed back into the catalog. When transactions occur, performance signals inform the next set of improvements. 

That full loop runs on one of the largest commerce datasets available: $50B+ in annual GMV, billions of SKU updates, and 3 out of 4 AI-driven optimizations accepted by clients. The scale and breadth to cover the full funnel from product data to AI-driven sale, in one system. 

The post How Rithum supports every stage of the agentic commerce funnel appeared first on Rithum.

Sellers across categories are adding Temu to their channel plans as they look for new ways to reach customers without rebuilding their operations from scratch. Temu ranked as the most downloaded shopping app worldwide in 2025 on the iOS App Store and Google Play, reaching more than 530 million global monthly active users by mid-year. 

Since Temu opened its marketplace to all U.S. sellers through the Local Seller Program in November 2024, sellers have moved quickly to test it alongside their existing channels. Zero registration fees, dedicated seller support, and a direct connection through Rithum give sellers a path to add the channel without overhauling how they work. 

Why add Temu: a new sales channel worth testing

Managing multiple channels gets expensive fast. Customer acquisition costs keep climbing, and adding a new marketplace typically brings new systems, new workflows, and more manual work. Temu’s Local Seller Program is built around a different model. 

Through the program, sellers of all sizes get access to Temu’s marketplace with zero registration fees, dedicated support, and built-in seller tools. Temu reports that 50% of new sellers make their first sale within 20 days of registration. Sellers can list products across more than 600 categories, including groceries, plants, books, and furniture. 

Sellers can work with their existing logistics partners or choose from Temu’s partner network for fulfillment and payment processing. Temu requires U.S. sellers to fulfill orders from domestic inventory, so local stock is recommended before activating listings.  

Temu’s discovery-based model surfaces products based on what shoppers are actively browsing and buying, with less reliance on paid placement to drive visibility. Both bestsellers and niche products get a chance to reach new audiences. According to an Ipsos survey commissioned by Temu, 1 in 4 shoppers said they are more willing to try new products on Temu, and 75% said they would recommend the platform to others. 

Rithum connection: adding Temu without adding manual work  

Rithum and Temu announced their integration in August 2025, giving Rithum sellers a direct path to add Temu without building a separate workflow for it. Sellers sync product catalogs in bulk, and inventory updates and fulfillment data move through the same workflows already in use across other channels. 

Rithum supports multi-location inventory and a range of fulfillment models, including owned warehouses, 3PLs, and marketplace fulfillment programs, so sellers can manage Temu alongside existing operations from one platform. Rithum connects sellers to more than 420 marketplace and retail integrations worldwide, including Amazon, Walmart, Target Plus, and TikTok Shop — Temu joins that network as another channel sellers can activate and manage from a central hub, without adding headcount or rebuilding operations. 

Pennsylvania-based Book & Mortar Record Store joined Temu in August 2025 through Rithum’s integration. Founder Eric Auth runs a small team selling hundreds of thousands of books, vinyl records, and other collectible media products online. Within months, the business listed several hundred thousand SKUs on Temu through the integration and moved more than 31,000 units without paid advertising yet. Auth credited Temu’s discovery-based shopping experience with helping the business reach new audiences and drive sales volume. 

For sellers already on Rithum, the Temu setup guide in the Knowledge Center walks through configuration, catalog sync, and listing activation. 

How one seller put it to work on Temu through Rithum 

Pennsylvania-based Book & Mortar Record Store sells books, vinyl records, and collectibles online, with hundreds of thousands of products in its catalog. Founder Eric Auth runs his business with a small team, so every new marketplace has to be manageable. Rithum already supported day-to-day operations across channels, and Temu came into that same workflow when the integration launched in August 2025. Within months, Book & Mortar listed several hundred thousand SKUs on the platform, moved more than 31,000 units without paid advertising, and reported daily GMV in the $5,000 to $10,000 range.  

Temu’s discovery-based model was a practical fit for a catalog that spans bestsellers and niche titles alike. The platform surfaces products based on shopper interest, giving individual titles a better chance of finding the right audience and turning catalog breadth into steady sell-through. 

“Rithum’s integration with Temu made selling on Temu possible for us. We can load a large catalog in bulk and cut down on errors and hours of manual work. Adding Temu as a new channel has enabled us to scale to several hundred thousand SKUs, and it’s now making up over 15% of our sales within just six months.”  

— Eric Auth, Business Owner, Book and Mortar Record Store 

Getting started on Temu 

For sellers looking to evaluate and scale on Temu, here are a few steps to get started: 

• Review Temu’s seller policies: Check marketplace rules before listing products, including prohibited categories and compliance requirements. Consistently meeting product quality standards helps build customer trust and earn positive reviews. 

• Meet shipping requirements: Follow expected delivery times and packaging standards to ensure smooth fulfillment and customer satisfaction. Temu requires U.S. sellers to fulfill from domestic inventory, so confirm your logistics setup before activating listings. 

• Provide excellent customer service: Respond promptly to customer inquiries and handle returns and refunds efficiently to support repeat business and maintain good standing. 

• Track performance metrics: Monitor order fulfillment rates, shipping performance, and customer satisfaction scores to identify areas where to improve. 

• Tap into Temu’s tools and support: Stay up to date with requirements and policies through Temu’s seller center resources and use tools such as the IP Protection Portal and Brand Registry to help manage intellectual property. 

• Connect through Rithum. If you’re already a Rithum seller, add Temu directly through the Rithum Connection. The Temu setup guide in the Knowledge Center covers configuration, catalog sync, and listing activation step by step. 

The post Why sellers are adding Temu to their channel mix appeared first on Rithum.

At a glance: 

• 53% of shoppers trust AI tools, including an AI shopping assistant, as much as brand websites, according to a Rithum and Retail Dive survey. This trust is reshaping AI shopping verification behavior and expectations for AI in retail. 
• When shoppers verify an AI recommendation, retailer and brand sites rank near the bottom at 5%, showing how AI shopping verification is happening away from owned channels. 
• 64% of shoppers ages 18 to 27 say they’re likely to purchase based on an AI recommendation without verifying it anywhere else. 
• AI-referred visitors now convert 42% higher than non-AI traffic.
• Retailers like Walmart, Target, and Instacart are enabling purchases directly inside AI conversations, accelerating a future where the entire shopping journey happens in one place. 

