Executive S&OP and operational S&OP- Should the data be the same?

addtoany linkedin

Here is part three of an interview I had with P.J. Jakovljevic from Technology Evaluations Centers (TEC) on sales and operations planning (S&OP). If you missed them, check out part one and part two. As mentioned yesterday, the entire Q&A along with PJ’s introduction and commentary on Kinaxis can be found here (free registration required).

PJ: Can the S&OP process be carried out without technology? Does this relate to the aforementioned S&OP maturity model?

TM: This is the perennial debate. Even Microsoft Office Excel is technology, as are the underlying ERP systems from which much of the data to carry out the S&OP process is extracted. Phone or conference calls and e-mail, both of which allow people to communicate more effectively as a group, are also technology. I find the more interesting question to be: “How can we change the process to better accomplish our goals using technology as an enabler?” The people who ask this question are the ones who achieve greatest process maturity. As I have already stated, if the S&OP process is designed without the available technologies in mind, then it will inevitably be less mature. Much of what we read as S&OP best practices was designed 30 years ago, before the advent of Excel, the PC, or the Internet—the times they are a-changin’.

PJ: Is it possible to have an effective S&OP process that only looks at the aggregate or “volume” level? How important is it to consider the operational and tactical feasibility of the S&OP plan?

TM: It depends, but I would say that the operational and tactical feasibility of the plan needs to be considered, and this is usually carried out at the so-called product mix level. This is especially true in markets that experience rapid introductions of new products, such as in high-tech/electronics, or when a new product introduction (NPI) takes a long time and has high uncertainty, such as in the pharmaceutical industry. This point is also a good illustration of the point I made above about the planning continuum. The classical definition of the S&OP time horizon is 6−18 months. What happens when you have a nine months’ or less product life cycle? What happens when a large part of the revenue for a product family will come from new products being introduced over the next nine months that require unique parts, due to the introduction of new technologies, and that are in short supply with 14−18 week lead times? Running S&OP for this product family at the ‘volume’ level only is nuts.

PJ: How do you differentiate between executive S&OP and operational S&OP, and what related solutions do you provide for each?

TM: I don’t. The differentiation is spurious. Executives may want to view data at a more aggregate level, but why would they want to view different data? Why wouldn’t executives want to drill into the details of something that intrigues them? Why wouldn’t they want to test the consequences of making a change? When I meet prospects and customers, I often hear of the Chambers Report, the Lazaridis Report, or whatever other CEO/COO report. These are reports that are inspected on a daily basis. But senior executives not only inspect them, but also want to get quick answers to the financial and operations consequences of changing assumptions or values. Essentially, these executives want to perform ‘what-if’ analysis in real time and to know that the values reported are based upon feasible plans. Feasibility cannot be evaluated at the aggregate or volume level. Perhaps feasibility can be tested at the volume level in slow moving industries with relatively few NPIs and fairly stable demand. But in high-tech/electronics, feasibility can only be tested at a fairly granular level. With key component lead times of 14−18 weeks, product life cycles of six months or less, and a forecast accuracy of greater than 60 percent difficult to achieve, S&OP has to be performed at both a lower level of granularity and over a short horizon. Of course, executives will want to view results and change data at the more aggregate level, but why does this require a separate data model and separate analytics? This is merely a matter of presentation of the information. Kinaxis RapidResponse is a single solution that covers multiple time planning horizons and supports planning as a continuum, not as a set of unique and divorced processes supported by islands of information. Scenarios can be created in a fraction of a second, which are then used to make changes at both the ‘volume’ and ‘mix’ levels. RapidResponse provides alerting capability for early detection of key metrics that are diverging from the planned values. For example, the marketing folks may make a change to the assumption of market share for a product family over the next 6−12 months. Based upon historical or projected mix ratios, the change in market share may lead to a shortage of capacity for one item and excess and obsolete commodities for another item. Clearly, several people need to be alerted that this change to the market share assumptions has been made because of issues that come up at the mix level. As described, RapidResponse alerts users to not only events, but also the consequences of these events. More importantly, RapidResponse supports the notion of responsibilities, which are used to identify whom in the distributed supply chain should be notified of the consequences and pulled into a new scenario used to propose and evaluate different alternatives for resolving, or at least reducing, the impact in minutes if not seconds. Scenarios can be compared side-by-side in RapidResponse to understand their impact of corporate targets for both financial and operational metrics, such as revenue, margin, inventory levels, customer service, etc. Stay tuned for the finale tomorrow!

Leave a Reply