Find Research
Our World View
Industry Perspectives
Research Program
Parallax View Magazine
> stay tuned
View our collections of research around key subject areas:
>
>
>
>
ERP
>
>
>
>
>
>
>
SRM
>
>
WMS
 
 
Article
Collaborative Demand Planning Technology

Everybody wants to collaborate. But is this actually done -- technically?


Full Article Below -
Untitled Document

More than a decade has passed since the advent of the Collaborative Planning Forecasting and Replenishment (CPFR) programs launched by Walmart and my colleague, Robert Bruce (who is the author of another article in this issue of the brief: “The State of Collaboration.”) Collaboration may be defined as an ideal state in which multiple parties engage in business processes and information sharing, working together to produce mutual goals—and results. The work of achieving collaboration between retailers and their suppliers has gone on for a decade. And the technologies used to achieve that have significantly evolved within the last decade to become very effective, affordable, and easy to adopt—technically, that is.

In this article, I would like to review some technical concepts associated with collaboration and also review a few solution providers who, today, do a great job in providing collaborative solutions for their customers. So firstly, let's talk about the underlying architecture and technology for collaborating. A historical perspective takes us back to the days of EDI and the ubiquitous 830 and 856 transactions used between retailers and their suppliers to create forecasts and confirm orders (and advanced ship notices). Basically, it is demand and supply planning via EDI.

With the advent of the Internet came the opportunity to upgrade that approach and provide more context to the forecast, moving from a hands-off transaction-oriented approach to one that was more collaborative, i.e., shared. (Thus the advent of collaboration was applied to supply chain.) It was as much the process as the technology, I might add.

Initially, XML use was examined. However, XML, fundamentally, still was a transaction-oriented approach. In other words, a file was created and sent from one system to another. It was not yet real-time data sharing. It was a send-and-receive messages approach to solving the problem. Then many companies started dusting off their spread-sheets and MRP files and making them available in web portals. A step forward. These methods—EDI, XML, file-sharing and web portals—are often what is used today when people talk about CPFR programs. Approaches have gotten pretty sophisticated in terms of their ability to provide standardized data, context, intelligence, workflow, and analytics to help trading partners refine information and processes. Figure 1 provides a historical context to data sharing, and shows how far we have come.

Now we have the cloud!1 This introduces yet another approach toward sharing information between trading partners. Data sharing from the cloud—whether transaction-based or single database single instance (where we are all looking at the same database) offers a range of possibilities.


Figure 1

As we think about collaboration between trading partners from the demand-management side, several key processes come to mind such as:

  • collaborative forecasting and replenishment
  • trade promotion management
  • joint supply planning, or S&OP
  • multi-party fulfillment

What has been interesting about the progress we've made in CPFR are the initiators, if you will, of the collaborative relationship. It used to be the power players (the anchors) who drove the relationship, for example, HP or Merck—manufacturer-to-supplier examples; or Walmart or Best Buy—retailer-to-supplier examples. However, we recently heard a great case study in which the supplier, Bayer, recruited retailers for collaboration. So it doesn't just have to be the OEM or the retailer who initiates a collaborative network with trading partners. That is an important concept that we think about work flows and data synchronization between partners; and as we think about joining or adopting networks, even multiple networks (often required in our complex supply chain world), where we have multiple trading partners of varying authority, and different types of technology to integrate to.

Cloud-based solutions share the same data; they don’t shuttle net changes between multi-party databases, a process which introduces data latency and errors. It is an ideal state, and one we see being adopted more as cloud becomes a more common approach. However, it just won’t be possible for all players at each stage of the forecasting process to be using the same database, all the time. We will still need methods and processes to harmonize the processes. Several challenges present themselves when we think about sharing demand data between trading partners in the cloud. For example:

  • The security of the information: What kind of information is viable to actually share between trading partners? Often, the enterprise with the end-customer data is very reluctant to share it with their downstream supply chain partners. However, providing at least some view of the customers, such as demographic information, lifestyle, and the channel(s) through which the customers prefer to shop clearly would provide supply chain partners with important insights.
  • What software and functions to use: What I mean by this is that even though one company may be the anchor—i.e., the catalyst for the use of a collaborative solution (say, the retailer), participating trading partners generally have some of their own software, as well. So the question becomes, what elements of my existing software solution will I continue to use? And what elements of the newly provided collaborative platform software will I use?
  • How and when to integrate the information between the trading partners: This is not a trivial concern, since today there are so many different types of methodologies that de facto represent a kind of integration. All the trading partners have their own enterprise forecasting systems, and also want to potentially share information. Each has their own business events which inform them about the state of their market, and impact how they interpret the forecast. This process all needs to be orchestrated between the partners to ensure that somehow they get everything synchronized.
  • Granularity of information to share: Many companies collaborate at an MPS level. In this, information can be updated daily. But in certain industries where information needs to be shared on an hourly basis, for example, build-to-order, real-time synchronization may be required.
  • Planning versus execution: Initially, when people think about collaboration, they think about the planning element. And although many people haven't done CPFR, those who have done it tackled what we perceive to be the easier problem, which is the planning. It's easier to orchestrate planning exercises between trading partners than it is to orchestrate execution, which tends to happen in real-time. However, truly meaningful information informs the system during the execution phase: How well is the promotion doing? What changes to orders and forecasts are occurring?

Who Can Help?

There are challenging problems to solve. So it behooves us to work with companies that have experience in dealing with those kinds of challenges. One such company is Logility, one of the most experienced CPFR providers. Their multi-tier value-chain collaboration software has been used by large and mid-sized companies, demonstrating that collaboration is not just for the giant power players.2

Software companies have been doing collaboration since the mid- to late-1990s, though. NeoGrid (formerly Agentrics, also an early pioneer in CPFR) uses a cloud solution. Their approach creates a “grid-like” integration between trading partners. Retailers such as Best Buy and manufacturers such as Merck, who collaborate with their suppliers, use NeoGrid. As does Bayer Healthcare, collaborating from supplier to customer, the way many suppliers would like to work.

Newer kids on the block, Vecco International (previously called Sockeye Solutions), use a cloud approach, as well. Their customer, HP, has thousands of suppliers in their collaborative portal. However, Vecco can provide these services for mid-market companies as well, who may have around fifty partners with whom they might want to collaborate. Of note is that the collaboration is not just ‘for the big boys’ and is accomplishable by a broad range of companies.

Supply chain performance has improved only for the few in the last few years, as suppliers struggle with complexity and poor data. It would behoove everyone to gain more collaboration with their customers. 

Conclusions

Businesses are getting more complex—not less. And one would think that this makes the challenge of collaborating more difficult. Surprisingly, collaboration within companies tends to be a bigger problem than collaborating with trading partners. Trading partner collaboration often has clear motivators; whereas internally, collaboration can involve a very complex set of interpersonal relationships, the politics surrounding competing priorities, and competition among teams and organizations. In addition, trading-partner collaboration has standardized information approaches. Whether the underpinnings of EDI help get us there or not may be debated, but the fact remains that inter-enterprise information somewhat conforms to recognizable standards. The cloud, of course, takes us to the next level. The challenge for companies will be to change the way they think about collaboration. But technically, they will have to adapt to the coexistence of cloud and on premise. Our future will have a variety of methods and we need to harmonize them—make them interoperable—in order to ensure error reduction and increased forecast accuracy.  

________________________________________________________

References:

Collaboration Library

The State of Collaboration


Collaboration Suite - Collaboration at the Crossroads

Retailer-Supplier Collaboration

________________________________________________________

1 For more on Cloud and the definitions and pricing models associated with them read: A Current View
and for pricing read SaaS Pricing.
-- Return to article text above
2 Examples are Lance (Lance brands and Cape Cod Potato Chips, to name a few), Caribou Coffee, Berry Plastics, Stanley Black & Decker and Verizon Wireless. -- Return to article text above


To view other articles from this issue of the brief, click here.





MarketViz powered.