A year ago, shoppers arriving through AI tools browsed but left without buying. A year later, those same shoppers are 42% more likely to buy than shoppers arriving through traditional channels¹. In the same month, Walmart deployed its AI shopping agent inside ChatGPT², joining Target, Instacart, and DoorDash in letting shoppers browse, compare, and buy products directly inside the conversation. 

In Rithum and Retail Dive’s survey of 1,046 online shoppers across the U.S. and U.K., 53% already trust AI tools as much as brand websites. And when they want a second opinion on what AI told them, they’re going everywhere except the brand to get it. 

When shoppers double-check, they go everywhere but the brand site 

When shoppers verify an AI recommendation, they’re choosing channels outside the brand’s control. Search engines are the top destination at 28%. Online reviews come next at 19%, followed by friends and family at 17%. Retailer and brand sites rank near the bottom at 5%. 

A brand’s own product page, no matter how thorough, is still the brand talking about itself. Shoppers want independent voices, and they’re finding them everywhere else. 

That puts more pressure on the product data traveling through those channels. If the search engine is the second stop after AI, the data you’re pushing into Google, Bing, and other platforms needs to be accurate and complete. If a shopper pulls up a review site and finds specs that conflict with what the AI told them, the brand absorbs that cost. In the same survey, 58% of shoppers said trust in the brand decreases when an AI recommendation contains incorrect product information, and 16% abandon the purchase entirely. 

Brands have the answers, but shoppers are asking somewhere else 

A shopper asks an AI tool to recommend a running shoe for flat feet under $150. Three options come back. The shopper likes one but wants to confirm the arch support claim before buying. 

They type the product name into Google. They scan a couple of review sites. They text a friend who runs. The brand’s product page may have the most detailed answer to their question, but the shopper has already moved on to other sources. 

Product information accuracy across your entire distribution footprint now carries more weight than the quality of your own site experience. Feeds, marketplace listings, third-party retailer pages, and structured data that AI tools can parse all shape what the shopper encounters during verification. The brands investing in that full footprint are the ones staying in the consideration set. The ones focused primarily on their own site are building for a shopping journey that fewer customers follow. 

A growing share of shoppers skip verification entirely 

Among shoppers ages 18 to 27, 64% say they’re likely to purchase based on an AI recommendation without verifying it anywhere else. Higher-income shoppers are twice as likely to trust AI without visiting another site. And across all demographics, 32% say they spend less time browsing other sites after using an LLM. 

For these buyers, the AI recommendation from an AI shopping assistant is the decision. The brand site is largely absent from it. And shoppers who verify through search and reviews encounter a broader set of options than they would on a single brand’s site, giving unfamiliar brands a real opening to enter the consideration set with accurate, well-distributed product data. Shoppers who skip verification altogether are relying entirely on whatever the AI already knows about your product. 

The verification step itself is disappearing 

64% of shoppers already take AI at its word. The platforms coming next are built to make that feel even more natural.

AI agents could mediate $3 trillion to $5 trillion of global consumer commerce by 2030, according to McKinsey and Co. Two competing open protocols are already live and processing transactions end to end: OpenAI and Stripe’s Agentic Commerce Protocol and Google’s Universal Commerce Protocol. The AI agent handles product discovery, comparison, checkout in one place. 

Now picture that same running shoe shopper six months from now. They ask ChatGPT the same question. A product card appears with an image, price, and a “Buy” button. They tap it, confirm their saved payment method, and the order ships. The entire transaction happened inside a single conversation. 

The shoppers who still verify aren’t going back to the brand to do it. They’re checking reviews, search results, other people. And as AI agents take on more of that process, the brand’s window to influence the answer gets smaller. The data has to be right before the question is ever asked.”

Your product data is now your pitch to an AI buyer that will never visit your homepage 

Getting product data right across every channel is the minimum. It’s expected. The question is where that data lives: search engines, review platforms, marketplace listings, and the structured data feeds that AI agents read when they decide what to recommend. AI-readable product content needs to be complete, consistent, and built for machines to parse. For a growing number of shoppers, that content is the only version of your brand they’ll ever see. 

The distance between discovery and purchase is collapsing. Sometimes it’s a single conversation with an AI agent. The brands feeding that conversation with accurate, well-distributed product data are the ones the agent recommends. 

For a full breakdown of the data, download The New Discovery Engine report. 

Sources:
1: https://www.retailtouchpoints.com/features/the-agentic-commerce-paradox-its-already-here-and-its-also-still-evolving/618945/  
2: https://www.cbsnews.com/news/ai-agentic-artificial-inteligence-what-is-it/

The post Shoppers are verifying elsewhere (away from brands and retailer sites)  appeared first on Rithum.

At a glance:

• More than a third of surveyed commerce leaders say key tasks are partially automated but still require manual checks. Many workflows still depend on a human proof step before go-live, according to the 2026 commerce readiness index.  
• Another 29% said data is scattered and stockouts show up after orders, turning routine pricing and inventory updates into avoidable cleanup work.  
• 34% of commerce leaders said dashboards are available, but teams are still forced into manual exports to confirm what they’re seeing.  

Retailers and brands have spent years layering ecommerce automation into pricing, inventory, listings, and media. But many still stop for a manual proof step before go-live.  

The double-check has not gone away. It often comes just before a team is ready to act. A price looks right in one system but off in another. Inventory looks stable until the channel view suggests otherwise. The team stops to verify the basics before moving ahead. That pause turns automated work back into manual work. 

Ecommerce automation still needs a proof step  

In the 2026 commerce readiness index, 34% of commerce leaders said key tasks are partially automated but still require manual checks. Nearly half of executives said 26%-50% of workflows still depend on spreadsheets, and repetitive tasks like manual data entry, and approvals.  

Promotions punish messy handoffs  

Marketing campaign promotions don’t leave much room for hesitation. Price, inventory, product content, and media all move at once, yet teams are often working against slower decision cycles: more than half of retailers said they act on meaningful performance signals within 48 hours, while brands are more likely to take three to five business days.  

