How are data sent between different standalone systems within your organization?
Most business will need data routed in at least these following ways:
- If payment is received up front, cus
tomer and payment information will be sent to your payment gateway, which processes the payment and credits your account. If payment is not up front, an invoice will need to be generated and applied to the correct account.
- The order data will need to be routed to your warehouse (or 3PL partner). Employees on the warehouse floor will need to locate, pick, pack, and ship the products in the order.
- Stock levels will have to be adjusted accordingly. If stock levels get low enough, your purchasing department (or manufacturing site) will have to be alerted so that items can be replenished.
- If you use multiple carriers, the correct carrier (offering the best service at the best price) will have to be selected. Shipping information will need to be sent to that carrier.
- All of this information will need to be stored in case the customer returns the item/s or a recall is issued.
- Purchase data will need to be aggregated to create reports that your sales and marketing teams can use.
In essence, each of these is a potential transfer that your data need to make between different systems. This list thus represents the minimum data that need to be sent to different departments; depending on your products and business mode, there could well be more jumps than are listed here.
Why list out these data transfers between systems? Depending on your technology and processes, each transfer can be a potential problem for your business. If data transfer between two systems or two silos is slow, that transfer process can represent a bottleneck in your business—either slowing down order fulfillment or delaying business decisions you need to make. And if sending data requires human intervention, each represents another potential source of error.
Why Data Transfers Need Automated Data Flows
Businesses vary widely in the ways in which they transfer data. Some download and upload files from a central server (using FTP, for example) with each department working on files independently. Some use a batch process to extract, transfer, and load data between systems. Some use Electronic Data Interchange (EDI) systems. And some use APIs to allow their in-house applications to access data in the cloud.
Not all of these methods are equally as good as the others. Generally, the more automated the data transfer, the better. When done right, data from different systems are “tied together” so that working with that information is seamless across applications, departments, and locations. This maximizes the value created for customers while using the minimal amount of resources, ultimately leading to competitive success.
If your organization does not have the best automated data transfer available, you risk the following:
- Bottlenecks: Faster throughput means higher productivity. If data need to be routed to a certain process, but the data are held up in transfer, bottlenecks occur. These are especially likely during times of high order volume.
- Error-prone processes: The more human beings are involved in entering or routing data, the more likely it is that errors will occur. Auditing also becomes more difficult.
- Versioning problems: What happens when sales is working with one set of order numbers, purchasing with another, and the C-suite is receiving yet a third set of numbers? This is a recipe for chaos.
- Sudden failure: Systems with less automation are more likely to fail without warning. Some examples: An item can go Out of Stock, a recall could be missed, an inbound order can be lost or misplaced, or a massive billing error can occur. Worse, when certain processes require human intervention, the entire system can come to a halt if the people with the knowledge to perform a step are suddenly unable to do so. To be blunt: If the one person who knows how to get your order data to accounts receivable is suddenly hit by a bus, your cashflow comes to a sudden halt until you can train a replacement.
- Propagation of problems: In highly automated systems, anomalies are caught and reported as soon as they occur. When data jumps are not automated, a problem can go unreported for some time—which cascades and creates other problems.
These are some of the reasons the engineers here at Infoplus prefer to use APIs. Yes, APIs take time to develop. And their adoption is not yet ubiquitous. Still, APIs currently represent the best way to link differing standalone systems and automate data transfer between them. For example, suppose you run a business with an online store. An order comes in with payment information. Where does that data need to go?