By industry
Customer stories
See how our clients are using Zipchat and the awesome benefits it's bringing to their businesses!
AI Customer Support
Sales Conversion
Analytics & Insights
Platform Integrations
Enter your store URL to see how Zipchat would behave.
Every new marketplace you add brings its own order flow, return policy, and customer expectations, and your support volume grows with each one.
The most common support questions from multichannel buyers are order status (WISMO), returns, and product compatibility, all of which can be partially or fully automated.
Copy-paste support responses break down quickly because Amazon, eBay, and your own web-shop have different policies, timelines, and language requirements.
AI support tools can identify which channel an order came from and apply the correct policy automatically, without agents switching between dashboards.
Setting up your support infrastructure before you launch on a new channel is significantly cheaper than fixing it after tickets pile up.
Adding a new marketplace feels like a sales win, and it is. More channels mean more reach, more orders, and more revenue. What most guides do not tell you is that it also means more support tickets, more complexity, and a customer service operation that can break under the weight of its own growth.
This article breaks down exactly why multichannel selling multiplies your support volume, what the most common problems are, and how AI tools are helping small and midsize merchants manage it without hiring a full support team.
The relationship between channels and support volume is not linear. When you go from one channel to two, your support load does not increase by 50%. It often doubles, sometimes more.
There are a few reasons for this.
Amazon buyers expect next-day responses and prepaid return labels. eBay buyers are accustomed to a more negotiable, message-heavy experience. Customers from your own web-shop expect personalized service and may know your brand. The same product sold across all three channels generates different support behavior from three different types of buyers.
Without a centralized order management system, your support team has to log into three separate dashboards to answer a single question about order status. That is three logins, three interfaces, and three times the chance of giving the wrong answer. The cognitive load alone slows response times significantly.
Your web-shop might offer a 60-day return window. Amazon enforces its own 30-day policy. eBay has its own buyer protection rules. When a customer asks, “can I return this?”, the correct answer depends entirely on where they bought it, and getting it wrong creates a second ticket, a complaint, or a negative review.
Hiring support staff takes time. Ticket volume grows the moment your first order ships on a new channel. Most growing merchants spend their first few weeks on a new marketplace in reactive mode, dealing with a backlog they did not anticipate.
Note: Industry estimates suggest that support ticket volume increases by 4070% when a merchant adds a second marketplace, depending on the category and average order value. These figures vary significantly by product type and return rate. Track your own baseline before launching on a new channel so you have a reference point.
Across marketplaces and direct channels, the same categories of questions account for the majority of inbound support volume. Knowing what they are lets you prepare answers before your customers ask them.
WISMO, short for “Where Is My Order?”, is the single largest category of support tickets for ecommerce merchants. It typically represents 30-50% of all inbound contacts, depending on the category and shipping speed. On marketplaces where customers cannot easily track orders themselves, this number is higher.
Return questions are the second most common category. The complication for multichannel sellers is that the answer changes depending on which channel the purchase was made on. An Amazon return goes through Amazon’s return center. A return from your web-shop follows your own process. Giving the wrong instructions adds a follow-up ticket and frustrates the customer.
Pre-purchase questions about compatibility, sizing, materials, or specifications arrive on every channel. These cannot always be automated, but they can be routed quickly if your product data is accurate and complete across all channels. Incomplete product data on one marketplace generates more of these questions than a well-optimized listing would.
These require fast resolution and usually involve a replacement or refund. The process differs between channels. Amazon has a specific resolution workflow. Your own web-shop gives you more flexibility. Having documented processes for each channel before the first complaint arrives is faster than figuring it out under pressure.
Customers do notice pricing differences across channels. This is more common than most merchants expect, especially if promotional pricing or channel-specific discounts are applied inconsistently. Clear pricing logic and consistent product data across channels reduce the volume of these questions.
Many growing merchants handle multichannel support by writing one response template and adjusting it manually for each channel. This works when you have 20 orders a week. It stops working at 200.
When you sell on Amazon, you agree to Amazon’s seller policies, including their return and refund rules. Sending a customer an incorrect policy from your web-shop is not just unhelpful; it can expose you to a marketplace dispute you will lose. The same logic applies to eBay’s Money Back Guarantee and bol.com’s seller terms.