The readiness assessment helps explain why: 32% said signals are spread across tools and reports, while 29% said they have dashboards and alerts but unclear workflows. When product feeds, pricing, availability, and catalog updates refresh on different schedules, a promotion can look ready until something slips and the team has to stop to confirm what is actually true. Rithum’s retail media guide describes the same problem from the campaign side: by the time the dashboard reflects it, time and budget may already be gone. 

Scattered data turns ordinary work into manual work  

The trouble is often small at first. A price changes here but not there. Inventory moves, but not everywhere at once. Reporting can show that something changed without showing where it began. In the index, 29% of respondents said stockouts appear only after orders come in. Routine work turns into cleanup.  

Dashboards lose their authority when something shifts  

A dashboard can feel reliable until something starts moving faster than your reporting can explain. When a product moves faster than expected, a promotion performs differently across channels, or a price update appears in one system but not another, the dashboard can flag the issue without showing its cause. The team has to look elsewhere to confirm what changed.  

The readiness assessment points to the same problem: 34% said core dashboards are standardized, but edge cases and new channels still depend on manual exports they know are unreliable. Another 26% said the data they work from is incomplete, late, or manually tweaked, but they still use it because it is all they have. The gap is not only in dashboard coverage, but in confidence in the data underneath it. 

Rithum’s retail media guide points to what’s missing: product context alongside campaign performance. When teams can see which products absorbed the budget, what changed in price or availability, and which issues need attention first, reporting stays useful while teams are still deciding what to do.  

The window to act is getting tighter  

The gap becomes more expensive during peak shopping events. In the report, more than half of retailers said they act on meaningful performance signals within 48 hours, while brands are more likely to take three to five business days.  

Rithum’s Prime Days 2025 data shows how timing can be problematic. One brand held spend until conversion and AOV recovered, then pushed harder. Another brand pivoted mid-event toward back-to-school assortments, bundles, and sharper titles and keywords, finishing 15% above the prior year’s Prime Day. The advantage was a timely read on what had changed, while there was still room to respond.  

What teams should standardize now to achieve the benefits of ecommerce automation  

Instead of adding more ecommerce automation, retailers and brands should look at where to cut back on verification work. Start with the handoffs that break trust most often. Set clear source-of-truth rules for product data, pricing, and inventory so a routine channel change does not trigger a manual review. Exceptions should surface early, with enough context for teams to understand what changed without going back into spreadsheets.  

Where Rithum helps by providing automation tools for ecommerce  

Rithum helps cut down the verification work that piles up between systems by offering automation tools for ecommerce. Error tracking flags mismatches earlier. Automated tools and workflows reduce the same reconciliation loop playing out over and over. Connected commerce and media insights bring pricing, listings, inventory, fulfillment, and performance into a view teams can actually use.  

For retail media teams, product changes and campaign performance sit closer together, so it’s easier to spot what shifted, what needs attention, and where to act first. As a result there are fewer exports, less back-and-forth, and less time spent confirming what should already be clear.  

Download the full 2026 commerce readiness index to see where retailers and brands are still losing time, and what it takes to move faster without adding risk. Then take our readiness assessment to see where you stand in comparison.  

The post The hidden cost of ecommerce automation is verification work appeared first on Rithum.

TL;DR 

• Product journeys, from discovery to decision, are shifting to AI. Among AI-active shoppers, 90%+ use LLMs to research products and compare options, and 53% use them to decide where to buy. 
• AI is diminishing buyer loyalty. 19% say they now buy from brands or products they had not heard about before, and 13% say they are more likely to switch retailers or products after using an LLM. 
• Brand and retailer sites have less time to influence the decision. 32% of shoppers spend less time browsing other sites after using LLM tools, and only 5% verify AI shopping information on a retailer or brand site. 
• Product information now plays a bigger role in trust. 49% say a clear explanation would do the most to increase trust in an AI recommendation, 67% say price is the most important detail to get right, and 58% say trust in the brand drops when an LLM provides incorrect product information. 

Rithum’s new report, The new discovery engine: How consumers are using AI to find, trust, and choose brands, and what’s at risk for those they never see, has a clear message for retailers and brands: the shopping journey is no longer confined to shelves, search results, category pages, or product detail pages. 

Based on a survey of 1,046 online shoppers in the U.S. and U.K., the report shows how LLMs have become the entire shopping journey and a one-stop shop where products get researched, compared, narrowed down, and chosen. Among AI-active shoppers, more than 90% use LLMs to research product information and compare options, while 53% use them to decide where to buy. By the time a shopper lands on a product page, the filtering may already be over. 

That shift to AI is creating room for brands that were not previously in the mix. In the survey, 19% of shoppers say they now buy from brands or products they had not heard about before. Established brands are facing a tougher version of the same market. Recognition still helps, but it has less force when AI is doing more of the sorting, ranking, and explaining before the click. 

The shortlist is forming earlier  

AI shopping adoption is the starting point for 80% of shoppers ages 18-to-27 and 80% among shoppers ages 28-to-43. Among households earning $100,000 to $150,000, it reached 84% adoption. These are commercially important shoppers, and they are already weaving LLMs into their purchase journey. 

The filtering that used to happen across tabs, retailer sites, and review pages is now happening inside LLMs. A product that appears high in the response moves ahead. One that doesn’t can disappear before the shopper has seriously considered it. 

ChatGPT’s product comparison feature adds to that shift. Shoppers can compare products side by side inside the chat, with price, features, reviews, and other details presented in one place instead of scattered across multiple retailer tabs. 

More than half of shoppers already trust AI tools as much as brand websites, and among high-income households, confidence in AI accuracy climbs as high as 80%. That trust gives LLM recommendations real weight early in the decision process, according to the report. 

Retailers now have more riding on how products are represented on their page, even before a shopper lands on the site. Brands face the same pressure, but with fewer natural intervention points. A retailer may still appear as the place to buy, even if a brand is filtered out earlier. If a brand’s product information is incomplete, inconsistent, or hard for AI to explain, it can be dropped from consideration before its own product page or brand story has a chance to influence the decision. 

