Introduction to Automated Order Management for E-Commerce Startups

Fabian Wesner
Fabian Wesner CTO Spryker
16. September 2014 in

Technology

Automated order management becomes a major issue when an e-commerce startup scales up. Since existing solutions often don’t meet the requirements in terms of scalability, customization, and transparency, we implemented our own solution to fulfill those needs. Prior to providing information on our solution, this post will firstly point out existing problems regarding order management in a startup environment.

Problem to Solve

When an e-commerce shop becomes successful you sooner or later need to deal with automated order management. Popular shop systems like Magento, Shopware and Oxid don’t provide a lot here, therefore it’s up to you to find a solution that scales.

Usually every order lifecycle consists at least of the dimensions payment, logistics, and fulfillment. Each dimension has different characteristics: for instance, there are pre- and post-payment methods, or you can ship from your own warehouse or use drop-ship services. On top of these default dimensions there can be many more like multichannel orders, transactional mailings, fraud prevention, integration with CRM systems, etc. In practice, every e-commerce company has its own, very specific order workflow which needs to be modeled.

During the first month of a startup it is sufficient to focus on the happy case, which usually is as simple as this: the customer buys, pays, and then the order is shipped. Depending on the business model this simple process should work for up to 90% of all orders. But when it comes to thousands of orders per day, you need to solve the 10% of unhappy cases like returns, cancelations, frauds, over- and underpayment, shortfalls, and others. This can result in heavy workflows which are hard to grasp. If you don’t have a rock solid solution, there’s a risk of operational chaos and a loss of agility. Even if your business runs smooth in Germany, you may get into trouble with specific requirements in other countries, e.g. in Brazil customers are used to pay with two credit cards (due to low credit limits) or via Boleto, while in Russia most of the customers prefer ‘cash on delivery’. Your order processing solution should be able to support any kind of workflow without expensive modifications.

Existing Approaches to Automated Order Management

Most of the startups begin their e-commerce business with manual order processing. Very often, there is no explicit workflow. Whenever something unexpected happens (like a return debit note) the operations team has to deal with it individually. This is the best way to start because it allows you to deeply understand your order workflows. Later, when the business scales up, limits will appear. For instance, if every order needs 15 minutes of manual processing, then with 100 orders per day you need 25 hours of processing time every day, which in turn means you need at least three dedicated employees in the operations team.

The common approach to scale up is to add automated hardcoded subprocesses. This very often results in hundreds of ‘if-then-else’ statements, whereas each of them is based on a single real life experience from the operations team, e.g. ‘if the customer did not pay for three days then we need to send a reminder’. In comparison to a pure manual approach, this approach allows the processing of a higher amount of orders, but at the expense of maintainability and traceability. In the end, there’s a high chance to end up with tons of spaghetti code.

Very often this is the time to introduce a professional solution for order management. On the one hand, there are big ERP systems which can do the job. These systems are created with multichannel enterprise companies in mind. For a startup, the integration of an ERP-system is a costly task and there is a risk of becoming dependent on expensive experts. On the other hand there are specialized SaaS-solutions that provide complete solutions for automated order management. Although some of them look promising, we strongly believe a startup must master its business processes and should not outsource this core competence.

From our point of view, none of these approaches really solves the issue of highly scalable, reliable, customizable, traceable, and measurable automated order management for high-performance startups. That’s why we implemented our own solution as a part of our shop system Yves & Zed which will be presented in a future blog post.

Still got questions?
Ask the author for further information.