AI in Ecommerce: 7 Ways to Get Started in 2026

For many businesses, using AI in ecommerce has moved from “nice to have” to “core infrastructure.” It can do everything from writing product descriptions to flagging fraudulent orders in milliseconds. Three out of four ecommerce business owners now use AI tools, according to the Shopify Merchant Survey.*

The impact is measurable. McKinsey found that 88% of organizations use AI in at least one business function—up from 78% a year earlier. Generative AI alone could add $400 billion to $660 billion a year in annual value for retailers. And you don’t need a computer science degree or a massive budget to capture some of that value for your own online store.

This guide covers what AI for ecommerce actually looks like, the seven highest-impact use cases, and exactly how to add AI to your store’s toolkit. Below, you’ll find real-world examples, fresh data, and practical next steps.

What is AI in ecommerce?

Artificial intelligence (AI) lets machines perform tasks like reasoning, learning, predicting, and decision-making—tasks that typically require human intelligence. AI in ecommerce lets you use data you already collect (like clicks, purchases, or supply chain activity) to make smart, real-time decisions across every stage of the online shopping experience.

Shopify senior developer Alex Pilon points out that this democratizes tech for non-coders: “Lowering opportunity cost means more people can participate in the economy. … AI really opens the door for anyone of any technical background to bring their ideas to life.”

The financial impact is substantial. The latest AI statistics show smart product recommendations powered by AI can triple revenue, more than double conversion rates, and increase order values by half. AI can also create personalized shopping experiences and reduce operational costs. 

“The cost of any software effort is essentially trending toward zero,” says Alex. “If you’re a Shopify merchant, you can redesign your site for Valentine’s Day and revert it the next day. What seemed absurd only a few years ago would now seem normal.”

With AI, your ecommerce teams can:

• Write product descriptions and generate images that match your branding
• Recommend the right products to each shopper to increase average order value (AOV)
• Predict demand and manage inventory to prevent stockouts or overstocking
• Spot and stop fraudulent transactions in real time

AI ecommerce tools like Shopify Sidekick plug directly into your storefront without code. You can use them to start, manage, and grow your business.

Types of AI technologies used in ecommerce

AI isn’t one technology—it’s a collection of powerful models. The main types driving results in ecommerce include:

Generative AI and large language models (LLMs)

Traditional artificial intelligence follows rigid rules: if X, then Y. Generative AI is different. Powered by large language models (LLMs), it creates new content rather than simply classifying or routing existing data. 

That distinction matters. AI generated content unlocks use cases that rule-based systems can’t handle, like drafting a product description in your brand voice or generating a personalized email subject line for each customer segment.

The Shopify Merchant Survey found that among merchants already using AI, content generation is the dominant use case, at 69%, followed by marketing initiatives, at 38%. This shows that generative AI for creating product descriptions, emails, and campaigns has become the entry point for most ecommerce businesses.*

Here’s how you can use generative AI in ecommerce:

• Write SEO-friendly product descriptions in multiple languages.
• Power 24/7 chatbots that recommend products and answer pre-purchase questions.
• Generate personalized emails, SMS messages, and ad creatives for campaign launches.
• Create lifestyle or hero images for product pages and social media.

These capabilities make generative AI the most accessible entry point for ecommerce businesses looking to scale AI-generated content without a dedicated engineering team.

Generative AI can even generate unique brand name suggestions for your business. Open Shopify’s AI-powered business name generator and type a few words describing your business idea, product, or industry. You’ll see a list of brand name suggestions based on your input:

Read more: AI for Business: Applications, Benefits, and Risks

Computer vision and visual search

Computer vision helps machines interpret the content of images and videos. In ecommerce, this powers visual search. A shopper uploads a photo of a dress they spotted on Instagram, and your store instantly surfaces similar items by matching attributes like shape, color, pattern, and style. This goes far beyond what a traditional search bar or text-based search results can capture.

Visual search reduces friction at the top of the funnel. Instead of struggling to describe what they want in words, shoppers simply show you. Image recognition also helps retail teams catch damaged returns before they hit the warehouse, streamlining operations and reducing reverse logistics costs.

Voice search is a related capability worth watching. As more shoppers use voice assistants to discover and reorder products, ensuring your product data is structured for voice queries becomes another way to meet customers where they are.