Amazon support interactions are typically formal and resolution-focused. eBay has a more conversational culture. Your own web-shop customers expect a brand voice that reflects your store. Copy-pasting a response written for one context into another creates a mismatch that customers notice, even if they cannot articulate why.
Amazon requires sellers to respond to customer messages within 24 hours, including weekends. eBay has similar expectations. If your support process treats all channels with the same priority queue, you risk missing marketplace-specific deadlines and receiving performance warnings.
An Amazon order number looks nothing like a WooCommerce order number or a bol.com reference. A copy-paste template that includes “your order #[ORDER_ID]” requires manual editing every single time. Automated systems can inject the correct reference, tracking link, and carrier information for each channel without human intervention.
The foundation of effective multichannel AI support is channel identification, knowing before the first response is sent, where the customer made their purchase. Without this, the AI cannot apply the correct policy, tone, or resolution path.
Marketplace order confirmation emails come from recognizable sender addresses. Amazon uses @amazon.com domains. eBay uses @ebay.com. An AI system trained on these patterns can identify the originating channel from the customer’s forwarded email or the context of their message before a human reads it.
Every marketplace uses a distinct order number format. Amazon order numbers follow a pattern like 12345678901234567. bol.com uses a different numeric structure. An AI system can be trained to recognize these formats and route the ticket to the correct policy and response set automatically.
The most reliable method is a direct integration between your support tool and your order management system. When a customer contacts support, the AI queries the order database using their email address or order number and retrieves the purchase channel, order status, tracking information, and relevant policy, all before a human agent sees the ticket.
This is where the connection to your product and order data infrastructure matters. Tools like Koongo centralize order data from multiple marketplaces into a single system, which gives your support tool or AI agent a single place to query. When an Amazon order, a bol.com order, and a Shopify order all land in the same order feed, your support AI can retrieve the right details from one source rather than switching between three separate integrations.
WISMO tickets “Where is my order?” are the most automatable category in ecommerce support. The information needed to answer them (order status, shipping carrier, tracking number, estimated delivery date) is almost always available in structured data. The challenge is retrieving the right data for the right channel.
When a customer asks about an Amazon order, the tracking information lives inside Amazon Seller Central. An automated system needs either an API connection to Amazon or a data feed that pulls order and shipment status regularly. Once that data is accessible, the AI can respond with the correct tracking link and estimated delivery date without human involvement.
The same logic applies to bol.com, eBay, and any other marketplace. Each has its own order API. Each requires a separate connection unless you are using an order management platform that aggregates them.
Direct-to-consumer orders placed on your Shopify or WooCommerce store are typically easier to automate because you control the data. Your store’s order management system holds the tracking number, carrier, and status. An AI agent can query this directly using the customer’s email address or order number and return a complete answer within seconds.
Automation fails when order data is incomplete or delayed. If your marketplace orders are not syncing into your central system quickly enough, the AI will return stale or missing information, which is worse than no automation at all. This is why the quality of your order sync infrastructure directly affects the quality of your support automation.
Note: Merchants who automate WISMO replies typically report handling 3045% of their total inbound volume without human involvement. This figure comes from industry benchmarks and will vary based on your product category, average delivery time, and carrier reliability. Measure your own WISMO rate before implementing automation so you can track the actual impact.
Return policy management is where multichannel support gets genuinely complex. Each marketplace sets its own rules, and those rules override your own preferences as a seller. An AI agent that does not account for this will give customers the wrong instructions and create more work than it saves.
Before deploying any AI support tool, document the return policy for every channel you sell on. This means:
This policy map becomes the knowledge base your AI agent draws from when a return question arrives.
Once the AI identifies which channel the customer bought from (see the section above on channel identification), it applies the corresponding policy automatically. A customer who bought on Amazon gets Amazon’s return instructions. A customer who bought from your web-shop gets your web-shop’s return process. The agent does not need a human to make that distinction.
Not every return question fits a standard policy. A customer who bought a product that arrived damaged, or who is requesting a return outside the standard window, needs human review. A well-configured AI agent identifies these exceptions and routes them to a human agent with the full context already attached, channel, order details, policy, and the customer’s message. The agent handles the straightforward cases; humans handle the edge cases.
