What Is Overselling? Causes, Prevention and What to Do Next

Overselling in e-commerce is when a store accepts more orders for a product than it has stock available to ship. The orders look normal at checkout, so the gap usually surfaces later, at pick, or when the shopper asks where the order is.

This page is about overselling in e-commerce, not an airline overbooking a flight or a salesperson pushing too hard.

Most oversells start the same way: the storefront's stock number and the warehouse's stock number disagree, and the storefront keeps selling. I ran e-commerce at SurfStitch and P.E. Nation, and the systems behind every order there did their own jobs well. As I put it later, "they're not designed to detect when a customer does not get what they want on time". An oversell is the clearest case of that. Every order is a promise. Keeyu keeps it.

What is overselling?

Overselling is selling stock you do not have: the storefront shows a product as available, a shopper buys it, and the warehouse cannot fill the order. Cin7 defines it as a business accepting "more orders for a product than it has available in stock".

It is different from a backorder. A backorder is a sale you chose to make against stock that is coming, ideally with a ship date the shopper saw before paying. An oversell is a sale the store did not know it was making. The order is the same, but the promise behind it is different: one was made on purpose, the other by accident.

That is why I describe overselling in terms of the promise rather than the stock count. When someone checks out, the store has told them they will get this item, on a date. If no system can see that the item is not there, nobody owns that promise until the shopper notices it broke. Keeping that promise is what we mean by the delivery promise, and an oversell breaks it before the parcel has even been packed.

Why overselling happens

Overselling happens when the number the storefront sells against is out of date or wrong. Deposco lists the usual causes as "inaccurate stock counts, insufficient safety stock, a buying surge, or slow inventory syncs to other systems".

My take is that sync lag explains how an oversell happens, but not why it stays hidden for days. As I said on The Breakout CEO podcast, customer service is "rewarding itself for fixing problems fast versus preventing problems". Nobody is measured on catching an oversell at the moment of the sale, so it waits until a shopper reports it. Peak days make it worse, which is why stock sync belongs in any plan for peak-season operations.

In the calls our team has had with retailers, the pattern repeats. The storefront reads its stock from an inventory system, an ERP or a 3PL feed, and that feed updates on a delay: twice a day, on a scheduler, or one way only. Or the feed is current but does not match the floor, because of mislabeled boxes, stock adjusted by retail staff, or transfers between locations still in transit.

The common causes, starting with the most frequent:

• Sync lag between storefront and warehouse or ERP. Every minute between syncs is a window in which the store sells against a stale number.
• Multichannel selling. The same unit is listed on your site, a marketplace and a retail POS, and two channels sell it before either hears about the other.
• Flash sales and launches. Ryder points out that a promotion puts a high volume of orders into a very condensed period, which is when sync lag stops being theoretical.
• Bundles and variants. Cin7 notes that a product sold as part of a bundle can still show as available on its own after the bundle stock is used up.
• Preorders and backorders. Shopify has a "Continue selling when out of stock" option, described in its help center. Switched on and forgotten, it turns every sale past zero into a promise with no stock and no date behind it.

A worked example of the exposure window

You can size your own risk per SKU with two numbers. Orders exposed per sync window = minutes between syncs x orders per minute for that SKU. Then compare it with what is left: if the exposed orders are larger than the units on hand, you can oversell before the next sync. A SKU with 30 units left, selling 4 a minute on a store that syncs every 15 minutes, has 60 orders exposed against 30 units, so it can oversell by up to 30 before anyone sees it. The same SKU on a twice-daily sync at 1 order a minute has 720 orders exposed, which is why a slow feed matters even on a quiet day.

What to do when you oversell: cancel, backorder or substitute

When an oversell has already happened, you have three options for each affected order: cancel and refund, backorder with a new date, or substitute a similar item. The right choice depends on when stock is coming and what the shopper wants, and in the US the shopper's consent is required for a delay or a substitute.

The last row is a split shipment, and it needs its own message, or the shopper reads a half-empty box as a mistake.

Decide once per group of affected orders, and tell the shopper first. The email that works names the item, says it is no longer in stock, and gives real choices. On The E-Commerce Edge I described the shape we use: you bought this, it is no longer in stock, here is an alternative we do have, or a full refund, or we have one on the way and it can ship then. The shopper picks, and the order is changed to match. What fails is the generic apology with a discount code and no order number, because it creates the second wave of contacts.