💡 Pro Tip: Install AI tools like.  ViSenze or Snap Search into your Shopify store to activate visual search for your customers.

Predictive analytics and machine learning

Predictive models powered by machine learning algorithms use real-time data—site traffic, promotions, weather, social media trends—to guide daily decisions. These AI algorithms help you avoid stockouts or overstocks. They can analyze customer behavior and identify patterns in historical sales data that humans miss. Six in 10 retail buyers say AI has already improved their demand forecasting accuracy.

These same models drive end-to-end supply chain optimization. They can help by anticipating shipping delays, recommending the fastest fulfillment route, adjusting safety stock levels during promotions, and triggering automatic reorders when thresholds drop.

“Our AI system detected viral TikTok trends and unseasonable weather patterns that spreadsheets failed to identify,” says Daniel Lewis, CEO at legal AI company LegalOn. “The system correctly forecasted a 47% increase in linen dress demand, which led us to redirect inventory from regions with excess stock before the trend peaked. We avoided $2 million in dead stock and experienced 32% fewer stockout-related lost sales.”

Benefits of using AI in ecommerce

Spending less time and coming up with more ideas—that’s what you can expect to see as primary benefits from AI adoption, according to the 2025 Shopify Merchant Survey.* Businesses earning more than $100,000 annually are significantly more likely to also recognize AI’s ability to scale operations with a small team and reduce operating costs.*

Increased sales

AI creates a more efficient sales process by gathering and analyzing customer data—including customer purchase history, browsing history, and preferences—to personalize your sales funnel. With more data, you can reach the right prospects with the right message at the right time.

Better and more personalized customer service

AI analyzes customer feedback and behavioral data from multiple touchpoints to enhance customer service and improve customer interactions. This goes beyond basic customer segments into hyper-personalization—delivering individualized offers and personalized shopping experiences based on each shopper’s real-time context.

Collecting this customer data helps you identify shopper preferences so you can create custom offers that encourage purchases. Brands like Ruti have implemented virtual sales associates, leading to higher conversion rates and average order values. The result? Enhanced customer satisfaction and stronger customer engagement across every channel.

Reallocation of time and resources

AI automates tasks and processes like email marketing, order fulfillment, customer service, and payment processing. These automations help you reduce labor costs and improve operational efficiency so you can spend less time on maintenance and more time innovating.

How to use AI in ecommerce: 7 applications and real-world examples

1. Personalized product recommendations
2. Conversational commerce and AI assistants
3. Fraud detection and prevention
4. Predictive inventory management
5. Dynamic pricing and revenue optimization
6. Customer retention and lifetime value prediction
7. Generative AI for content creation

You can use AI across every part of your ecommerce operations—from helping customers find products to optimizing prices. Here are seven use cases:

1. Personalized product recommendations

AI recommendation systems analyze customers’ shopping carts, past purchases, and browsing history to deliver personalized product recommendations. The best systems go beyond collaborative filtering into hyper-personalization, using natural language processing (NLP) to understand how shoppers describe products and computer vision to match desired items.

AI-powered features like “People also bought” or “Customers also viewed” suggest complementary products based on size, color, shape, fabric, and brand. These cross-sell and upsell opportunities appear on product pages, in checkout flows, and in post-purchase email follow-ups—each touchpoint tuned to the individual shopper’s behavior in real time.

Here’s how AI-based product recommendations appear in fitness brand Gymshark’s “People also bought” carousel on its checkout page:

Here are specific ways to target customers with personalized recommendations:

Read more: How Ecommerce Product Recommendations Drive Sales

2. Conversational commerce and AI assistants

AI chatbots and virtual assistants work as customer service representatives for your ecommerce business. Using NLP, generative AI, and real-time store data, chatbots greet visitors, recommend products, and start orders. They can also answer post-purchase customer queries like, “Where’s my package?” handling the ecommerce customer service tasks that would otherwise require a live agent.

Consumer demand is surging. A Nosto/Censuswide survey found that 72% of consumers expect AI shopping assistants to help them shop online, with that figure rising above 80% among consumers under 45. The most-wanted features include deal and price-drop alerts, personalized recommendations, and gift inspiration.

Voice search is an increasingly important channel within conversational commerce. As shoppers grow comfortable asking voice assistants to reorder household staples or compare product specs, stores that structure their product data for voice queries gain an edge.