Marketplace return policies change. Amazon updated its return policy terms multiple times in the last two years. When a policy changes, every automated response that references it needs to be updated. Assign one person to monitor policy updates for each channel you sell on and update your AI’s knowledge base when changes occur.
Most support problems after a marketplace launch are not caused by a sudden surge in difficult customers. They are caused by a support infrastructure that was not prepared before the first order shipped. Here is what to have in place.
The numbers below are based on industry-reported benchmarks and patterns observed across multichannel ecommerce operations. They are not specific to any one business. Use them as directional estimates when planning your own expansion, and measure your actual baseline before and after launch so you have real data to work with.
Marketplaces with strong buyer protection programs (Amazon, eBay) tend to generate higher support volume than direct channels because buyers are more likely to contact the seller directly before initiating a marketplace dispute. Platforms with clear self-service tracking (some courier integrations) reduce WISMO volume.
Merchants who implement WISMO automation before expanding to a new channel report a meaningfully different experience than those who add the channel first and automate later. The key difference is that automated WISMO replies handle a predictable, high-volume ticket category from day one, which keeps the human queue manageable during the launch period.
Note: All figures in this section are industry estimates based on publicly available benchmarks and merchant-reported data. Your actual numbers will depend on your product category, average order value, carrier performance, and buyer profile on each channel. Establish your own baseline metrics before and after launch for accurate comparison.
Merchants who expand to a new marketplace without setting up support infrastructure first consistently report the same pattern: a manageable first week, a difficult second week as order volume grows, and a third week spent firefighting while trying to simultaneously manage the backlog and serve new customers. The time cost of catching up is typically higher than the time cost of preparing.
Industry estimates suggest 40-70% for a second marketplace like Amazon, though this varies significantly by product category and return rate. Measure your baseline ticket volume before launch so you can track the actual change for your business.
WISMO stands for “Where Is My Order?”, a customer asking for an update on their shipment. It is the most common type of support ticket in ecommerce and typically represents 30-50% of total inbound volume. It is also the most automatable category because the information needed to answer it (tracking number, carrier, status) is structured data.
Yes, if it is configured correctly. The AI needs to identify which channel the customer bought from first, then apply the correct policy for that channel. This requires mapping your policies per channel before deploying the agent, and keeping those policies updated when marketplaces change their terms.
Through order number format recognition, email domain matching, or a direct query to a centralized order management system. The most reliable method is an integration between your support tool and your order data, so when a customer provides their order number or email, the system can retrieve the channel, status, and applicable policy automatically.
Before. Ticket volume arrives with the first order, not weeks later. Setting up WISMO automation and policy documentation before launch means you start from a manageable position rather than spending your first month in reactive mode.
Koongo is not a support tool, but it solves a problem that directly affects support quality: fragmented order data. When your marketplace orders from bol.com, Amazon, eBay, and others sync into a single system, your support team or AI agent can query order status, tracking, and channel details from one place rather than switching between multiple dashboards.
Order status questions (WISMO) are consistently the highest-volume category, typically accounting for 30-50% of all inbound support contacts. Return and refund questions are the second most common category.
Every new marketplace you add multiplies your support complexity. The volume increase is predictable. The policy differences are structural. And the fragmentation of order data across multiple systems is the root cause of most support errors.
The merchants who manage this well are not the ones with the largest support teams. They are the ones who set up their infrastructure before the first order ships, documented policies per channel, centralized order data, and automated handling for the ticket types that do not require human judgment.
WISMO automation alone can handle 30-45% of your inbound volume without a human agent. That is a significant reduction in cost and response time, and it becomes possible the moment your order data is accessible in one place.
If you are expanding to a new marketplace and want your order data centralized before your first ticket arrives, Koongo’s Marketplace Manager pulls orders from 40+ marketplace platforms into a single feed which your support tools can query directly. You can explore it on a free plan at koongo.com, with no credit card required.
koongo.com 500+ channels Plans from €24/month Free plan available
Author: Jiří Zahrádka
Jiří Zahrádka is the CEO of Koongo and an expert in feed management and marketplace connections.