New brands are finding room where familiar brands once had an easier ride 

You can see the LLM effect far beyond just initial research, with ripples into what shoppers buy. Nineteen percent of shoppers say they are more likely to buy from brands or products they had not heard about before if an LLM suggests it. Another 13% say they are more likely to switch retailers or products after using an LLM. Together, those numbers create a shopping environment where familiar brands have less room to rely on recognition alone. 

That creates an opening for challenger brands. A newer brand does not need years of broad recognition to get in front of a shopper. It needs usable, consistentproduct information and enough context for AI to present it clearly and convincingly. 

Established brands have less room to lean on familiarity alone. Customer loyalty still helps, but it no longer ensures that they go to your site first. Nearly half of shoppers say a clear explanation of why a product or brand was chosen would do the most to increase trust in an LLM recommendation. What carries weight here in an LLM recommendation is not name recognition but whether the recommendation feels specific, informed, and ready to act on. 

Brand-owned sites get fewer chances to influence the outcome 

The shopping journey used to leave more room for second thoughts. A shopper could open a few tabs, compare prices, read reviews, leave, come back, then change course. LLMs have shortened that process.  

In the survey, 32% of consumers say they spend less time browsing other sites when using LLM tools to shop. Another 36% say they make faster decisions, while 34% say they feel more confident about their purchases.  

These three stats don’t live in a vacuum. They indicate a continual trust-building experience for the shopper: they’re saving time, they’re finding what they need faster, and they feel better about their purchases. Why would they leave that experience to go back to a retailers website? 

The same pattern appears in how people verify what they see. Shoppers who double-check an LLM recommendation rarely begin with looking for confirmation on a brand or retailer site. Twenty-eight percent turn to search engines (which is likely also relying on AI tools), 19% specifically look for online reviews, 17% ask friends and family, and only 5% go to a retailer or brand website. A beautiful, brand-forward website won’t convince them to buy your product. They won’t even see it. But a PDP with in-depth specifications, GEO-optimized keywords, and highly relevant descriptions will impact consumers’ decisions, even if they don’t see the page. 

The recommendation is only as strong as the product story behind it 

Ask an LLM why it chooses one product over another, and it has to build that answer from the product facts it can find: materials, dimensions, compatibility, intended use, etc. The recommendation that an LLM givesis assembled from those pieces in real time. 

For brands, that raises the standard for product content. Copy, attributes, use cases, and supporting details are no longer sitting off to the side as content maintenance. They are becoming part of the recommendation itself. When the product story is thin, generic, or inconsistent, the answer reads that way too. 

Retailers feel the same pressure across the assortment. Pricing, inventory, attribute completeness, and feed quality all shape how products are represented before a shopper ever reaches the site. Anyone who has spent time inside a catalog has seen how quickly that can start to fray. A bad price, a missing dimension, or stale availability can make a solid product look less reliable than it is. 

The harder question is whether the product story still holds together everywhere that LLMs are pulling from. This includes product content, syndication, pricing, availability, and the systems that keep those details aligned. It also includes sources brands and retailers cannot fully control, such as reviews, forums, and social discussion. When those external signals surface alongside structured product data, inconsistencies become more visible. That makes it even more important for the information you do control to be accurate, complete, and easy for AI to explain. LLMs only give recommendations they can trust, based on the information that holds it together. 

Trust is moving closer to the data itself 

The survey leaves little ambiguity on price. In an AI shopping recommendation, 67% say it is the most important detail to get right. Reviews, availability, where to buy, and technical specifications all come after it in the list of prioritization 

That order will feel familiar to anyone who has watched shoppers abandon a cart over a mismatch or lose confidence over a number that does not look right. A wrong price or stale detail does not stay in the background. It becomes part of the recommendation, which means it becomes part of the shopper’s impression of your brand. 

The report puts numbers behind that. When an LLM provides incorrect product information, 58% say trust in the product or brand decreases, and 16% say they leave the purchase altogether. 

At that point, the issue is no longer confined to data quality. The recommendation may come from the model, but shoppers are not spending time sorting out where the error began. They decide whether the information feels reliable, and the brand lives with the result. 

The next phase is close enough to shape decisions now 

The report also looks ahead to a shopping flow where the model takes on more of the decision itself. More than 25% of AI-active shoppers say they are already very likely to hand purchasing decisions to AI, and another 39% say they are somewhat likely to consider doing this, if and when it’s available. Among the most AI-active shoppers, 65% say they are very or somewhat likely to use an AI agent that would buy for them. 

What AI sees already shapes what shoppers buy. Thin product content, stale pricing, patchy attributes, and inconsistent availability all weaken the recommendation before the shopper has done anything beyond type in a prompt. 

The priorities are clear. Keep your product story consistent. Keep pricing accurate. Keep availability current. Make products easier to compare, easier to explain, and less likely to be misread. New brands already have more room to enter the conversation. Established brands have less room for weak information, stale details, or missing context. 

For more details on the survey and a full breakdown of the results, download the report here.  

Methodology 

Rithum’s 2026 report is based on a survey of 1,046 online shoppers in the U.S. and U.K. Some questions look at behavior in the last 3 months, some category questions use the last 6 months, and some trust and behavior questions are broader and are not tied to a single recall window. 

The post A brand your customer had never heard of just won the sale appeared first on Rithum.

• Snapshot: eTail Insights surveyed marketplace and ecommerce leaders for Mastering the marketplace in retail and ecommerce to benchmark marketplace trends including what’s working, what’s breaking, and where budgets are going next. 
  Why it matters: These benchmarks show how marketplace teams describe their day-to-day reality right now. 

• Listing optimization no longer separates anyone. In the survey, every team rates its listing optimization capability as good or better, which tells you where the “easy wins” have already been harvested. 
  Why it matters: Listings may look “done,” but they stop being the lever that moves growth. 

• Retailer and brand teams lose time and margin on coordination. Inventory coordination across platforms ranks as the top challenge at 40%, followed by marketplace data integration at 33%. 
  Why it matters: When systems disagree, the fallout lands in fulfillment, customer service, and returns. 