Conversational AI tools also boost your bottom line. Implementing AI voice and chat agents into contact centers cancut cost per call by nearly 50%. Generative AI assistants have helped agents resolve 14% more tickets per hour while reducing handling time by 9%.

You can use chatbots and virtual assistants for:

• Handling efficient customer interactions. Chatbots can process simple transactions, take orders, and provide personalized offers across multiple POS channels.
• Collecting customer data. Chatbots can gather customer needs and information like sizing preferences and inquiry reasons, guiding product development strategy decisions and revealing customer preferences at scale.
• Improving checkout experience. Integrate a chatbot into your checkout page so customers can ask about product details, stock levels, and shipping without leaving their cart.
• Providing 24/7 customer service. AI assistants can respond instantly around the clock, freeing your live agents to handle complex issues.

Set up Shopify Inbox on your store to support customers through live chat and boost revenue without increasing headcount.

“Shopify Inbox is a powerful tool,” says Rennie Wood, founder of Wood Wood Toys. “It helps me rescue sales after spending time, money, and energy getting a customer to that point. It pays huge dividends.”

3. Fraud detection and prevention

AI in ecommerce detects and prevents fraud by analyzing data, spotting anomalies, and monitoring transactions in real time. The technology identifies unusual patterns—like high-value transfers, multiple transactions within minutes, or purchases from unfamiliar locations—and flags them for investigation before they cost you revenue.

Machine learning (ML) models create user profiles based on customer behavior data like browsing habits, transaction history, and device information. They compare current behavior with historical patterns to catch fraudulent behavior the instant it deviates from the norm.

AI also powers smart payment optimization. AI-enabled payment routing selects the processor most likely to approve each transaction, reducing false declines and recovering revenue that would otherwise be lost. These systems monitor payment anomalies in real time—chargeback spikes, repeated failed attempts, unusual refund patterns—to proactively flag emerging risks before they escalate.

The emphasis here is on speed: real-time monitoring means threats are caught in milliseconds, not hours.

Also read: Shopify Protecting Millions of Merchants From Fraud

4. Predictive inventory management

AI can help you optimize inventory management by analyzing historical sales data and forecasting future demand. Real-time insights from sensors and RFID tags show what’s selling, where it’s going, and whether it’s coming from a store or warehouse.

Using AI demand planning tools, you can reduce inventory by 20% to 30% without hurting service levels. This frees up significant working capital you can reinvest in business growth. AI also brings end-to-end supply chain optimization. It can help with predicting shipping delays, recommending store-to-store inventory transfers based on regional demand, automating restocks, and keeping both your team and customers informed.

Here are specific ecommerce use cases for AI-powered inventory management:

Automate inventory management, fraud prevention, and order fulfillment workflows using Shopify Flow. You can also add a third-party forecasting app or your own machine learning model to improve accuracy.

“Usual stock management involves human involvement, but by setting up Flow, we’re able to save a huge amount of time and eliminate human error,” says Panos Voulgaris, a creative strategy director who partners with children’s brand Cozykids). “When dealing with a catalog of 6,000 to 7,000 products, that’s incredibly helpful. Flow makes it easy.”

Read the case study: Cozykids uses Flow and Launchpad to automate its processes, remove human error, and dramatically increase efficiency.

5. Dynamic pricing and revenue optimization

Instead of manually checking competitor pricing and adjusting your rates, AI-powered dynamic pricing solutions do it automatically. These systems watch real-time signals—site traffic, competitor prices, customer behavior, inventory levels—and adjust prices for every product to maximize your profits.

The real-time dimension of AI in ecommerce is critical: AI doesn’t run overnight and update prices in the morning. It responds to surges, competitor moves, and inventory shifts as they happen across every channel. You can set different strategies for your ecommerce website, Amazon, and marketplaces simultaneously, capturing margin where you can and volume where you need it.

Here are ways AI pricing works in practice:

This level of pricing optimization was once available only to enterprise ecommerce retailers; AI technology now makes it accessible to stores of any size.

Related reading: Price Optimization: A Definition and Complete Guide

6. Customer retention and lifetime value prediction

AI can spot which customers will stick around and which ones are about to leave—before they do.

It analyzes browsing history and habits, purchase frequency, user behavior on-site, and support interactions to score each customer for churn risk and future lifetime value. This means you can send the perfect offer at exactly the right moment, driving customer engagement and ensuring customer satisfaction.

