Part 1 of this blog series articulated the acute need to bring supply chain planning and execution together so that enterprises can react quickly in an informed and confident fashion. The Boston Red Sox‘ September 2011 collapse was used as a poignant example of how even the best long-term planning can be rendered useless if there is no responsiveness during crunch time.
In general, if we know that our plans are inherently wrong to start with – because we can’t forecast and predict accurately – why do we still insist on religiously executing that plan? On the other hand, if you need to make a change, shouldn’t you be able to evaluate the holistic consequences of your decision, especially in these days of scarce credit and working capital?
Getting Ready Closer to the Crunch Time
Where supply chain planning (SCP) systems have customarily come up short is in trying to bring the timeline down closer to execution. Here they have failed because their data isn’t current enough to be effective. In environments that have complex multilevel sales structures and no real constraint of inventory availability, demand planning (DP) systems’ complex forecasting (i.e., management of the forecast) is perhaps more important than real-time “what if” planning and execution. But how many companies have the luxury of no inventory constraints these days?
While all companies need to plan, the real problem to solve is responding profitably to actual demand, not the projected forecast. Since we know that the forecast was inaccurate by default, and therefore the plan was inaccurate, responding profitably to real demand requires a constant rebalancing of the supply chain.
In other words, the merging (or better aligning) of planning and execution is needed. I think that the desire to better connect planning and execution comes from the recent increase in use of sales and operations planning (S&OP) applications. Planning and execution have long been handled by different groups within the organization: planning was most often done in enterprise resource planning (ERP) systems by analysts while execution took place in operations by departmental managers and their staff without much interconnection. Companies have been slow to move these departments out of their respective silos so that they can begin collaborating on these issues
An S&OP process, the place where demand and supply are reconciled with financial departments and engineering departments (in terms of introducing new product portfolios and phasing out the old ones), is a good start towards merging planning and execution and keeping everyone on the same page. A recent Kinaxis blog post discussed the need for faster and more responsive and interactive S&OP processes, although top executives can engage in long-term strategic business planning via so-called “executive S&OP” tools.
S&OP has always been important. Now with the heightened economic uncertainty and volatility in both demand and supply, it could mean the difference in a company’s ability to survive or thrive in the current global environment. Multi-echelon Inventory Optimization (MEIO) is critical to intelligently right-size inventory investments with a keen eye on customer service goals. To drive margin improvements, one must look beyond single location finished goods inventory.
Needless to say, this process has a whole different feedback process than what traditional SCP applications were designed for. On the other hand, some supply chain execution (SCE) applications have had certain planning aspects for some time - the ability to sense real-time data, occurrences, replenishment points, etc., and react accordingly.
It is the middle ground between these two software categories that is starting to be looked at more closely. As one example, most transportation management systems (TMS) offer interaction between transportation planning and execution (decision making). This is constant with companies’ need and ability to continuously optimize solutions based on real-time input.
I am not yet sure quite how cloud-based and/or composite (service-oriented) applications intend to overcome the issue of different data models and different analytics. Without a doubt, there are ‘natural’ hand-offs between applications where this may make sense, such as between, for example, demand planning tools and TMS applications. But we are a long way from an all-encompassing semantic data model that would make this endeavor highly scalable. Thus, unified supply chain platforms by individual vendors will be the way of the world for some time to come.
Some Good SCM Product Examples
To bring planning and execution closer requires all the capabilities of generating the original plan, but also the ability to monitor reality against the plan and to respond very quickly when a divergence is detected. Needless to say, this is equally true on the supply side.
Because the physical supply chain has limited ability to be as responsive as required, resolving issues during execution often requires compromise. To reach compromise across functions and trading partners requires the ability to evaluate and compare different courses of action across a number of operational and financial metrics to understand the impact in both the short-term and the long-term.
As its name suggests, Preactor, a combination of “predict” and “react,” has long been able to marry planning with pragmatic execution on the shop floor, with as little or as much optimization as necessary, based on the manufacturing environment. For its part, Logility has long been an unsung supply chain management (SCM) hero that is a master of both planning and execution within a single platform, with S&OP as its core. Along similar lines is Manhattan Associates’ platform play, although focusing more on distribution of finished goods, i.e., not necessarily on exploding multi-level bills of materials and operations.
RedPrairie’s Flowcasting application solves the problem of disparate planning and execution data sets and time horizons by using near real-time point of sale (POS) data in retail stores (and the data on what is exactly on shelves) to adjust the plan on a daily basis. Actually, the updates could be presented more frequently, but such rapid adjustments cause too much fluctuation and system “nervousness.”