• Leaders are budgeting for trust optimization above all else. Over the next 12 months, the top investment is customer service and review management systems at 51%, ahead of marketplace advertising and content expansion at 46%. This makes sense, as credibility optimization is key to meeting both customers’ and agentic shopping tools’ high standards. 
  Why it matters: Teams are funding the places where customers notice problems first and where reputations change fastest. 

• Profit pressure is shaping the stack. Competitive and dynamic pricing tools sit close behind at 41%, a signal that fees and price competition are forcing sharper math. 
  Why it matters: If teams can’t see the true cost per order, pricing tools can push them into underpricing. 

• Marketplaces are moving from “channel” to “center of gravity.” 81% expect marketplace sales to become more important to overall strategy, and 73% expect that shift within the next year. 
  Why it matters: Marketplace execution is becoming core operations, not an add-on project. 

A customer adds a refrigerator water filter to their cart. They’ve bought the brand before, and everything looks familiar. One click from paying, they pause. The model name is close, but not identical.  The customer taps the chat bubble in the corner of the screen to make sure they’re purchasing the right part.  

That bubble used to connect them to a live agent. Now it’s often an AI assistant, trained to answer product questions quickly, so the shopper doesn’t abandon checkout. 

“Is this the filter for model X?” 

The assistant replies instantly: “yes, it’s compatible.”  

But it’s not.  

The filter fits a similar model with a near-identical name. The product listing doesn’t clearly distinguish the two, so the AI fills in the gap with the most plausible-sounding answer. The customer buys anyway. The filter arrives, doesn’t fit, and gets returned. 

The return reason isn’t “wrong model.” It’s “defective.” The frustration isn’t just with the product. It’s with the guidance the shopper trusted. 

This is the kind of mistake customers try to avoid. When it happens anyway, they don’t blame the model, they blame the seller. Marketplaces penalize getting it “almost right” with returns, bad reviews, and support costs. The warning appears in performance long before you see it on a dashboard. 

The new marketplace problem: the storefront that moves faster than operations 

Most organizations aren’t dabbling in marketplaces anymore. Nearly everyone sells on Amazon. Most also sell on Walmart.com, according to eTail’s Mastering the marketplace in retail and ecommerce report. 

But while most respondents say their marketplaces strategy is at least somewhat effective for revenue, there is a deeper problem. Each marketplace comes with its own rules and requirements. Retailers and brands don’t describe their biggest marketplace challenges as “getting listed” (the industry has evolved past that point). They describe them as keeping platforms aligned. 

Inventory coordination across platforms ranks as the top challenge at 40%. Marketplace data integration follows at 33%. 

Those aren’t minor headaches. They sit underneath the moments customers remember: the delivery promise that never was, the unwelcomed out-of-stock surprise, the placement part that “should have worked,” or the one-star review that deters others from buying your product. 

Meanwhile, marketplace importance keeps rising. 81% say marketplaces will become more important to overall strategy, and 73% expect that shift to happen within twelve months. So, as opportunity grows, the friction grows along with it. 

Where marketplace execution starts to separate teams

Listing optimization is not the weak point: every respondent rates their listings as good or better. But when everyone feels strong in the same area, it is no longer a competitive advantage. Instead, the advantage shifts to what’s more difficult to keep consistent at scale, such as inventory accuracy, data consistency, review health, margin control, and service quality. 

It also moves to the places where organizations admit weakness. In this case, one-third rate “review and reputation management” as fair or poor. 

That gap matters because reviews aren’t just social proof. They shape ranking, conversion, and what shoppers learn before they buy. They also feed the summaries shoppers increasingly rely on, whether that’s a marketplace recap, a review highlight, or an AI-style shopping assistant answer. A shopper who doubts your listing doesn’t read your FAQ. They read your reviews. 

Marketplaces reward speed, but trust decides who keeps the sale 

Marketplaces raise expectations without lowering complexity. Shoppers expect quick answers, fast shipping, and painless resolution even when the catalog and fulfillment reality are messy.  

Trust decides whether a retailer or brand keeps the sale. On marketplaces, customers judge trust through what happens when something goes wrong and what other buyers say happened to them. That’s why the report’s investment priorities matter. Leaders are investing in customer service and review management systems at 51% over the next year, ahead of advertising and content expansion at 46%. 

Competitive and dynamic pricing tools follow close behind at 41%, underscoring how quickly the fees and shifting costs can erode margins. Teams don’t lead with service systems unless service continues to break. 

The catalog is the experience now 

The story of a mismatched water filter isn’t really about AI. It’s about what happens when product information fails under pressure and gets repeated across channels. 

Marketplaces have always punished unclear product information, but the old failure mode looked different. A customer read the listing, guessed, and learned the truth when the package arrived. 

Now the failure happens earlier. Customers ask questions mid-journey and even when the catalog is incomplete or inconsistent, the assistant still answers. When two SKUs look nearly identical, the assistant still picks one. When variation structure is messy, the assistant still tries to resolve it. If your product data leaves room for interpretation, the assistant will fill it, and you will own the consequences. 

What breaks first when teams scale marketplaces 

Most marketplace failures don’t start with a dramatic crash. Then one day, someone notices reviews shifting in tone. Support tickets spiking. Returns rising. A marketplace rep flags an issue. A finance team sees margin slip and can’t tie it to one decision. 

Coordination can be the biggest bottleneck. Inventory coordination and marketplace data integration top the list. If you can’t keep systems aligned, you can’t keep promises aligned. 

Three shifts that make marketplace growth easier to scale 

Most teams don’t need a radical reinvention. They need to tighten the parts that create expensive downstream effects. 

These three easy shifts  map directly to what leaders say is slowing them down. 

1) Treat inventory coordination like a revenue problem 

Inventory coordination across platforms ranks as the top marketplace challenge. This sits at the center of marketplace performance. 

When inventory is wrong, you don’t just lose a sale. You lose ranking and trust. You waste customer support time. You pay for returns. You create reviews that never should have existed. 

Start by making your inventory view boring and reliable. Establish one source of truth for what’s sellable. Push that truth to every platform on a cadence you can defend. 

That shift prevents the “invisible losses” teams learn to tolerate, including cancellations, refunds, angry reviews, and time spent explaining mistakes. 

