Seven Things Brands Get Wrong About Order Management

When I ask retail leaders what an order management system (OMS) is for, I usually get one word back: omnichannel. Buy online, return in store, see stock everywhere. It isn’t a wrong answer, but it’s only part of the picture, and I’ve watched it lead brands to buy the wrong thing, put off the right thing, or decide they don’t need an OMS at all.

I’ve spent most of my career on the vendor side of order management, and I now work with brands as a Solution Architect at AbsoluteLabs. These are the seven misunderstandings I come across most often.

1. Thinking an OMS only matters if you're omnichannel

Omnichannel was the main reason retailers bought an OMS over the last decade, and since the pandemic most enterprise retailers either have one or know they need one. The side effect is that “OMS” now sounds like a solved problem, and brands that sell mainly through one channel, such as wholesale, assume it isn't for them.

In my experience, the conversation changes when the label does. When I talk instead about controlling when payment is taken, how each order is routed, and what happens when something goes wrong, teams who told me they didn't need an OMS recognise the problem straight away.

What people forget is that an OMS isn't just about omnichannel. It's about control and visibility of the order all the way through. In practice, an OMS is the single record of an order's life: when it was placed, when and how payment was taken, where each item shipped from, and what changed along the way. Any brand selling at volume needs that, whatever its channel mix.

2. Expecting the ERP to do the job

Most brands already have an ERP that holds stock and orders, so it's fair to ask why they need another system. The difference is what each is built for. An ERP is designed to keep the books right: what stock you have, where it is and what it's worth. It isn't designed for the complexity that comes with keeping the customer happy, such as making fast decisions about individual customer orders.

ERP routing tends to be fixed: orders from this region go to that warehouse. An OMS decides order by order, using rules the business sets and can change. For example, below a certain order value, don't split the shipment, because extra parcels cost more than the margin. Above it, choose whatever reaches the customer fastest. The OMS then passes the outcome back to the ERP so finance stays accurate. The two work best side by side, each doing its own job.

3. Judging an OMS on the good days

Demos show orders flowing smoothly. The real test is what happens when they don't: stock is damaged in the warehouse, a store count is wrong, a carrier misses a collection. Those items were already promised to customers.

Without an OMS, no single system can see every open order alongside every unit of stock, so the usual fix is to cancel and apologise. An OMS treats the problem as a new routing decision, finds the stock elsewhere and re-routes the order, often before the customer knows anything happened.

When you evaluate platforms, ask to see this. How a system handles exceptions tells you more than how it handles a perfect order.

And sometimes the stock was never really there: if availability on the website is out of date, you've made a promise you can't keep before anything has even gone wrong.

4. Treating returns as an afterthought

Most OMS projects focus on getting orders out. Returns get far less attention, even though they're often where a customer decides whether to buy from you again. If returns sit outside the OMS, the store taking back an online order can't see what was paid or how, refunds wait on manual checks, and the stock sits in a back room instead of going back on sale.

An OMS that links every return to the original order can refund the right amount to the right payment method, update stock straight away and get the item back where it can be sold. It's worth checking how a platform handles the awkward cases, such as a return from an order that shipped from two locations and was paid partly with a gift card.

5. Adding AI to customer service before fixing the data

Customer service is where many brands are starting with AI, and “Where's my order?” is the obvious first question to automate. But the answer is only as good as the information behind it. If order details are split across the ecommerce platform, warehouse system, carrier sites and ERP, the AI has to piece them together, and customers notice when it gets that wrong. You can't layer AI over bad data; it will just give you bad answers.

My advice is to start small. Point AI at one complete, reliable order history in the OMS, and leave decisions about orders to your rules and your people for now. That alone improves service answers and reporting.

It starts with the customer. When one order ships in three parcels, the OMS knows the status of every item, so it can send one clear update through your email platform instead of three emails and three tracking links.

The same is true for human service teams. As a director at an apparel retailer told Forrester:

“There's one interface, just one place to look. My customer care team doesn't need to think about which website the order came from.”¹

6. Keeping B2B orders in a separate system

B2B and B2C orders are complicated in different ways. In B2C, the hard part is deciding where an order ships from and keeping the customer informed. In wholesale and trade, fulfilment is often simpler; the complexity is in the order itself. Business customers agree quantities up front and call them off over time, commit to minimums, add one-off orders on top, and amend orders in ways a consumer never would. Each account can have its own terms.

Many brands handle this by keeping B2B in a separate system, often the ERP. Joanna sees three reasons to bring it into the OMS instead. You get one stock pool: if a style is slow in wholesale but selling online, you can move it without updating two systems by hand, while still reserving stock for each channel. There's less to reconcile, because one system makes the change and updates finance. And all your order data sits in one place for reporting, so you can respond quickly when demand shifts.

B2B orders do need more care in setup and integration, particularly around account-specific terms. That's where implementation experience makes the difference.

7. Assuming there's one way to roll it out

Most OMS programmes I have worked on take one of two shapes:

• Region-first: deliver everything you need in one smaller, lower-risk market, then roll out to others with local adjustments.
• Capability-first: fix the most painful problem everywhere, such as cancelled orders, then add capabilities until the OMS runs everything and old processes can be switched off.

Neither is right for everyone. It depends on where the pain is, how similar your markets are and how much change the business can absorb at once. Whichever route you take, a few things consistently help, and they match what retailers told Forrester¹: roll out in stages rather than all at once, plan around peak trading, and bring the right people from the business in early so the requirements are clear from the start.

And plan for after go-live. An OMS keeps improving only if someone in the business owns it and watches a few numbers, such as cancellations, split shipments and on-time delivery.

The bottom line

An OMS is about knowing and controlling what happens to every order, whoever placed it and wherever it ships from. Omnichannel is one result of that, not the reason for it.

At AbsoluteLabs we design, implement and integrate order management for B2C and B2B brands, on platforms including Kibo and Salesforce. If any of these sound familiar, get in touch with our team to see how we can help.

¹ The Total Economic Impact™ Of KIBO Order Management, a Forrester Consulting study commissioned by KIBO, November 2025.