E-fulfillment solutions allow companies engaged in direct-to-customer operations to handle the challenges of fulfillment over the Internet
Two Christmases ago, Nancy Simpson-Banker could not find the Pokemon Yellow Hint Book, a Christmas present for her youngest son, Dean. No one seemed to have it. So she became excited over the Thanksgiving holiday when a well-known Internet toy retailer reported that the item usually shipped in two days. She placed the order and canceled her backorder at the local Toys ‘R’ Us. When canceling the Toys ‘R’ Us order, she was informed that they had just received the shipment, was she sure she wanted to cancel? She said she did. It was more convenient to have it delivered than to make the 20-minute drive to the local toy store.
Two days later she received the following e-mail from the Internet retailer: “Unfortunately, when your order was processed, we discovered that the item listed below was not available for shipment. This is a rare exception (to) our normal inventory policy, and we apologize for any inconvenience or disappointment this may have caused you.” The Pokemon book would not arrive by Christmas. She canceled the order. When she went to the local toy store, she found that the item was sold out. No Pokemon book for Dean this Christmas. Bad mother! She vowed never to buy from that company again. Following Christmas 2000, the Internet retailer reported demand much lower than expected, and admitted that without a capital infusion and drastic downsizing the firm was unlikely to survive through the spring. This formerly high-flying stock now trades for pennies.
Such stories of flawed order fulfillment and customer frustration are all too common. They are not limited to new Internet retailers selling to consumers, but include traditional catalog houses and direct-to-customer manufacturers. Brick and mortar companies that have just started selling goods over the Internet, as opposed to virtual merchants, usually believe that their existing order fulfillment processes and systems will enable e-business. They often find out they do not. Spending millions of dollars putting up pretty Web sites will not help companies succeed in e-business if they cannot execute on the back end with solid order-fulfillment capabilities. Laying out the massive amounts of capital required to build e-commerce operations and attract new customers is a bad investment if those customers refuse to come back.
To market, to market
Over a year ago, the ARC Advisory Group, based in Dedham, MA, put out a report stating that it believed that a new market was forming for e-fulfillment solution (EFS) software. This solution set allows companies engaged in direct-to-customer operations to handle the challenges of fulfillment over the Internet.
E-fulfillment solutions are designed to improve a company’s order fulfillment capabilities. Specifically, EFS will enable companies with Internet sales to improve their e-business perfect-order fulfillment standard dramatically. This measurement defines a perfect order delivery as an order delivered on time, in the quantities ordered, with the correct value-added services, with no unauthorized substitutions, and billed correctly.
Different companies will require very different e-fulfillment solutions. The breadth, depth, and specific functionality needed will depend upon whether the company engages in business-to-consumer (b-to-c) or business-to-business (b-to-b) transactions, the kind of supply chain it has, and the sales channels it employs. The complete EFS suite is composed of:
- online order management;
- online product configuration;
- constraint-based capable-to-promise engines;
- warehouse management;
- transportation management; and
- fulfillment process management.
| E-Fulfillment Solution Suppliers | Online Order Mgmt. | ATP/CTP | Online Config. | Business Process Mgmt. | WMS | TMS |
|---|---|---|---|---|---|---|
| Baan | 3 | 3 | 3 | 3 | ||
| irista | 3 | 3 | 3 | 3 | ||
| i2 Technologies | 3 | 3 | 3 | 3 | 3 | |
| Manhattan | 3 | 3 | 3 | |||
| McHugh | 3 | 3 | 3 | |||
| Optum | 3 | 3 | 3 | |||
| Provia | 3 | 3 | 3 | |||
| WebPlan | 3 | 3 | 3 | |||
| Yantra | 3 | 3 | 3 |
Not all companies will need all these applications; however, all will need some sort of fulfillment process management capability. Because of space constraints, the two solutions best understood by logistics personnel, warehouse management and transportation management, will not be discussed here.