2) Treat review management as a system, not a reaction 

One-third of survey respondents rate “review and reputation management” as fair or poor. That’s one-third of teams saying they still treat reviews reactively—and know that it’s hurting them. 

Reviews deserve a better approach because they aren’t random. They’re patterns. 

A shopper complains that the filter “didn’t work.” Another says the product “felt defective.” A third says the listing “misled them.” The issue is expectation. 

Route review themes back to the source. If customers keep confusing models, fix the variation structure. If customers misunderstand compatibility, tighten attributes. If customers misread product claims, update bullet copy. 

When review management becomes a loop instead of a fire drill, you stop treating bad reviews as a cost of doing business. 

3) Align marketplace data before you scale ads 

Marketplace data integration ranks as the second-biggest challenge at 33%, and it’s easy to underestimate how costly those integration speed bumps become once you push harder on growth levers. 

Incomplete integration creates quiet inconsistencies: price mismatches, delayed updates, missing attributes, inventory errors that propagate across systems. 

Then advertising scales and performance doesn’t follow. You spend more to acquire customers you can’t keep, and you blame the campaign when the problem lived upstream. 

Tighten the data path before you scale spend. Your future self will thank you in fewer support tickets and cleaner margin. 

Where AI fits without hijacking your strategy 

AI can help marketplace teams move faster. It can spot inconsistencies, flag missing fields, draft content variants, and speed up service workflows. 

It cannot rescue messy foundations. Give it incomplete data and it produces confident answers built on gaps. That’s why product truth has become a performance lever: AI increases the speed of both accuracy and error. 

What mastering the marketplace looks like in 2026 

Most teams feel good about listings. The report shows the friction elsewhere: inventory coordination, marketplace data integration, and review management. That’s why next year’s spend tilts toward customer service and review systems ahead of content and advertising. 

Marketplace growth requires consistency. Keep one version of the product true across platforms: what it is, whether it’s available, what it costs after fees, and what the customer should expect. The report shows why this is hard at scale. Inventory coordination ranks as the top challenge at 40%, followed by marketplace data integration at 33%. When those basics slip, the cost surfaces as returns, heavier support volume, and reviews that live on the listing. 

Micah McGuire is Senior Product Marketer, Brands, at Rithum. 

The post Mastering the marketplace: Build credibility with accurate inventory, data, and reviews appeared first on Rithum.

During peak season, more than $1.5B in Cyber 5 sales moved through Rithum across channels, regions, and categories. That’s the first wave. The second arrives after the shipping notifications stop: returns season. Calendars reset, staffing normalizes, and inventory shifts back toward warehouses. For the shopper, the decision often happened earlier—at checkout, when they relied on whatever the product page made clear and whatever it didn’t. 

And as shopping shifts from product pages to AI agents that summarize and recommend, returns get harder to manage. When an agent gives the highlights, shoppers may never scan the full product detail page (PDP), compare specs side by side, or catch details that prevent a mismatch. If the summary is wrong, the purchase can still go through, and the retailer still absorbs the return. The harder part is accountability and diagnosis. If shoppers never see the original product page, you lose the trail of what they were told and you can’t easily trace a return back to a missing detail, a bad summary, or a mismatched listing. 

How can you prepare for returns season? Use January to audit the promises you made during peak: which products were misunderstood, which channels dropped key details, and which fixes would have prevented repeat returns. 

What returns show 

In the U.S., online returns will top $363B, driven by a 24.5% return rate, according to Rithum’s 2025 global returns & profit impact report.  

Rithum’s consumer research shows the same drivers repeating: 61% cite poor fit and over a third say the item didn’t match the description or photos. 

Fit plays out differently by category. The “didn’t match” driver applies across categories. 

Electronics returns can trace back to compatibility or what was included. Home returns follow scale and finish. Parts and industrial returns surface around fitment, specifications, or installation assumptions. Across categories, the same thing happens: unclear listings leave customers to guess. 

In January, returns cluster. The same products come back through the same channels for the same missing details. In a Rithum returns fireside chat featuring Fabian Ortmann, Head of Returns, ZEOS; Kevin Brown, Director, Sales and Strategic Partnerships, Essendant Fulfillment Services; and Louis Camassa, Director of Product, Rithum, Ortmann argued that returns should be treated as a data source, not just a cost center, one of the clearest ways to pinpoint where expectations broke by product, channel, and market. 

That framing changes how brands and retailers should use January. For returns season, the work should shift from cleanup to finding out what went wrong. 

Where channel listings diverge 

Channel differences start costing money in January. During peak, products are listed, reformatted, and rewritten to fit each marketplace’s format. That’s expected. The risk is what gets lost. 

One channel may carry a compatibility field. Another may not. One listing spells out what’s included, while another assumes it’s obvious. One channel enforces short titles that drop (important) qualifiers where another keeps them in tact.  

Louis described this in the fireside chat: when listings leave room for interpretation, customers answer the question themselves. Sometimes they guess right. Sometimes they don’t and that can lead to a higher rate of returns. 

Why timing changes the math 

Returns also don’t happen on your schedule. In the same conversation, Essendant’s Kevin Brown described the gap between decision and action. Customers often decide quickly whether they’re keeping something. The return itself may take weeks to arrive. 

That delay narrows options. Inventory comes back later. Resale windows shrink. Seasonal goods lose flexibility. The cost isn’t only the refund. It’s what the business can no longer do with the product. 

January is also when slow fixes compound. If product information is inconsistent across channels and updates move slowly, the same mismatch keeps shipping while teams debate which version of the truth is “correct.” 

Rithum’s 2026 commerce readiness index points to why this is common. 91% of retailers and 78% of brands report poor data quality. Nearly 75% say inaccurate data leads to bad decisions. 

When product records aren’t consistent or trustworthy, January becomes more triage than correction. 

What policy can and can’t do 

Return policy shapes demand, but it doesn’t explain repeat returns. Policy still matters because shoppers read it before they buy, especially when they’re unsure. 

Rithum’s research shows return policies influence 41% of purchase decisions, 88% expect free returns, and 47% won’t click “buy” without them. 

