< Back to Archive

OEMs and Suppliers Share Best Practices in Collaborative Forecasting

COLUMBIA CITY, Ind. (May 12, 2020)—We hosted a virtual roundtable to gain insight from both device companies and contract manufacturers about collaborative forecasting in today’s orthopedic space. Julie Sutton, Supply Chain Director for NN Life Sciences, led the discussion on ways to manage information, processes and behaviors to overcome obstacles and attain mutual success, particularly in the context of ambiguous timelines such as with the current global crisis.

Here in Part 2 of the recapped conversation, we focus on how companies are using forecasts and best practices for collaborating with manufacturing partners. Part 1 focused on companies’ responses to COVID-19.

The participants included:

  • John Kennedy IV, General Manager, Autocam Medical
  • Shannon Mills, Supply Chain Director, Zimmer Biomet
  • Michele Roskowiak, Demand Planning Manager, Wright Medical
  • Greg Stalcup, CEO, SITES Medical
  • Julie Sutton, Supply Chain Director, NN Life Sciences

Forecast Expectations and Utilization

Sutton: For device companies, how often do you give suppliers a forecast, and how do you expect them to use that?

Mills: We send a refresh of our forecast out to suppliers on a weekly basis. The weekly forecast that we give is essentially that netted supply plan that combines planned orders with purchased orders.

Our time horizon is 18 to 24 months out, and we’d like to see suppliers do two things with this data: first, put that into their planning system to compare their future month’s supply needs with their capacity; and second, look at any peaks or valleys that need to be smoothed out. We are not actively master scheduling yet, so the ability to see those spikes at our supplier or valleys that we need to fill in is critical, especially out in those future months where we have time to react.

If a supplier needs more capacity, I would expect that they would use this data to elevate that into their supply planning process and elevate that to us. Ultimately, we’d like them to inform us of a potential issue so we can vet that into our supply planning process. We would work together with our supplier to determine whether the forecasted issue will come to pass based on various factors, and then they can make their investment in either labor or capital as a result.

It is critically important when you have demand and supply gaps that the contract manufacturer and the OEM work together to decide the best way to close those gaps, because it’s not a one-sided answer.

Roskowiak: We focus on key partners – around 20% to 25% of our volume is with those customers – and give them forecasts about 12 to 18 months out. We also have S&OP meetings with our top five to 10 suppliers. The expectation is that our partners use these for their capacity planning. The larger companies are good about using it; the smaller ones, not so much. We want to make sure that they have the information if they need to add machines and build plans for their supply.

Sutton: For Contract Manufacturers: How often are you receiving a forecast from your customer, and how specifically are you using that forecast to navigate the short-, mid- and the long-term?

Kennedy: For Autocam Medical, I would say that varies depending on the customer. We have portal data and the ability to on-demand pull forecast data for 18-month visibility into the future. We use that to perform capacity planning and ensure that we have enough equipment and staffing in place to adjust to and accommodate what that data may show.

To that end, it’s only as good as the data that’s in there, and it certainly takes active management by an alignment with our customer. In some cases, we’re receiving a spreadsheet either on a weekly or monthly basis indicating demand planning by our customers. We’re able to utilize that data and do some planning around that, usually out about 12 months.

Having the forecast from our customers gives us a touchpoint and an opportunity to have a partnership, and if we see an issue in the forecast, we can bring that up on a more proactive basis. This allows for a more informed discussion and dialogue with our customers.

Stalcup: For SITES Medical, we receive forecasts from most if not all of our customers. However, the accuracy of those forecasts varies and depends on factors like whether the forecast is for a mature product vs. a new product launch, or whether the company is growing slow vs. fast.

We typically request and use forecast information monthly. We ask for it more often when there is a significant change in demand. We also apply a confidence factor based on what we know about our customer and the product. Even when we use the best available information and techniques, forecast inaccuracies can have negative impacts on our plans. Because of these forecast inaccuracies, we continually look for technologies and processes to improve our ability to execute in an environment where information changes are frequent.

We use short- and medium-term forecasts for two primary purposes; first for capacity planning, including the traditional elements of people, equipment, raw materials and space, and then for our financial planning. This allows us to allocate our financial resources where they’re needed to meet demand. For the long-term, we use forecasts as an input to our strategic plan.

Sutton: From my experience, getting those forecasts in a monthly cadence is important. At NN LifeSciences, for companies that do not supply us with a forecast, we are creating a statistical model and covering that same 18-month time period. That’s proved to be very helpful in two ways. One, it fills a gap for our clients that don’t send us a forecast. Two, it helps us model different scenarios, which helps us understand variability and change as well.

When our customers show a gap in demand and supply, there are times when we may choose to pull forward, times when we may need to add resources, and times when we flex our capacity. Making those decisions jointly and understanding the risks and benefits to each one of those has been beneficial and in a partnership spirit.

Best Practice Adoption

Sutton: What best practices can industry implement to better forecast?