A few ways you can use AI to improve retention:

• Churn alerts. AI algorithms spot red flags like repeat cart abandonment or longer gaps between purchases. When a valuable customer’s score drops, it triggers loyalty perks or targeted offers.
• Smart upsells. Using lifetime value forecasts and product preferences, AI suggests relevant add-ons, like offering a shaker bottle to your monthly whey protein subscribers.
• Win-back campaigns. AI re-engages at-risk customers with automated retargeting messages and emails, then stops outreach once they respond.

Related reading: 14 Customer Retention Strategies That Help Increase ROI

7. Generative AI for content creation

Generative AI can quickly produce marketing collateral like product copy, images, videos, and voice-overs. You can also use it to test whether your brand messaging resonates with your target audience.

Below are some ideas for using AI prompts for writing, SEO, and visual content:

💡 Pro Tip: Use Shopify Sidekick to write, edit, or translate product descriptions, headlines, and store content directly in your admin. It learns your brand voice and creates quality copy in minutes.

Implementing AI in your ecommerce business

Start by looking at your current resources, customer data, and workflows. Then decide what you want AI to accomplish.

Here’s how to approach it:

Assess your AI readiness

Before you invest in AI solutions, check these four areas:

• Strategic fit. Identify a specific business problem AI in ecommerce can solve (like “reduce stockouts by 15%”). Get clarity by asking every stakeholder, “Why do we need AI?” Everyone should have a clear, specific answer.
• Data quality. You need at least 12 to 18 months of clean, labeled data covering orders, web traffic, and your product catalog. If less than 10% to 20% of your data needs manual cleanup, you’re ready.
• People and process. Make sure you have a product owner, data lead, executive sponsor, and an agile workflow. Test this by mapping a process like pricing.
• Tech stack. Confirm your ecommerce platform supports AI solutions and APIs for inventory, pricing, and CRM. Whether you’re running a single ecommerce site or a multistore setup, this makes AI integration easier.

Start with small, low-cost AI implementations

The biggest barrier to AI adoption isn’t cost or complexity—it’s uncertainty. According to the 2025 Shopify Merchant Survey,* among merchants not using AI, 29% aren’t sure what AI tools can do, 29% don’t know where to start, and 26% don’t know which tool to use. Starting with built-in tools like Shopify Sidekick and Inbox eliminates these barriers.

Many early wins come from AI tools that won’t break the bank:

• Instant copywriting. Shopify Sidekick lets you write or translate product descriptions directly in your admin—at no extra cost.
• Live chat that sells. Turn on Shopify Inbox for a basic FAQ bot, then layer in generative AI once you have real transcripts to train on.
• Simple automation. Use Shopify Flow to auto-tag low-stock items or email suppliers.

Browse more AI tools for business to find the right fit for your store.

Measure ROI on AI investments

Merchants who measure results rigorously are more likely to see AI benefits, and are thus more likely to expand their AI usage. Shopify’s November 2025 Merchant Survey found that AI users demonstrate stronger financial discipline overall: 50% conduct regular P&L reviews compared to 39% of non-AI users.*

Here’s how to track returns on your AI investment:

• Choose one measurable KPI like gross margin, reduced refunds, or added revenue.
• Record a baseline for at least four weeks before introducing AI.
• Run an A/B test, e.g., 50% of traffic sees AI pricing, 50% sees manual pricing.
• Track both upside and cost, including app fees + team hours.
• Calculate payback using the formula net benefit ÷ monthly cost = months to break even. Aim for less than 12 months.

Challenges of using AI in ecommerce

The upside is real, but so are the hurdles and potential dangers of AI. Treat these as realities to plan for, not reasons to avoid AI.

Some key challenges of AI in ecommerce:

• High upfront and ongoing costs
• Data challenges
• Technical integration and legacy systems
• Talent shortage and skill gaps
• Bias and ethical risks
• Organizational resistance

High upfront and ongoing costs

AI implementation requires upfront investment in software, hardware, data infrastructure, skilled professionals, and third-party platforms. Ongoing expenses include model updates, data storage, and subscription fees.