It is the realization of enterprises that they need to “plan, monitor, and respond” continuously that has lately brought many customers to the Kinaxis RapidResponse offering (and to its S&OP edition much more recently). The combination of the speed of Kinaxis in-memory analytics and the single data model covering demand and supply, finance and operations, long-term and short-term horizons provides the breakthrough of agility for its customers.
SAP has recently realized these trends and came up with the SAP Response Management solution, in close partnership with ICON-SCM. In process- and distribution-intensive industries, Infor has been able to link tactical planning with execution with Infor Advanced Planner (AP) and Infor Advanced Scheduler (AS). Reflexis Systems and RedPrairie would be good examples of unified planning and executions in retail stores, while Click Software’s solutions do that in the realm of field service.
Companies can use execution in terms of both decision processes and the physical processes. In other words, rapid response solutions focus on taking an order and deciding when and how it can be satisfied. Moreover, Kinaxis focuses on industries with configurable products in which capable-to-promise (CTP) plays a much bigger role than does available-to-promise (ATP). CTP necessitates a rapid analysis of material and capacity availability and the balancing of competing objectives across multiple levels of the supply chain.
Once the execution-level decision has been made, the physical supply chain springs to life, including a warehouse management system (WMS) and TMS. While SAP, Logility, RedPrairie, Manhattan, and Infor customers can benefit from these vendors’ WMS and TMS products, most of Kinaxis and Progress Software customers outsource these capabilities and consider them to be ‘black box’ operations.
Getting a Bigger Picture
The aforementioned concepts can be applied to both the long term and the short term horizons, and go beyond supply chain to overall operations planning or true integrated business planning (IBP). Indeed, to enter a new market one must have facilities, personnel, and a marketing budget in place. To launch a new product one must have research and development (R&D), market analysis, and marketing, which departments must in turn be coordinated and synchronized with the supply chain.
This is where an aggregated “corporate control tower” (“situation room” or “control center”, if you will) cockpit application comes into play. Servigistics, Progress Software, and Kinaxis would be some vendors with such situation room-like solutions. The recent Kinaxis Blog post explains the vendor’s control tower offering.
Bundled with rapid response SCM solutions are in-line analytics, which use the wealth of real-time captured data in the normal course of operations to adjust execution on the fly based on what is happening. I think saying that SCM solutions are moving from historical to predictive (data mining) analytics is perhaps misleading because the purpose of business intelligence (BI) has always been to use historical data to predict the future. Why else would you bother? What I think is more plausible is the use of more timely (inline) data and transactions to more accurately reflect on what is happening right now. This makes predictions more accurate bearing in mind that some heavy-duty stochastic predictive analytics could still be used to detect potential risks in the supply chain, but not really for tactical planning and execution.
The “Response Management” software category is really all about evaluating multiple variables to present scenarios given a set of business conditions. As with Kinaxis, there are in-memory solvers for several components in the Logility Voyager Solutions suite – Manufacturing Planning, Supply Planning, Inventory Optimization, Transportation Optimization. This allows these vendors to solve complex problems with multiple variables, constraints, scenarios, etc. in a quick and efficient manner.
With the Logility Voyager Solutions suite, Logility has also the built in performance management backbone that allows constant monitoring of the customer’s plans and execution at each stage/step of the supply chain. When conditions change – short supply, for example, users can quickly re-plan and apply their business goals (e.g., least cost, highest margin, preferred customer, etc.). Likewise, when someone over forecasts, he/she will be notified that “actual demand is lower by XYZ% than forecasted demand” which gives the team the opportunity to “shape demand” in the form of promotion, sales incentives, etc.
Process Innovation Remains Crucial
Social tools and networks is where the real enhancements should come to facilitate execution in the near future. Thus far social media has missed a real opportunity to improve processes by being too focused on sales, marketing, and customer service automation. But the evolution of social media does provide a good way to describe the next phase.
Initially the focus was on the “social graph” (Who do I know?) which morphed into the “interest graph” (Who wants to know me and do they like me and my offerings?) to facilitate social media marketing. We need to move to the “responsibility (action) graph” (Who can help me resolve an issue?) in order to capture the value of social media in business. This action graph is related directly to ‘time to detect’ and ‘time to correct’ because without knowing who can help little action can be taken. Needless to say, taking a long time to identify who can help reduces the time that one has to act.