This is the core of proactive instead of reactive support: the shopper hears about the problem from you, with options, before they have to ask.

The FTC Mail Order Rule and the delay notice

In the US, the FTC's Mail, Internet, or Telephone Order Merchandise Rule sets the legal floor for an oversold order. If you cannot ship within the time you stated, or within 30 days if you stated none, you must get the shopper's consent to the delay or promptly refund them without being asked.

What the FTC's business guide to the Rule requires:

• A reasonable basis for any shipping time you advertise, or for shipping within 30 days if you advertise none. The Rule text is at 16 CFR 435.2.
• A delay notice sent "reasonably quickly", by any means including email, so the shopper has time to decide.
• In that first notice, a revised ship date or a statement that you cannot give one, a statement that the shopper can cancel for a full and prompt refund, and a free way to cancel.
• If the first notice gives a definite revised date 30 days or less away, it must tell shoppers that not responding counts as consent to the delay. A renewed notice for any further delay cannot treat silence as consent.
• If the revised date is more than 30 days away, or you cannot give one, the notice must say that the order will be canceled automatically unless the shopper agrees to wait within the original shipping time plus 30 days. With no date, also give the reason for the delay and say the shopper can cancel any time before you ship.
• Refunds within seven working days of cancellation, or within one billing cycle where you are the creditor. Rule-required refunds cannot be paid as store credit, vouchers or scrip.
• No substitutes without the shopper's express agreement beforehand. For a backordered item, the FTC says the Rule allows only two responses: get the shopper's consent to the delay, or give a full and prompt refund.

That last point changes how an oversell should be handled. Quietly shipping a different color is not an option under the Rule unless the shopper agreed. Telling the shopper first, with a choice, is the legal minimum as well as good service.

This is a summary for US merchants, not legal advice. State consumer law and card network rules may also apply, so check them with your own counsel.

What overselling costs a store

An oversold order costs more than the refund. The store pays for the cancellation or the substitute, the time to contact the shopper, any goodwill discount, and the follow-up contact when the shopper does not understand what happened.

A simple way to cost it: cost per oversold order = refund handling + agent minutes x loaded cost per minute + discount or credit given + the cost of any second contact. Multiply by the number of affected orders. In a demo of our oversell workflow, I showed five oversold orders that take about 12 minutes each to resolve by hand, an hour of work, against 22 seconds automated. At that rate, 150 affected orders is 30 hours of manual work before anyone has fixed the cause.

In the retailer calls we analyzed, the oversell was typically found at pick, days after checkout, and the CX team then emailed each affected shopper by hand with a refund, a substitute, store credit or a discount code. Bulk email tools could not include the order number and SKU, so shoppers wrote back asking what had been canceled. The fix created a second wave of contacts. Subscription products had their own version: when a subscribed SKU hit zero on the storefront, the renewal failed and nobody told the subscriber.

The other costs are slower. Late or missing orders show up as where is my order tickets. And the shopper who was told yes and then no remembers it the next time they choose where to buy. Each of those second contacts is a ticket that exists only because the oversell was found at pick instead of at checkout.

How to prevent overselling

You prevent overselling by making the number each channel sells against smaller than, or equal to, what you can actually ship, and by keeping it current. The levers below each come with a decision rule.

The buffer is the most useful of these. Deposco recommends withholding a safety buffer from the available-to-promise number published to your sales channels, at the network or fulfillment location level. Shopify's guide to overselling recommends syncing stock across channels and auditing physical counts against the system so the number stays honest. Shopify's guide also recommends setting reorder points so you restock before you run out.

Each lever narrows the window, and none of them closes it. A mislabeled box or a late feed will beat any buffer on the wrong day, and Ryder notes that manual methods struggle once hundreds of orders are placed in an hour. My frustration with the tools that sit after checkout is the one I wrote about earlier this year: none of them prevents issues, they all react to them. The useful addition to a buffer is an alert that fires when it fails. Every order is a promise. Keeyu keeps it.

