Stop retyping orders: joining your shop, your stock and your accounts
20 July 2026 · 7 min read · Optimum IT Solutions
It starts small. One person copies the day's online orders into the accounts package each morning. Then a marketplace gets added, then a till, and now it is somebody's job.
Almost every growing retailer or wholesaler we meet has a version of this. The website is on one system, the accounts are on another, stock lives in a spreadsheet or in someone's head, and a human bridges the gaps by typing. It works, right up to the point where volume makes it stop working.
The interesting thing is that the retyping is rarely the worst cost. It is the most visible one, but the expensive damage is done by the errors and the delay that come with it.
What the manual bridge actually costs
There are four costs, and only the first is obvious.
- The hours. Someone is paid to move numbers from one screen to another, which produces nothing new.
- The mistakes. Manual entry is exactly the sort of repetitive work where attention drifts. A digit out on a price, an order entered twice, an order missed entirely, a customer address transposed.
- The overselling. If stock is only reconciled once a day, you can sell the same item on your site, on a marketplace and over the counter. Somebody then has to make an apologetic phone call, and on some marketplaces repeated cancellations affect your standing.
- The lag. Nobody can answer where are we on stock, or what did we actually sell this week, without waiting for the person who does the typing. Decisions get made on figures that are a day or a week old.
What joined up actually means
The phrase gets used loosely, so here is what we mean by it. For each important piece of information there is one place that is the truth, and everything else reads from it or writes to it.
Stock levels live in one place, and your website, your marketplaces and your till all reflect that one number rather than keeping their own private version. Product information is written once and pushed out to each channel in the format that channel wants. An order placed anywhere lands in one order list. Customers are one record, not four.
That is the whole idea. Everything else is detail about which system holds which truth and how the messages travel.
The joins worth doing first
You almost never do all of this at once, and you should not try. This is roughly the order that gives the most relief for the least disruption.
- Shop to accounts. Orders, refunds and fees flow into your accounting package automatically, with the right VAT treatment. This usually removes the single largest chunk of daily typing.
- Stock across channels. One stock pool that every sales channel draws down from, so you stop overselling.
- Product data out to channels. Write the title, description, attributes and images once, then map them to each marketplace's own format rather than rekeying for each.
- Shipping and couriers. Labels generated from the order, tracking numbers written back to the order and sent to the customer without anyone copying a reference.
- The till. If you have a physical shop, the counter sale should reduce the same stock number the website sees.
Sort your product codes out first
This is the part people skip and then regret. None of the joining above works unless each product has one code that means the same thing everywhere. If the website calls it BLU12, the accounts call it 12BLUE and the marketplace listing has no code at all, no integration can match them and you will spend the project reconciling by hand anyway.
Cleaning up the product list is dull and it is the highest value preparation you can do. It also tends to surface things worth knowing on its own, like duplicate lines, products that have not sold in two years, and prices that disagree between channels.
How the systems actually talk
There are broadly three approaches and each is right in different circumstances.
- A direct connector. Many shop platforms and accounts packages have an official app that links them. When one exists and fits how you work, use it. It is the cheapest option and somebody else maintains it.
- A middleware or integration platform. A service that sits between several systems, holds the stock and order truth, and talks to each channel. This is the usual answer once you have more than two or three channels.
- A bespoke integration. Built for you, when your process genuinely does not match what the products assume. More expensive up front, and the right call when the alternative is bending your business to fit software.
The edge cases that decide whether it works
The happy path is easy. What determines whether an integration actually removes work is how it handles the awkward cases, and these are worth asking about explicitly before anyone builds anything.
Returns and partial refunds. Cancellations. Orders paid partly by gift card or voucher. Marketplace fees and commissions, which need to land in the accounts as costs rather than vanishing. VAT on sales to other countries. Bundles and kits, where selling one item should reduce several stock lines. Backorders. And what happens when a channel is briefly unreachable, because it will be.
It is also worth deciding deliberately how much stock you expose per channel. Some businesses hold a small buffer back rather than publishing the last unit everywhere, precisely because no synchronisation is instant.
Start with one join and prove it
The failure mode for this kind of project is trying to connect everything at once, discovering the data is messier than anyone admitted, and stalling halfway with two systems half joined and the manual process still running alongside.
Pick the join that removes the most typing, do that one properly, run it in parallel with the manual process for a couple of weeks until you trust it, then stop doing it by hand and move to the next. Slower to plan, considerably faster to finish.
If you want to talk it through, tell us which two systems you currently retype between and roughly how many orders a week. That is usually enough for us to say honestly whether an off the shelf connector will do the job or whether it needs building.
Want a straight answer on this?
Tell us the job that is costing you the most time. We will look at it and tell you honestly what, if anything, is worth doing about it. The first conversation is free.
Get a quote
