In the world of on-premise software, asking about order volumes made sense. Could an application handle your volume or not? But times have changed. Modern enterprise application has evolved. Behind any ‘transaction’, like an order, there are tens or hundreds of API calls and events firing in the background. Some directly related to the transaction, and others to support it. Let’s look at some examples.
Inventory position updates
Before a customer can buy from you, you need to show them what’s available. And to show customers what you either have on hand or Available to Promise (ATP) online, you need an aggregate view of all your inventory across all systems and locations. This means that an Order Management System (OMS) needs to serve as an enterprise inventory hub. One that can consume data from one or more ERP systems, and a stream of sales transactions from Point of Sale systems or other internal sales systems. Then, it must recalculate what you want to make available to customers based on your safety stock and allocation rules. For large enterprises, the number of inventory position updates can easily total tens or hundreds of millions per day. Far in excess of order volumes. What else happens more often than orders?
Stock availability checks across the buying journey
Just because a customer checks if an item is in stock doesn’t necessarily mean they will place an order, and the ratio of stock availability checks to orders can vary a lot. In fact, we’ve seen ranges from 2:1 to 700:1. And today, stock availability isn’t just shown on the Product Details Page, it’s often surfaced for multiple products on the Product Listing Page and used to drive more accurate search results as well. Once again, this volume is typically many times the total number of orders processed. And in the world of modern event driven architecture, it’s important to remember that an order isn’t just a single transaction, it consists of many events. What kind of events?
Events during the order lifecycle
Events triggered during the order lifecycle may include payment authorization, fraud checks, order sourcing, allocation to a location, consuming order updates from other systems, triggering customer notifications, the list goes on. In fact, many orders can consist of 150 events or more. This is especially true of a flexible OMS that lets you extend order workflows to orchestrate the entire order lifecycle across other backend systems and manual processes. So, when it comes to scale in an OMS, what are some questions to ask?
Scalability questions to ask an OMS vendor
If you’re evaluating OMS vendors, consider the following questions:
- Can you provide real-world examples of peak season volumes? Including not just orders, but API calls, inventory position updates, and stock availability check volumes?
Ask about the total volume over peak periods, plus peak volume per second. But don’t just ask about volumes for a single customer, ask about volumes across an environment, especially if the vendor offers multi-tenant and single tenant deployment. - What is the maximum number of inventory position updates a single customer sends to your order management in a 24 hour period? How often are they sending updates?
This will give you a sense of how quickly they can process high volume loads. Also, ask how they optimize inventory ingestion from systems that can only send a full data refresh rather than deltas. - What sort of API response time do you provide for live stock availability checks across multiple products on the Product Listing Page?
Showing product availability earlier in the buying journey can increase conversions. So ask about response times across multiple products as well. - Can you provide real-world examples of dynamic scaling?
And remember, it’s not just about raw volume but the percentage volume change as well. - What investments have you made in cloud scalability over the past 24 months?
Are they on the latest and greatest databases and services offered by their cloud provider? Ask what actions they have taken to ensure peak performance in the future.
Summary
When it comes to scalability in modern architecture, it’s no longer enough to ask about basic transaction volumes. So much happens behind the scenes in event driven architecture. So dig deeper. Ask about API calls, inventory position updates and stock availability checks as well. And make sure the vendor you choose is continuing to make investments so they can continue to scale and support your future growth.
For real-world examples of how Fluent Order Management can support your peak season volumes, contact us today.