Knowledge is power in doing (executing) as well, and not only on the planning side (as mentioned in Part 1). During his Kinexions 2011 keynote speech, Kevin O’Marah, currently an independent pundit (formerly a top executive at AMR Research and Gartner), asserted that the 2000s have experienced a SCM drought in terms of innovation by technology vendors. To back this up, he mentioned the advent of materials requirements planning (MRP) in the 1960s, manufacturing resource planning (MRPII) in 1975, ERP in the 1990s, advanced planning and scheduling (APS) in 1995, and business to business (B2B) e-commerce in 2000.
Well, I might disagree slightly here. Perhaps there is a drought in terms of no new sexy & sizzling three letter acronyms (TLA) in the SCM space, although Progress Software did come up with the Responsive Process Management (RPM) suite (see TEC’s related article here).
The reality is that much of innovation takes place within enterprises at the process level. Loss of talent (or human capital if you will) is a huge loss in process intellectual property. The goal is to marry responsive decision-making SCM tools with human intelligence and experience. Those so-called supply chain “maestros” with a proper mix of skills (e.g., analytical, interpersonal, gaming, etc.) are often the key success factor for any company.
To that end, below is the list of some pragmatic and innovative actions that some planners have espoused during their talks at Boston APICS Chapter’s meetings that I’ve attended recently. Some specific process and policy changes that these companies have instituted in response to the recession and volatility challenges are as follows:
Contrary to the conventional wisdom of “battening the hatches” and weathering the storm, companies that are not faint of heart can view the recession and volatility as an opportunity to drive greater agility and integration into their supply chain planning and execution processes. To do so, they will need to maintain a strong focus on the critical issues and be willing to drive behavioral changes at all levels of the organization. Alternatively, supply chain staffers can always look into a career in lifeguarding (or even pitching for the moribund Red Sox)…
Dear readers, what are your comments and opinions? We would certainly be interested in your experiences with these software tools and your best practices in this economic volatility and uncertainty.
Great piece higlighting both the need for Response Management and the organizational impacts, principally the aspect of compressing the decision cycle time.
My only disagreement i to the reason why the merging of planign and execution is becoming so important. It is market volatility, principally demand volatility, but also supply volatiltiy, that is exposing the short-comings of a cadence-driven, monthly S&OP. Byt the time the S&OP plan is published, much of the input that was used to create the plan is already a month out of date, and the market continues to change during the next month when the plan is being executed. In many environments there is too much change in this period, which is driving the convergence of plannign and execution.
These two posts highlight some important issues and a very balanced overview of some of the solution providers. While current information is always preferred, few companies will be able to achieve this goal. Even if companies could have current information, forecast accuracy and cost of service levels will always be 800 lb gorillas in the room. With high rate of churn in process industry marketing churn, the originator of the forecast is rarely around to face the music. Most service levels are based on historical norms but the actual cost and profit impact is not well understood. This leads to stock overages, deeper discounts and higher carrying costs. When you are producing excess production, you are wasting production capacity and agility. In uncertain times, reduced agility can really hurt your ability to profitably react. I have seen several companies reduce service level from 95 to 90%, and they are not losing business, are more profitable and are able to react to market swings.
In terms of dealing with uncertainty, I have seen customer use multiple simultaneous models with work capital constraints. The base line forecast is often high service levels with good margins. They have other models for supplier viability risks, key customer risks, several growth levels, reduced production and increased costs. As the market moves, they can continuously evaluate the best model and improve their financial results.
I agree that there is a base level of uncertainty in any model that means one is never sure of the likelihood that the plan will be achieved.
But that is the very reason to develop Response Management capabilities. Planning is only ythe beginnign of the process. You have to monitor the plan continuously to determine the significant mismatches, and then re-align the supply chain quickly.
Doing so as part of a decision support tool is a lot cheaper, and quicker to implement, than trying to achieve the same agility and responsiveness in your physical supply chain. Every second you save by knowing sooner and responding faster adds to the time the physical supply chain has to respond.
We have spent the past 20 years focusing on supply chain planning. This was time well spent, but it is time to progress to the next stage of maturity of accepting that the plan is never 100% correct and will never be 100% correct. In fact, count yourself lucky if your forecast accuracy is above 70%, especialy for new products.
So what do you do now? As PJ writes, you can no longer afford the inventory buffers, and for that matter the capacity buffers. this si where monitoring for significant disconnects int he supply chain followed by rapid response - and profitable response - come into play.
Thanks for great insights and a healthy debate, Trevor and Rory. Please keep those good comments coming :-)