Complex activity
E-fulfillment solutions (EFS) are needed to handle order fulfillment complexity. For example, a company selling only a few products directly to customers can use an order entry system with secure payment capabilities with no adverse effects on order fulfillment. This solution would not be considered an e-fulfillment application. A more complex situation would require an online order management system (OMS) for billing accuracy.
Setting up a central processing unit might preclude the use of high-end CD/DVD drives. One company that manufactures engineer-to-order process automation equipment told ARC that before implementing this solution, 15% of its committed orders were, in fact, not capable of being manufactured or would not work properly if manufactured.
A future application of online configurators will be in ensuring not only robust order promising, but profitable order promising. For example, if a company knows what its activity-based costs are for value-added services, the configurator can price those value-added services for customers. Some Internet sites now do this in relation to gift wrapping, but the concept can be extended to cover the more complex forms of value-added services required by b-to-b customers.
Can do
Most companies engaged in direct-to-customer fulfillment order promising base that process on seeing whether finished goods are in a warehouse. If they are, then those goods are available-to-promise (ATP). These items are frequently sold on a first-come first-served basis. In theory, ATP is easy. All it requires is good integration between order management systems and warehouse management systems with real-time, up-to-date inventory levels.
In practice, this integration can cost millions, because companies may have multiple ordering systems supporting multiple channels. In such situations, debiting large orders from inventory that is available-to-promise is more complex than it first appears. That is why leading e-merchants still send the message, “This item usually ships within X days” instead of informing customers whether the item is in stock and precisely when they can expect to receive it.
Make do
If ATP is deceptively complex, constraint-based capable-to-promise (CTP) is downright complicated. CTP applications seek to ensure reliable promising, while reducing operational costs, for make-to-order manufacturers with direct fulfillment models. The primary savings come from being able to meet customer commitments with less work-in-progress and a smaller finished goods inventory. CTP is one of the more expensive EFS applications, but the payback period is generally less than a year. In this environment, commodity goods are not immediately available. What can be promised is production capacity.
Most CTP engines provide real-time commitments to customers by looking for unused capacity in a production planning engine and checking to see if the components are available. The engine then looks at the order size to see if sufficient capacity exists to fill the order in the desired time frame. Typically, a more granular scheduling is run at the end of the day. The engine looks to see if orders can be combined to create longer production runs and fewer change-overs. This can free additional capacity that can be promised the next morning. In essence, this is the form that most supply chain CTP implementations take.
Manugistics, the current CTP market leader, offers one product for high-volume order promising and another version for highly configurable products. Its product suite has functionality that goes beyond that of many other suppliers. It has the ability to include component substitution logic in promising situations. This is very important for configurable product industries in which shortages can occur in key components. Manugistics also has the ability to reoptimize several times a day, creating more capacity that can be promised. It can examine production, distribution, and logistics constraints before making customer promises, and it uses management tools for tracking and adjusting the performance of the CTP engine.
CTP projects can range from fairly straightforward to highly complex, depending on the number of systems integrated with the CTP. The CTP model can include the constraints of one factory or multiple factories, and even distribution constraints. Each integration point increases the magnitude of the potential payback, but at the cost of a more painful implementation, slower performance, and an increased chance of project failure.
Delayed gratification
Users commonly report that the tools of leading supply chain vendors have terrific ROI once implemented. However, getting them implemented usually costs much more and takes significantly longer than expected. Leading suppliers have tried to respond to these criticisms by redesigning their products, forming tight alliances with leading enterprise application integration (EAI) suppliers, creating proprietary EAI tools, and using vertical industry templates.
The next advance in constraint-based promising will be the move from capable-to-promise to variable pricing based upon capacity constraints. Currently, if we order an airline ticket during the holidays, we pay more because of airline capacity constraints. The same logic could, and should, be applied to make-to-order manufacturers in industries with variable or seasonal demand.