Businesses earning more than $100,000 annually are significantly more likely to recognize AI’s cost-reduction benefits, according to a recent Shopify Merchant Survey, making AI ROI clearer as operations scale.* For smaller businesses, cost concerns can be significant. The key is starting with no-cost tools like Shopify Sidekick and Inbox to prove value before investing.

Data challenges

Ecommerce businesses face several data-related roadblocks when adopting AI tools:

• Data silos and integration. Your ecommerce data is often spread across CRMs, ERPs, web analytics, and marketing tools. Merging this data into a single, AI-ready system is complex and time-consuming.
• Data quality and governance. AI relies on clean, accurate, and consistent data. This requires you to have strong data practices, clear ownership, and access controls—all difficult to both establish and maintain.
• Limited data volume and variety. Some AI models need massive datasets to perform well. Smaller or newer businesses may not have enough data, or enough variations, which can skew results or limit model accuracy.

Technical integration and legacy systems

These technical challenges can make AI adoption difficult:

• Legacy systems. Many ecommerce companies still run on outdated platforms that aren’t designed for AI. They need to upgrade their retail technology first.
• Interoperability issues. New AI-powered tools must work smoothly with existing systems like inventory, payments, and marketing automation. This often requires custom development and can lead to unexpected technical problems.
• Ongoing model management. AI models need constant updates—development, testing, deployment, monitoring, and retraining. Managing this life cycle (also known as MLOps) requires tools and skills many ecommerce teams lack.

Talent shortage and skill gaps

It’s not just about hiring a data scientist. You need a team with skills in machine learning, data engineering, AI ethics, and business strategy. This kind of talent is hard to find and expensive to recruit. Training your existing team to understand and use AI in ecommerce is also a significant undertaking.

Bias and ethical risks

AI can reflect or reinforce biases found in historical data. This is especially concerning in personalized pricing, product recommendations, and fraud detection, where biased models can alienate customers or create discriminatory outcomes. Building trust requires transparency. 

Clearly communicate how AI affects customer experiences—what data you collect, how recommendations are generated, and what decisions are automated. Consider publishing an AI usage policy that explains your approach in plain language. Give customers control over their personalization settings so they can adjust or opt out.

The Nosto/Censuswide survey found that the top reasons consumers stop using AI for shopping are concerns about data handling (24%) and feeling AI is making decisions without their input (21%). Transparency and customer control directly address both concerns.

Organizational resistance

AI often changes how people work. Some employees may worry about losing their jobs or struggle to adapt to new tools and workflows. This requires strong change management, clear communication about how AI will augment their roles rather than replace them, and hands-on training.

The future of AI in ecommerce

AI is quickly becoming a key piece of how customers actually shop. Instead of browsing endless product pages, they’re talking to AI that understands exactly what they want. As Shopify’s Alex Pilon puts it, “We’re living through unprecedented technological change. … All the software development we did over the past 25 years is now accessible at your fingertips, in real time, to solve problems and boost efficiency.”

Four major AI trends will reshape ecommerce in the coming years:

Agentic commerce

Agentic commerce is the next evolution beyond chatbots and assistants. Instead of responding to prompts, AI agents act autonomously on a shopper’s behalf. This can include researching products across multiple stores, comparing prices, negotiating bundle deals, and completing checkout, all within a single conversational flow.

McKinsey estimates that by 2030, agentic commerce could orchestrate up to $1 trillion in US B2C retail revenue alone, with global projections reaching $3 trillion to $5 trillion. The technology is moving fast: Shopify is developing agentic shopping infrastructure that allows AI agents to tap into merchant catalogs and build carts across stores. OpenAI, Google, Amazon, and PayPal are all building agentic shopping services.

For business owners, agentic commerce means your product data, pricing, and policies need to be machine-readable and agent-friendly—not just optimized for human browsers. Structured data, clean APIs, and transparent pricing will determine whether AI agents recommend your store or skip it.

Read more: AI Agents: Harnessing Agentic AI for Ecommerce Businesses

Autonomous commerce

Autonomous commerce means shopping journeys that run themselves. AI agents can detect demand, curate products, set prices, answer questions, and handle fulfillment—all without human oversight.

Alex sees AI assistants as leveling the marketing playing field. “AI will slash the cost of entry to marketing and ad campaigns. … Having an assistant who understands your business and helps design, execute, and tweak strategy is a massive power-up. With access to your business data and tools, it becomes a marketing expert wired into your systems—a superpower.”

