Are you an IBM Order Management user challenged to find a way to control the rising costs of fulfillment? Are you getting ready to add another layer of variables to your already complicated fulfillment rules? Do you struggle with fulfillment network disruptions, available capacity or other changes to your fulfillment strategy during peak season? The good news is that there is a cloud-based solution that can help – IBM Sterling Fulfillment Optimizer with Watson. In this blog, I’ve put together answers to some of the most frequently asked questions so you can take advantage of this AI-powered solution to optimize fulfillment. But first, let’s take a quick look at how it works.
About Sterling Fulfillment Optimizer
Sterling Fulfillment Optimizer was designed to help retailers address rising customer expectations around fulfillment, while maximizing profits. If you’re dealing with complex orders and a complex fulfillment network, this can be particularly time consuming and difficult to do – yet this is exactly where Sterling Fulfillment Optimizer thrives. For every order, the solution simultaneously analyzes real-time data for these five costs:
- Shipping cost
- Load balancing cost
- Processing cost
- Stock-out avoidance cost
- Markdown avoidance cost
Sterling Fulfillment Optimizer works with IBM Order Management to consider each of these building blocks of total cost to serve (TCS) and balances them in relation to each other to arrive at the lowest overall fulfillment cost. Unlike traditional rules-based systems that are limited by programmed rules that must be set up to account for every factor, Sterling Fulfillment Optimizer uses big data analytics, AI and machine learning to look for trends and patterns to find the optimal recommendation, while constantly “learning” based on the data. [
Watch our webcast for more details on AI technologies.] The solution continuously captures sales and inventory information for every SKU and node (store and distribution center) combination available, figures out the potential risk for stockouts and markdowns, and optimizes fulfillment to meet your business objectives.
Answers to five common user questions
To help you jumpstart benefits from Sterling Fulfillment Optimizer here are answers to five of the top user questions.
- How should I configure Sterling Fulfillment Optimizer?
Sterling Fulfillment Optimizer is AI-powered to help you make better fulfillment decisions against multiple objectives faster and more easily. You don’t have to spend time creating and updating rules and priorities as required by traditional systems – just use sliders. The most advised configuration is to leave all the sliders to the right, the maximum of 100%, so that you optimize fulfillment considering all factors equally with every order [Figure 1]. The solution is always up to date with your latest information. As fulfillment network conditions change it naturally adapts and the numbers prevail, ensuring you always optimize for costs, while meeting your SLAs. For example, if a particular order has a higher shipping cost if shipped from further away, Sterling Fulfillment Optimizer knows to recommend the lower shipping cost location.
Figure 1. Considering all factors equally ensures Sterling Fulfillment Optimizer always optimizes for cost.
However, because fulfillment networks and sell-through patterns are always changing, at another time the results may be different. For instance, the lowest shipping cost location for this order may have had high sales and low inventory, creating a risk of being out of stock for walk-in sales if that unit were pulled for another order. Said another way, Sterling Fulfillment Optimizer dynamically identifies opportunities to reinvest some of the overall shipping cost savings to deliver additional margin via walk-in sales that may have otherwise been lost. The solution automatically calculates the TCS and optimizes for the overall enterprise value with each order in real time.
- Are there times when I should change the settings?
In general, think of the sliders as overrides rather than something to tinker with frequently. However, at certain times, you may want to prioritize one or more factors over the total cost and change the weighting. For example, during a holiday promotion if you’re guaranteeing shipping by a certain date, you may choose to put special constraints on Sterling Fulfillment Optimizer to make sure customers get orders faster, even though this increases shipping costs. This configuration may provide a sub-optimal TCS, but it enforces your specific business objective – meeting shipping promises [Figure 2].
Figure 2. Set sliders to reflect a specific business objective, for example to meet shipping promises.
- How long does it take for Sterling Fulfillment Optimizer to start delivering benefits?
Sterling Fulfillment Optimizer allows users to set up different configurations of the models based on their business objectives. When a user chooses to activate a specific profile, the change takes place in near real-time. As a result, most benefits occur immediately. Due to the predictive nature of the stockout avoidance and markdown avoidance models, benefits from those considerations may take several weeks as the solution must account for the possibility of future sales during a stockout or markdown period and be able to track reality against what it predicted would have happened.
- How is Sterling Fulfillment Optimizer better than rules-based processing?
Sterling Fulfillment Optimizer enables optimization at scale because it considers a broader view of the overall fulfillment network [Figure 3], as opposed to traditional systems that only consider individual rules and their artificial constraints. In many rules-based systems, the true optimal options may have been eliminated due to these constraints, and thus are blind to what could have been a better fulfillment decision. They also may need to go through several layers of rules, without considering something like markdown avoidance costs, and ultimately select a node because it is the only option available at that point. Additionally, most rules-based systems live only in the moment and are unable to consider the longer-term view to benefit. Sterling Fulfillment Optimizer’s predictive models look at benefits for “today” as well as benefits to having the network better positioned for “tomorrow”. The solution considers all fulfillment costs together and in parallel, without the need to setup and maintain complicated rules. Because it is free from incoming assumptions as to which path to take, it delivers immediate and long-term, lasting benefits.
Figure 3. Consider all fulfillment costs together and in parallel with a view of the overall fulfillment network.
- What can I expect longer term from the solution?
There are two aspects to this answer. First, as an AI-powered solution, Sterling Fulfillment Optimizer learns over time with more and better data. As the predictive components continue to improve in their ability to identify sell-through patterns and probability of sale based on your data and historic seasonal considerations, the solution makes better and better fulfillment decisions, delivering sustained, ongoing benefit. This means that the sooner you get started, the further ahead you’ll be in optimizing delivery experiences for customers and the further ahead you’ll be compared to competitors.
Second, with respect to future solution capabilities, additional predictive models are on the roadmap, including one targeting node performance, as well as other new features focused on the business user. It’s important to remember that because Sterling Fulfillment Optimizer is a cloud-based solution, clients typically can take advantage of new features as they become available.
How to get started
Learn how IBM client REI is using Sterling Fulfillment Optimizer to optimize for various business goals throughout the year.
To learn more about the solution, visit our product page or online knowledge center. More technical documentation and details are available here.
If you have more questions, we encourage you to post them in the forum and we’ll address them as quickly as possible.
#SupplyChain#OrderManagementandFulfillment