Most preventable returns aren’t caused by policy. They’re caused by uncertainty that existed before policy mattered. 

What to check first 

Start with the returns that teach you something. Look at SKUs that return quickly after delivery. Fast returns tend to signal expectation breaks, not end-of-season cleanouts. 

Compare the same SKU across your top channels. Don’t assume it matches your internal record. Check titles, key attributes, images, what’s-included language, and variants. 

Track “didn’t match description/photos” returns by channel. Rithum found that 33% of shoppers cite mismatch with the description or photos as a reason for returns. 

Then move from insight to change quickly. If updates take weeks to reach every listing, the same mismatch keeps shipping. 

Where AI fits 

AI won’t fix messy product data. The eTail report’s preparedness finding reinforces the same issue returns already expose: only 2% of organizations describe themselves as fully prepared with structured product feeds and governance. 

Rithum’s 2026 commerce readiness index shows how widespread the gap still is: 91% of retailers and 78% of brands report poor data quality, and nearly 75% say inaccurate data leads to bad decisions.  

Whether the buyer is human, a marketplace algorithm, or an emerging shopping interface, the requirement doesn’t change. Product truth has to be consistent, category-appropriate, and fast to update. AI will move decisions faster, but it will also scale whatever your data gets wrong. 

What to do next 

The goal is to reduce the avoidable ones and recover value faster on the rest. 

The brands and retailers who do this well don’t wait for the next peak. They use January to tighten product truth, align it across channels, and push corrections while the signal is still fresh. Learn how Rithum can help.

Jordan Christensen is Director, Client Experience, Retailers at Rithum. 

The post Returns season is here. And AI raises the stakes  appeared first on Rithum.

Retail teams adopted AI fast. But now, it’s about applying strategy. To learn how the industry was moving past the hype and into execution, eTail Insights surveyed senior retail and ecommerce leaders for The 2026 retail AI revolution in commerce. The report includes a snapshot of what retail teams are using now, what results actually support those choices, and what still looks promising but hasn’t proven reliable just yet. The result is a baseline that retailers and brands can use for planning, grounded in how peers describe strategy, effectiveness, and readiness.  

What might be surprising—or relatable—is that only 4% of organizations say they have a comprehensive AI strategy in place. Most teams still build the playbook as they go. 54% say they’re developing an AI strategy while 39% say they’re experimenting.  

Adoption is widespread, confidence is not. Only 7% rate their AI implementations as very effective. Most respondents land in the middle: they call AI “somewhat effective.” 36% say their AI has been minimally effective, and 4% say it has not been effective.  

In practice, AI projects can move faster than the plan that’s meant to guide it. What does this mean for AI effectiveness? 

Where AI already runs, and why results vary 

AI already runs through everyday ecommerce, especially in personalized product recommendations, inventory management and demand forecasting, and customer data analytics and insights. Most organizations rate their capabilities in these areas as at least somewhat advanced. 

That’s not the surprising part. Retail doesn’t debate whether to start with AI anymore. Many teams already operate with it in the mix but want to turn widespread use into repeatable results. The report starts to answer that by looking at where AI is already making a clear difference. 

Where leaders see results today 

Instead of predictions, the survey asked for outcomes that retailer and brand leaders can point to right now. The biggest reported impacts are in inventory management where 60% reported improvement, with 58% noting demand forecasting. Other areas of impact include customer value, customer service costs, and conversion. 

Only a small share rate implementations as very effective. Mariko Davison, a Senior Technical Account Manager at Rithum, puts a practical reason behind that gap: AI acts like a magnifier, so messy inputs create messy outcomes. One example she sees often is incorrect titles, descriptions, or attributes pushing an item into the wrong product type, which then triggers missing required fields, listing and variation errors, and sometimes even post-sale issues. 

The three challenges leaders cite most 

When leaders explain why AI progress slows after early wins, three issues come up most often: 

• 50% data privacy and compliance 
• 49% high costs 
• 47% customer trust 

These constraints help explain why strategy maturity remains low, with only 4% reporting a comprehensive AI strategy. 

Personalization stays a top customer experience priority 

Personalization remains a top customer experience priority, with 55% of the survey respondents ranking it as high-or-top tier. Personalized product recommendations also rank among the most advanced AI uses reported. Many teams already run them, but they still want stronger results. 

Agentic AI exposes a readiness gap 

Agentic AI refers to systems that can surface products and take shopping actions on a shopper’s behalf. The survey asked leaders how prepared they are to make products discoverable in these environments. The answer: Readiness for agentic discovery is low. 

Only 2% say they are fully prepared. 56% say they are partially prepared, and 42% say they are still in early exploration. 

Retail and brand leaders expect discovery and buying behavior to change, but most say they are not prepared yet. 

Where budgets go next, based on planned investment 

According to the survey, budget priorities are still fundamentals-first. 53% name inventory and demand management as a top investment area next year, with supply chain at 45% and customer analytics close behind. This focus matches the biggest reported gains in inventory and forecasting. 

What respondents expect next 

In open responses, respondents describe AI as speeding up operations, improving personalization, reshaping loyalty programs, and making the customer journey faster and more data-driven. The emphasis stays practical: efficiency and personalization first, then faster decisions. 

While AI adoption is high, readiness lags. Use the benchmarks to test where you are strong and where you are exposed before you scale. Download The retail AI revolution to pressure-test what comes next.  

The post Retail’s AI reality check: Adoption is high, readiness is not  appeared first on Rithum.

During Prime Days 2025, one client brand Rithum works with saw their conversion and average order value slide early. Instead of throwing more budget at weak campaigns, the team held spend, watched the data, and waited for cleaner signals. When performance improved later in the event, they pushed onward and finished strong. 

An apparel brand during the same period saw similar issues. But they decided not to wait. Reliable performance signals showed a shift toward back-to-school demand. So, the team pivoted mid-event and adjusted assortments to focus on back-to-school products, introduced bundle offers, and optimized product titles and keywords. That shift resulted in a 15% increase in sales year over year vs. the previous Prime Day. 

Both brands took different approaches, based on having the visibility and confidence to act on what the data was telling them. There wasn’t a one-size-fits-all approach, but the key was data and agility. 