Mills: I’d like to see progress with suppliers implementing internal capacity monitoring. From what we’ve seen, only about 30% are actively doing capacity planning, and that same amount seems to be utilizing the forecasts that are going out to plan and smooth their production schedule.

Roskowiak: Suppliers that have a dedicated person on our account who knows our business makes a huge difference. Also, related to COVID, we need to make sure that we’re not single sourced. If suppliers can have multiple site approvals to make something that will prevent a lot of risks if one location goes down.

Sutton: It’s a challenge; many of us would like to have multiple site approvals and few of us have the resources to validate dual sourcing, and that includes those at the OEMs. Even if the suppliers were interested in that, the ability to get resources at the OEMs has proven historically to be quite challenging.

Stalcup: For critical products, it is important to have redundant/multiple sites capable of manufacturing the product. While it can be hard to get multiple site approvals, we have two customers, in particular, that are proactive and work with us in that regard. They are doing creative things to help us get a second supplier approved. This has helped significantly for high growth unpredictable demand products. This is particularly helpful with new product lines where demand could be significantly different than the forecast. Legacy products don’t typically require the same level of redundant capacity because they typically have a known sales trajectory.

Kennedy: For us, we certainly do have cases where we’ve been able to source internally, and it’s allowed us to offer a lot more flexibility based on capacity planning. This does require additional validation resources on the front end, but we’ve been fairly successful, and it’s worked out generally well with our customers. We’ve been able to do some creative things with automation software to create our own version of Electronic Data Interchange (EDI).

As we look to continue to grow and develop best practices, we aim to have portals both for forecasting and actual order processing collaboratively, working with our customers to pass that data back and forth seamlessly and integrate that into our ERP.

Stalcup: We’ve been thinking about this problem for a long time. We’re getting ready to roll out some new capabilities that will help with variability in forecast, variability in planning, and many other variables each of us encounters. The solution that we are implementing involves both technology and process improvements.

As I mentioned, we’ve been thinking about this in a completely new and holistic way, one that lets us manufacture only the implants needed for an individual surgical plan instead of batch manufacturing for inventory. The new technology and processes allow us to build implants that are size-specific for the patient one at a time with very short lead times. We will be able to deliver a finished sterilized part in approximately three weeks, that is machining through sterile packaging, after receipt of the demand signal. We call this high velocity, single piece flow manufacturing.

Because of the short manufacturing lead time, the need for a detailed forecast is less critical. The OEMs can also move from an average of 12 months of implant inventory down to two to three months of inventory. This also frees up the OEM sales reps from the burden of inventory management and simplifies logistics for the hospital.

We believe that a high-velocity single piece flow manufacturing system is needed. SITES Medical and our startup sister company, Mach Medical, have developed technologies to enable this manufacturing for some of the toughest implants to make, such as porous femoral knees. We are rolling out this technology in knees first and will utilize the foundation we build for other implant categories in the future.  Some of the key technologies that enable this process for porous knee implant manufacturing include:

A high velocity precision casting process uses a proprietary 3D polymer printer to produce patterns for casting orthopedic implant components with tighter tolerances, in less time and with less expense than with traditional investment casting tools. This process is highly automated and scalable.

A proprietary CoCr Stabilization Process dimensionally stabilizes CoCr components which greatly reduces part movement, thereby eliminating the need for manual re-work and allowing process validation and automation, including zero set up time.

Our OsteoSync Ti proprietary technology offers improved implant-bone attachment without the risks or high costs of other materials and is robust enough to withstand automated handling in high velocity manufacturing.

Scalable transfer allows integrated systems to deliver rapid and consistent Device Master Records for manufacturing.

Single piece order manufacturing involves fully integrated, dedicated process flow with zero setup, common fixturing, automated transfer and digital inline inspection enabling single piece order manufacturing.

This system allows us to worry less about forecasts on an individual part number basis. Instead, we worry more about part family or macro forecasts to plan capacity for things like machines, team members, and materials we need to support demand.

Sutton: One best practice that industry can use in this space is the basic principle of S&OP or integrated business planning. That fundamental principle has been time-tested in terms of being able to control the exchange of information in your business quickly, and formalizing the rhythm of how we plan our business and involve the customers in planning our business. We have an extreme amount of data flowing through the organization that we have organized such that we can aggregate the data, dis-aggregate the data, and create insights for the business that we can use to engage our customers and suppliers.

Integrating that customer data into our S&OP process, creating business insights that we publish on a cadence all the way from our suppliers to our customers, gets everybody speaking the same language and making the same decisions about the same set of numbers and a single source of truth, a single set of data to make fact-based decisions. That’s been transformational in terms of keeping our organization aligned when there’s a significant amount of change, such as is occurring right now.

Discussion Takeaways

Manufacturing forecasts are routinely inaccurate due to several factors, including upstream market forces and internal human and software capacity. Device companies and contract manufacturers are adopting more sophisticated practices and tools to decrease the margin of error in forecasting. However, even as supply chain partners enhance their processes and better integrate their planning, the need for constant communication won’t be replaced.