In practice, this looks like:

• Auto-replenishment. Your coffee subscription refills itself when a smart bin detects you’re running low.
• Voice-powered checkout. A voice assistant compares sizes, applies your loyalty points, and completes payment—all through conversation.
• Hands-free merchandising. AI bundles new arrivals, writes product descriptions, and schedules posts.

Tools like Shopify Sidekick and AI Website Builder are making autonomous commerce accessible today, handling routine tasks so you can focus on strategy and growth.

Sustainable AI applications

As AI for ecommerce models grows larger and more data-hungry, it requires more energy. This worries both regulators and eco-conscious customers.

Deloitte’s 2025 tech forecast warns that global data-center electricity demand could double to 1,065 terawatt hours by 2030—nearly 4% of all global electricity use.

Here’s how you can shrink your AI footprint:

• Train AI models during low-carbon hours in your cloud region.
• Choose smaller, efficient AI models that deliver most benefits with less power.
• Let AI pick the smallest shipping box for each order to cut waste and emissions.

Generative engine optimization (GEO)

As shoppers shift from typing keywords into Google to asking AI assistants for product recommendations, a new discipline is emerging: generative engine optimization (GEO). Traditional SEO optimizes content to rank in traditional search results. GEO optimizes content to be cited and recommended within AI-generated answers from tools like ChatGPT, Perplexity, Google AI Overviews, and Gemini.

Gartner predicts a 50% decline in traditional search engine traffic by 2028, as users increasingly prefer AI-powered interfaces. For ecommerce brands, this means your product descriptions, FAQ pages, reviews, and buying guides need to be structured in ways that AI models can easily interpret, extract, and cite—not just index.

Here are practical GEO steps to optimize your online store for AI assistants:

• Write clear, factual product descriptions with specific attributes (materials, dimensions, use cases).
• Structure FAQ and comparison content in question-and-answer format.
• Build a strong foundation of authentic customer reviews that AI models can reference.
• Ensure your product schema markup is clean and comprehensive.

Stay ahead: 2026 Ecommerce Trends: How Brands Are Planning Ahead

Is AI in ecommerce worth it?

AI has moved past the experimental phase. It’s now core infrastructure for the ecommerce industry—the kind of capability that compounds over time as your data improves, your models learn, and your workflows get smarter. The question is no longer whether to adopt AI in ecommerce, but how disciplined your implementation is.

AI adoption correlates with broader operational maturity. Shopify’s November 2025 Merchant Survey found that merchants using AI are more likely to validate business ideas through MVP testing (11% versus 3% of non-users), conduct customer research (10% versus 3%), and cite brand reputation as their competitive advantage (45% versus 36%).* This suggests AI has become part of a larger pattern of systematic, data-driven decision-making.

ROI depends on disciplined AI implementation. Follow these best practices:

• Start with a specific, measurable business problem rather than adopting AI for its own sake.
• Use built-in tools (like Shopify Sidekick, Inbox, and Flow) to test value before investing in custom solutions.
• Measure results rigorously with clear baselines and A/B tests.
• Scale only what works, reinvesting savings into the next high-impact area.

Accessible tools have lowered the barrier to entry. You can test AI on product descriptions today and see results within a week—no engineering team required. Brands that apply AI strategically to real business problems—rather than haphazardly adopt it—will gain a competitive advantage.

What to do next

For newcomers, Alex suggests treating AI like a business partner: “If I was starting with AI today, I’d interact with it as a thought partner. … Ask questions, build intuition, and let it expand what you think is possible."

Here’s your next move, based on where you are:

• Just getting started. Pick one high-impact area and try a no-code tool for a month, then measure results. Try Shopify Sidekick to write product descriptions or turn on Shopify Inbox for live chat. 
• Ready to scale. Automate repetitive tasks using your platform’s built-in AI tools or external apps. Shopify merchants can use Shopify Flow to automate workflows.
• Already experienced. Let AI bundle new arrivals, A/B test pricing, or draft SMS campaigns. Track the impact on margins and conversion.

*Based on a 2025 survey of 500 Shopify merchants conducted in English across Australia, Canada, the United Kingdom, Ireland, New Zealand, and the United States. Respondents were established merchants with two or more years on the platform. Results reflect the experiences of this specific sample and may not be representative of all merchants.