But according to the 2026 commerce readiness index, a survey of 200 brand and retail leaders across the U.S. and U.K, many brands don’t have that level of confidence and visibility to make it work.

According to the survey data, economic instability and inflation are the top hurdles to market expansion for both retailers and brands. But close behind sit technology and data challenges, rising operational costs, shifting consumer behavior, and supply chain issues.  

Some of that drag is the market. But lot of it is self-inflicted. Using the survey responses from commerce leaders, these are the three internal drags you say you feel the most . . . and what you can about them. And one external headwind you can only prepare for. 

Hidden drag 1: Manual ecommerce operations that slow omnichannel growth 

This drag shows up in the everyday work that still runs on spreadsheets and email, even as your sales channels and partners multiply. 

According to the survey, leaders say they’re “stuck at spreadsheet speed.” Fully automated workflows are rare, with many retail and brand leaders saying that 26% to 50% of their workflows still rely on manual steps like spreadsheets and email. 

In practice, this might look like retail vendor analysts pulling late-order reports into Excel, pivoting them, emailing suppliers one by one, and manually editing product descriptions before listings go live. On the brand side, manual work often involves pulling performance signals from multiple platforms, reconciling conflicting reports in spreadsheets, and chasing analysts for validation before anyone can act. 

You can see the impact of that manual overload in the data. Nearly three quarters of leaders say they at least sometimes make decisions based on inaccurate or inconsistent data, and more than a third say it happens often or all the time. 

53% of retailers say they act on important performance signals within 48 hours; brands are more likely to need three to five business days. Even the fastest groups say they are still acting on incomplete or inconsistent data and largely manual processes. 

If you’re selling through marketplaces, dropship programs, retailer.com sites, and your own direct-to-consumer (D2C) website, this is more than an internal annoyance. Manual listing updates, inventory syncs, and routing decisions become the places where channels drift and small errors balloon across your entire network. 

To turn this drag into an advantage, teams are: 

• Automating onboarding of assortments, content updates, inventory synchronization, and order routing where possible. 

• Consolidating product, inventory, and order data so a change to a SKU is reflected wherever you sell it. 

• Replacing “hero” spreadsheets with shared rules and playbooks that run on current, accurate data. 

Hidden drag 2: AI running on messy product and inventory data 

This drag appears when AI is built on data that is incomplete, inconsistent, or scattered across systems. 

AI is already live for many of the retailers and brands surveyed. 41% of retailers and 29% of brands use AI-based automation across functions like pricing, inventory, and marketing, and another 57% of brands and 41% of retailers say they are getting ready to implement it. 

At the same time, nearly three in four leaders say AI is advancing faster than their organizations can apply it effectively. 

The gap is visible: 

• 49% of retailers and 62% of brands say they still struggle with too many manual processes. 

• 91% of retailers and 78% of brands say poor data quality is a challenge. 

The same leaders rolling out AI across pricing, inventory, and marketing are also telling us they don’t fully trust the data underneath it. When catalog attributes are inconsistent, stock numbers are unreliable, or order and return data live in different silos, AI trained on that information doesn’t fix the issues, it magnifies them. 

This can show up as: 

• Retail media campaigns bidding on SKUs that are already out of stock on key partners. 

• Pricing models making decisions based on incomplete fees or cost data in certain channels. 

• AI “optimizing” assortments based on stale sell-through and margin data. 

How to make AI actually useful 

Start by fixing the inputs. Clean up product data, improve inventory accuracy, and connect orders and returns back to their source channels so you know what really happened and where. 

Then shorten the path from insight to action. If every AI-driven price or bid suggestion still has to be pasted into a spreadsheet and debated in a meeting, you will never see the benefit.  

Apply AI where it matters most: margin pressure from fulfillment and logistics, inventory stock-outs, and wasted media on low-quality traffic. Those are natural places to focus AI, once the data is ready. 

Hidden drag 3: Margin erosion across marketplaces and retail media 

This drag shows up in the small gaps where money and customers slip away across channels. 

Brands say the biggest hits in the past year came from fulfilment, logistics, and product costs. Retailers point first to tariffs and trade disruptions. Both groups also call out discounts, paid media inefficiency, and listing errors or inaccurate product data as other margin drains. 

Retailers most often lose shoppers before checkout, especially when ads do not match the product experience or when payment fails. Brands are more likely to have problems after the sale, in customer care and returns or refunds. 

At the same time, 91% of retailers and 84% of brands say they have changed their marketing channel mix in the last year, often in response to shifting consumer behavior and strategy.  

Some examples where customers and profit are lost in operations: 

• Broken links or mismatched product pages that cause pre-checkout drop-off. 

• Ads driving to SKUs that are out of stock or unprofitable to ship once fees and costs are counted. 

• Returns and service policies that vary by channel, leaving some experiences noticeably worse. 

The fix is to connect performance metrics with operational and margin data to see where the problems really come from. Are you losing money because of traffic quality, or because of content, availability, or fulfillment issues that could be fixed centrally and rolled out across channels? 

The drag you can’t fix: External volatility and ecommerce expansion 

This force comes from outside your walls, but it still shapes how fast you can grow and where. 

When retail and brand leaders rank their top hurdles to expanding into new markets, economic instability and inflation come first for both. Close behind are technology and data challenges, rising operational costs, shifting consumer behavior, supply chain issues, regulatory complexity, and tariff or trade uncertainty. 

Tariffs in particular stand out. 46% of retailers and 60% of brands say they are at least somewhat concerned that tariff and trade shifts will disrupt their sourcing strategies. More than 60% of both groups say they are re-evaluating sourcing relationships to prepare, while many are also cutting business costs and investing in supply chain resilience. 

You cannot control that volatility. However, you can decide how much internal drag you stack on top of it. Focus on what you can change by building stronger operations, reducing technology barriers, and creating more flexible, integrated customer experiences so you can pivot channels, partners, and assortments when conditions change. 

For a deeper look at the data behind these drags and more benchmarks you can use in your own planning, download the full 2026 commerce readiness index report. 

The post Three hidden drags you can fix (and one you can’t) appeared first on Rithum.