In the CTP area, leading supply chain suppliers have been slow to capitalize on simulation. CTP engines base their promise on available capacity. The capacity that is available to be promised depends upon the business rules. For example, a manufacturer with guaranteed three-day delivery has smaller capacity availability than a manufacturer whose policy is 90% on-time delivery within a week. Simulation lets users examine the business rules that govern available capacity and determine when those rules should be relaxed and when they should become more stringent.
The right score
E-fulfillment process management (EFPM) applications coordinate the chain of Internet fulfillment actions that must occur across an extended supply chain composed of internal and trading partner activities. EFPM, also called supply chain process management and supply chain event management, mainly focuses on providing visibility into fulfillment problems. Visibility can be visibility into key performance indicators, also called performance metrics, or visibility into key events. These solutions can also contain alert resolution logic and cross-application workflow-enabled business rules.
Performance measurement is the first component of this solution. Performance measurements help keep a company focused on the major areas that drive business growth. These measurements also should be linked to such areas as employee evaluations, bonuses, and pay raises. Within fulfillment process management, these items can be displayed through a dashboard or portal. More technically advanced portals offer drill-down capabilities.
Similarly, within a company, measurement must be standard and uniformly applied. For example, a consumer packaged goods manufacturer may measure different brand managers on fill rates. Therefore, it is essential that the fill rate calculation be based on accurate and current data. The importance of the integration backbone of these solutions, and the data cleansing methodology employed, cannot be overstated.
Companies need the right measurements to be able to implement any of these solutions. For analyzing supply chain performance, you should start by examining the supply chain operations reference (SCOR) Model, published by the Supply-Chain Council. This “plan-source-make-deliver” model links measurements to each model component. While SCOR is great for setting a framework and providing the initial groundwork, it is not granular enough from a role or industry point of view. However, SCOR does provide a uniform set of measurements that different companies within an industry and across industries can use to benchmark their performance.
| TRIGGER | TYPE | EXAMPLES |
|---|---|---|
| Occurs | Event/object | Cancellation/decommit/hold |
| Counts | Event/object | By product/customer/operation |
| Impending occurrence | Event/object | Impending receipt/maintenance |
| Paired/events (duration) | Event/object | Order to shipment/on hold to off hold |
| Attribute Comparison | Event/object | Ship quantity vs. order quantity/price vs. quote |
| Compare to fixed threshold | Measurement/attribute | Yield/return levels/profit margin |
| Compare to another variable | Measurement/attribute | Revenue vs. plan forecast |
| Triggering mechanisms for the Vigilance alert program use an agent-based system to detect disruptions in the business process. | ||
Out of line
Businesses also need nonstandard key performance indicators (KPIs). The more role-specific and industry-specific the measure is, the better. Manhattan Associates has done a good job of developing a logistics portal with different KPIs that reflect different roles, such as, shipping manager, receiving manager, inventory control, warehouse manager, and logistics executive. In contrast, McHugh Software International is doing pioneering work defining what a logistics portal for the consumer packaged goods industry should look like. More robust dashboards offer the ability to slice and dice the data in a variety of ways, as well as the ability to explore deeper for more granular looks into the data.
| PERFORMANCE ATTRIBUTES | MEASUREMENT |
|---|---|
| Cycle time | Total source lead time |
| Cost | Material acquisition costs |
| Service/quality | Percent defective/ |
| Assets | Raw material days of supply |
| The Supply Chain Operations Reference (SCOR) model links appropriate measurements to each component to analyze comparative performance. | |
Fulfillment process management visibility also consists of visibility into key events. For example, has a shipment of key components left the supplier’s factory? To ensure the highest level of service for key customers, managers need to know the earliest point when a problem has occurred.
Management often does not know critical business process events until after the fact. Resolving an exception typically involves a series of communications among people in different departments and companies by phone, fax, e-mail, and face-to-face discussions. Even routine problems such as late orders can take several days to remedy. Alert functionality must detect an event and then notify all the right people. Typically, alert notification takes place via e-mail or pager. To close the loop, it must also track the response and resolution process, capturing the knowledge gained to avoid the problem or optimize its resolution in the future.