How to detect an oversell in minutes

You detect an oversell by comparing, continuously, the quantity ordered for each SKU with the quantity available across the storefront, the ERP and every fulfillment location, and alerting the moment orders exceed stock. Most stores only make that comparison once, at pick, which is why they find out days later.

A detection routine, in order:

1. Per SKU, compare open orders with available units at each location every few minutes, not once a day.
2. Alert when open orders exceed available stock, and when a SKU's sales rate would exhaust it before the next sync.
3. Group the affected orders by SKU so one decision covers the whole group.
4. Pause the SKU on the storefront, or switch it to backorder with a date, before the group grows.
5. Contact the affected shoppers before they contact you.

The oversell is the hardest case to catch, which is why we built for it first. When we started building Keeyu's workflows with our first two beta customers, my brief to the team was to build the easiest case and the hardest one: an order waiting for collection, and an order that is out of stock. If we could prove both, everything in between would follow. The out-of-stock case is hard because no single system holds the answer. You need what is left at each location, what is inbound, what the shopper bought, and what they would accept instead, all read at the same moment.

That is the gap I think the industry's tools leave open. The systems that manage an order, from the storefront and payment gateway to the warehouse, OMS, inventory system, carrier and helpdesk, each do a specific job, and none of them is built to detect a missed promise. Post-purchase is spread across many point solutions, and none of them is synced to the shopper's promise. Detection in minutes is the piece that sits across all of them.

Where Keeyu fits, and where it does not

Keeyu is not inventory software. It does not count, set or forecast stock, and it does not replace your inventory system, ERP or OMS. It sits across those systems and watches for the moment the storefront and the warehouse disagree about an order.

In practice, Keeyu compares quantity ordered with quantity available across the storefront, the ERP and your fulfillment locations, flags the affected orders, checks other locations and inbound stock, and runs the response you have approved: reroute, offer a backorder date, offer an alternative, or refund, and then update the order to match the shopper's choice. That is Detect. Decide. Act. applied to the oversell. It is how proactive e-commerce operations differs from managing the fallout in the inbox. Helpdesk manages complaints. Keeyu prevents them. We still write the update into Gorgias or Zendesk, so your team sees everything in one place.

If you are choosing software to prevent or resolve oversells, we compare the options in our list of tools that prevent or resolve oversells. You can also see how Keeyu detects order issues and acts across your existing systems.

Keeyu gets shoppers what they want, on time, as promised. If you want to find the oversells your current setup is missing, see it on your own store. Every order is a promise. Keeyu keeps it.

• 1. Cin7, "Overselling: How to Prevent It Across Channels", https://www.cin7.com/blog/overselling-ecommerce/. Definition of overselling as accepting more orders than available stock, and that bundles and variants are more prone to it.
• 2. Deposco, "Prevent Overselling - How it Happens and How to Avoid it", https://deposco.com/blog/prevent-overselling/. Causes (inaccurate stock counts, insufficient safety stock, a buying surge, slow inventory syncs) and withholding a safety buffer from the available-to-promise number published to sales channels.
• 3. Ryder, "Preventing Overselling During Sales", https://www.ryder.com/en-us/insights/blogs/e-comm/how-to-prevent-overselling-during-sales. Flash sales put a high volume of orders into a very condensed period, and manual methods struggle when hundreds of orders are placed.
• 4. Shopify Help Center, "Setting up inventory tracking", https://help.shopify.com/en/manual/products/inventory/setup/set-up-inventory-tracking. The Continue selling when out of stock option.
• 5. Shopify, "Overselling: What It Is and How to Prevent It", https://www.shopify.com/blog/overselling. Syncing inventory across sales channels and auditing physical stock against system records.
• 6. Federal Trade Commission, "Business Guide to the FTC's Mail, Internet, or Telephone Order Merchandise Rule", https://www.ftc.gov/business-guidance/resources/business-guide-ftcs-mail-internet-or-telephone-order-merchandise-rule. The 30-day default, delay option notices, consent by silence, refund timing, no store credit for Rule-required refunds, and no substitution without express agreement.
• 7. Legal Information Institute, Cornell Law School, "16 CFR 435.2 - Mail, Internet, or telephone order sales", https://www.law.cornell.edu/cfr/text/16/435.2. The Rule text: a reasonable basis for shipment representations, and the requirements for delay notices and revised shipping dates.