Over the past weeks we have been speaking with a large number of Fortune 500 companies to provide support and guidance on how to respond to and rebound from the disruptions caused by the global COVID-19 pandemic.

At many of these companies, CXOs have asked their teams to accelerate supply chain related digital transformation initiatives. Why would they ask for that? The COVID-19 crisis has exposed significant weaknesses in how organizations run their value chains today.

The biggest challenge has been lack of supply chain visibility. Many companies struggle to have one comprehensive view on all of their capacities, constraints, inventory and alternatives across the network. This has prevented most organizations from responding to sudden disruptions.

Secondly the majority of all large organizations are still very focused on what they sell to their customers, rather than focusing on what is driving end-consumer demand. For instance, large consumer product companies forecast their future demand based on historical shipments into the retail channel (sell-in). This is obviously not giving great results when the underlying demand changes dramatically. Instead companies would benefit from incorporating drivers of sell-out/consumption and adopting advanced analytics/machine-learning based techniques for a more responsive forecasting process.

One of the key questions that comes to mind: why have all these companies not been able to prepare for surprises way earlier? The most common answer to that question is “data”. In Gartner’s “Future of Supply Chain, Reshaping the Profession” report, one of the key roadblocks to embarking on a digital transformation  is the access to reliable and good  data. When we interact with large organizations, we receive similar feedback: “we do not have the data to take a next step in our digitization journey”, “we would love to move to driver-based forecasting but we do not have the data”, “we have been working on supply chain visibility for the past 3 years but the master data is in different systems and often not accurate”.

While we acknowledge that master data is often inaccurate, or might not exist, there is immense value in taking a fresh approach to solving the problem. First and foremost, nearly all organizations are data rich, but lack insights. This has been shown time and again in most of the deployments that the o9 team has been part of.

In this blog we would like to provide a few practical recommendations on common data challenges and how to solve these more effectively. Most organizations have “data cleansing” projects, assigning a significant workforce to start cleansing data. But where to start? Are there alternatives to how it has been done traditionally? How do you ensure that this is not a one-time exercise but a continuous one, as it should be?

Here are some practical approaches on how to do all this more intelligently, and how the o9 platform, with the underlying Enterprise Knowledge Graph (EKG), allows for it:

  • Multiple ERPs, or different systems naming the same entity differently
    For example, a product is named differently in different systems (Sku12, SKU12, SKU-12 etc.). o9 is able to create a planning model and map the transactional data from different systems to the single model. At a large electronics customer, we wrote intelligent algorithms to detect similar entities across systems and map them back to one thing in the EKG model.
  • System of record for forward looking master data
    Master data needs to be maintained for placeholder products when planning for them ahead of when they become fully realized products. In the o9 EKG,  we can create maps between placeholders and fully realized products.  This then allows forecasts and actuals to be reconciled over planning cycles.
  • Affinity graphs/relations inferred via analytics
    Often demand planning products have cannibalization/halo impacts on other products. When demand for one product goes up, demand for another product  could  go down or up. In most planning systems today, users have to maintain that relation so planners can enter/adjust the forecasts. With o9, such relations can be  inferred by analytics instead of having to be maintained by planners.
  • Collaboration over master data
    In some situations (for example outsourced manufacturing), master data (capacities, routings etc) may be coming from other participants in the supply chain, and it is important to review / accept these to develop meaningful joint plans. The o9 platform, with its high fidelity representation of the value chain, makes it possible for different participants to work together to come to agreement on master data before it influences planning.
  • Connect transactional data with smart analytics to update master data
    Example: supplier lead-times. In the master data the lead-time from a supplier to the Factory is 6 weeks. However, based on transactional data that lead-time seems to be 7 weeks. By applying smart analytics on transactional data, in the o9 platform that lead-time deviation results in an alert to 1) update the planning policies, in this case the lead-time, 2) send a message to the buyer to talk to the supplier and 3) update master data and inventory policies for future periods.
  • Connect real-time data, with smart analytics to update master data
    Organizations are increasingly investing in IoT capabilities that can provide real-time information from the supply chain network. Think about a sensor at a factory line or from an inventory location. o9’s cloud-first architecture allows for robust analytics and dataflow pipelines to be built with the relevant data warehousing and management services from the cloud providers. Pipelines to ingest real-time data as they become available is key to ensuring master data is always kept current.
  • Connect with tribal knowledge to establish a planning system of record
    In some cases, there may not be an evident system to maintain data needed for planning. For example, a retailer may need to maintain inbound capacity & QA/QC process lead-times for goods being procured, potentially wildly different based on the item. This information today is not captured in the master data but is often maintained in Excel spreadsheets or stand-alone systems (such as hand scanner system output). o9’s EKG, allowing for a true representation of the supply chain network allows for attributes and associated parameters to be built up over time, as necessary. This ability was actively used by one of the largest retailers in the world who connected this data to the o9 platform and was able to create a planning system of record close to where the planning decisions were being made.

In conclusion, master data management and maintenance is important to get right before embarking on a journey on digital transformation. It is essential to have a platform that has the richness of representation as well as the technical capabilities to allow this to be done effectively. The o9 platform, with the underlying Enterprise Knowledge Graph that enables a digital representation of the entire enterprise. Its interconnections with customers and suppliers as a digital twin of the supply chain, can support organizations looking to embark on these initiatives.

See below for an example of the digital twin of a supply chain powered by the o9 EKG, representing various data sources to provide E2E visibility

Are you intrigued? Let’s get in touch!

Stephan de Barse

About Stephan de Barse

EVP at o9 Solutions, driving digital transformation with some of the leading Fortune-500 companies. Intrinsically motivated to solve some of the most difficult challenges with technology with the aim to deliver business value.