Ever alert
The main advantage to this approach is speed. Automating the detection/notification of critical fulfillment events shrinks response times and improves performance. Monitoring resolution activities ensures that effective responses are implemented quickly. A key feature is escalation. If a situation is not resolved within a certain period of time, you can alert a manager’s manager.
This is a relatively new area, at least at the supply chain level, with most suppliers lacking strong alerting capabilities. One of the players that does feature robust alerting is Vigilance, which has introduced an agent-based system for real-time event detection and Web-based collaborative resolution of business process disruptions. By focusing on issue tracking and resolution, and not merely generating alerts, the system can improve business processes across enterprises and help optimize the value chain.
Alerting is different from the metric dashboards described earlier. Those dashboards are more historically oriented, the data tends not to be real-time, and while a KPI may show falling performance, it is often not immediately clear, without further analysis, why the value of that measurement is falling.
Place your order
FPM can also be used to coordinate geographically distributed and trading partner workflows. For example, several warehouse management solution suppliers have what is called order parsing capabilities. Order parsing allows companies that own a network of their own warehouses, as well as access to third-party warehouses for certain products, to determine which distribution center should fulfill a particular order. This determination can be based on geographical proximity, the availability of product, or least-cost shipping choices. Ideally, automation of this process includes e-mail verification to customers, as well as automatic updates to business systems and e-commerce systems that the order has been shipped.
In b-to-b transactions, there are often complex pricing arrangements that can include discounts based on order size and market/advertising rebates. Divisions of a company may qualify for corporate discounts but require delivery to remote locations. An order may even be split into distinct order lines with different delivery destinations.
And there are other complexities. Online order management systems come in different flavors. In some cases, the system provided is an entire online order management system for customers with no OMS. In some cases, it is a front end program with real-time hooks to a back-end OMS. In other cases, it is an extranet application for a large, special customer or a dealer network.
OMS has not traditionally been well adapted to process manufacturing industries — sectors such as metals, chemicals, and oil and gas — but that is changing as some suppliers begin to introduce systems based on attributes. In process industries, it is not unusual to have thousands or even tens or hundreds of thousands of stock-keeping units (SKUs). This can overwhelm many customers. Now customers will be able to order by attribute; for example, they can order pork sausage that is 75% fat-free. Instead of having to know the specific SKU, the system will automatically translate the attributes into an SKU with the proper pricing.
| Applications | Average Selling Price Order Management Systems (Software and Services) |
|---|---|
| Online order management | $238,000 |
| Capable-to-promise | $968,000 |
| E-fulfillment process management | $326,000 |
| Warehouse management | $376,000 |
| Transportation | $334,000 |
Customers rule
Online configurators are the most expensive of all EFS applications, costing a million dollars or more. These software programs enable customers to order based on allowable choices. The choices create multi-tiered business rule models that govern how products are configured, offered, and sold to individual customers, customer types, or market segments.
Online configurators prevent customers from ordering make-to-order or engineer-to-order goods that cannot be made. Configurators have a logic structure that precludes customers from choosing that are incompatible product features or functions.
Fulfillment should be viewed holistically, as process that requires the cooperation of diverse departments within an organization and even the cooperation of key trading partners. With the opening of the online sales channel, this need becomes more pronounced.
| Online | Online configurator | Capable-to-promise | WMS | TMS |
| Business intelligence/OLAP | ||||
| Integration hooks and connectors | ||||
| Tool set to extend & maintain solution | ||||
EFS enables the completion of the perfect order. It’s the Pokemon Yellow Hint Book delivered, as promised, soon after Thanksgiving and hidden in the hall closet in its festive wrappings well before Christmas.
Steve Banker, Ph.D., is the director of supply chain solutions at ARC Advisory Group, which provides strategic planning and technology assessment services. Steve, a former professor, can be reached at [email protected].