Walk into the back office of most growing UK retailers and you will find the same scene: an ecommerce platform, a till system, a marketplace connector, an accounting package and three spreadsheets that hold the real truth. Nobody planned this architecture. It grew, one channel at a time, and each addition was sensible on its own. The problem is not any single system. The problem is the gaps between them, and the gaps are where the money goes.

How retail ends up disconnected

Disconnection is a growth pattern. The website was added first, with its own stock table. Then Amazon, with its own feed. Then the warehouse moved to a new system because the old one could not handle volume. Each system works, and the integration between them is whatever the cheapest connector at the time could do. Often it is a nightly CSV export and a person who checks it in the morning. That person is not doing data integration. They are doing data archaeology, and they are the most expensive integration you own.

Cost 1: the stock record decays

When stock lives in several systems, the records disagree, and the disagreement compounds. The web platform sells from its table, the marketplace sells from its feed, and the warehouse ships from its own counts. Nobody's number matches, and the stock accuracy article explains what that does to every decision downstream. Disconnected systems are the most common cause of chronic stock inaccuracy, because every sync gap is a chance for reality to escape the record.

Cost 2: the copying tax

Count the hours people spend moving data between systems: rekeying orders, exporting and importing spreadsheets, emailing stock figures, rebuilding reports in Excel because no system has the whole picture. Multiply by the loaded cost of those people, and the figure is usually shocking. I have audited operations where the copying tax exceeded the licence cost of a proper ERP, which is the quiet argument for the selection process: the system you are hesitating to buy is often cheaper than the admin you are already paying for.

The Chartered Institute of Procurement & Supply makes the same point about purchasing processes specifically: manual rekeying between purchasing and finance is not just slow, it is where errors are born. Every handoff is an error opportunity.

Cost 3: decisions on stale data

The buying team needs a stock picture to plan the season. The merchandising team needs sell through to plan markdowns. The board needs a margin number to plan investment. With disconnected systems, every one of those decisions runs on data that is days old or manually assembled. Seasonal planning shows how much of the overbuying problem in retail is actually a data timing problem, and the margin analysis shows the same pattern in pricing and purchasing. Decisions are only as good as the freshness of the numbers under them.

Cost 4: the sync gap sells and cancels

The most visible cost of disconnection is the oversell. A nightly sync means a full day of exposure to double selling. The overselling problem covers the mechanics, but the financial point belongs here: every sync gap is a standing order to lose margin, and the loss grows with each channel you add. The retailer with five channels and hourly syncs is running a casino with better odds than the retailer with three channels and nightly syncs.

Cost 5: data where it should not be

Disconnected systems scatter customer data across platforms, spreadsheets and personal inboxes, and each copy is a compliance obligation. The Information Commissioner's Office guidance for organisations is clear that data protection responsibilities follow the data, not the system that holds it. A spreadsheet of customer orders on a buying manager's laptop is a data protection incident waiting for a laptop theft. Consolidation does not just improve operations, it shrinks the data surface you are responsible for.

The fix is consolidation, not more glue

The instinct when systems disconnect is to buy more middleware. Sometimes that is right, and the integration decision guide covers when connectors make sense. But for most retailers the better move is consolidation: one stock record, one order flow, one set of master data, and as few seams as the business can tolerate. The connected business model describes the target, and the gap between where you are and that target is the cost of disconnection. Price that gap honestly and the business case for change writes itself. Consolidation also means deciding what the ERP should own and what stays specialist, and the boundary for warehouse systems is a common fault line.

The final note from client work: nobody sets out to build a disconnected business. It happens by addition. The fix also happens by addition, but deliberately: connect the flows that cost the most first, measure the difference, and keep going until the spreadsheets have nothing left to do.