Unilever expert cites financial flexibility as sales & operations planning aim

Effective sales & operations planning (S&OP) not only addresses supply and demand concerns, but also encompasses a more flexible approach to financial resources, according to an S&OP expert with Unilever. Martin Jarvis is VP of global S&OP for the London-based manufacturer of consumer goods, ranging from home and personal care products to foods.

By Staff June 1, 2007

Effective sales & operations planning (S&OP) not only addresses supply and demand concerns, but also encompasses a more flexible approach to financial resources, according to an S&OP expert with Unilever .

Martin Jarvis is VP of global S&OP for the London-based manufacturer of consumer goods, ranging from home and personal care products to foods. Manufacturing Business Technology recently had the opportunity to ask Jarvis a few questions about S&OP best practices, supporting IT systems, and end goals.

MBT : What are some of your best practices for S&OP?

One of the things we put on the top of the list is that the process should be led by the CEO or head of the unit. It also should be fully cross-functional. Of course, there needs to be clear agreement on the forecast, but where we’ve seen some good work lately is in assessing the operational forecast alongside the financial numbers. We’ve moved finance [managers] quite a bit over the last couple of years from having a purely bookkeeping view of the world to more of a rolling view of resource allocation.

MBT : So you are applying financial resources more dynaimcally?

Yes. We have started to get rid of fixed budgets and formal allocations, and are planning on more of a rolling basis, with a process in place by which someone can in effect bid for resources for a category or brand. If proof of a good track record or opportunity can be shown, then it is more likely to attract investment.

MBT : Just how flexible should the process be?

We agree on a set of principles the process has to conform to, but allow local operating units quite a bit of freedom in how they choose to run it. For instance, we don’t have a precise date on which you must have your S&OP meeting. What we do have are clear reporting guidelines and a process that includes when the forecast numbers have to be submitted, so most of the meetings line up at roughly the same time of the month. In some cases, however, we run S&OP weekly, such as for margarine in the U.K.

MBT : Do you spend much time debating the forecast during meetings?

Pretty much after the first week of the month, the forecast is agreed upon, and everything after is about the consequences of those numbers. In the past, we might have had some debate over what the forecast was quite late in the process, but now, it’s agreed upon quite early and we spend most of the rest of the time dealing with gaps and exceptions. You should have a certain level of detail in place beforehand so you can accomplish the business you need to in a couple of hours.

MBT : What are some of your IT priorities regarding S&OP?

We are putting time and effort into understanding data sources available, and how to turn them into information. A balancing factor is to work toward completeness of the data set, rather than extremely high accuracy in all areas. That said, we find problems can arise if you overdesign the S&OP process, and the process needs data that’s not readily available; or you do the opposite, and ignore data that you have.

MBT : What sort of “what-if” analysis tolls do you use?

Most of what we do still ends up getting done in spreadsheets. The ERP systems can churn out data, but they aren’t good at what-ifs. The advanced planning tools we’ve been using aren’t particularly good at them either. We are beginning to look at other means of decision support, such as piloting software from the likes of Interlace Systems to get data in the right place, and be able to manipulate it to make decisions.

MBT: What’s more important: having the right master data, or the right decision support?

The answer almost depends on which day you ask me, because both are important. However, as long as you have a sense of where your master data may be wrong or incomplete, you can probably get better and quicker business value on the decision-support end.