Optimization tools are powerful additions to the arsenal of business systems for any company. The tools leverage the transactional advantages afforded by the discipline and interconnectedness of energy trading and risk management (ETRM) systems. They build on rich data in a continuous feedback loop where new objectives are set, and new forecasts are created frequently. They allow for thinking ahead and changing the overall mindset of a business from being reactive and focused solely on the “here and now” to a proactive organization and plans. The tools allow an organization to plan for scenarios and perform “what-if” analyses.
So, now what?
The biggest hurdle to overcome in any journey is the first step. As Thomas Jefferson once famously said, “If you want something you have never had, you must be willing to do something you have never done.” The fear of the unknown and perceived complexity prevent us from embarking on new initiatives. Companies are no different.
This article addresses a few myths and hurdles that prevent companies from taking the first step toward moving up the maturity curve.
Myth 1: Optimization Systems Will Never Integrate with Any Other System
The first myth about implementing supply chain optimization systems is that they require a tremendous number of interfaces, many of which are too complex to figure out. The worry is that the optimization engine will require massive amounts of data and then force users to have to manually enter the results in their enterprise resource planning (ERP) system. Nothing could be further from the truth. Tools these days have a standard structure and connectors to interface with other systems. Software vendors want the tools to connect to increase the overall effectiveness of the tool.
Typically, optimization tools are fed on a nightly basis. The optimization tool takes in master and updated reference data, contracts, physical inventory and orders. In return, it provides back an optimized schedule that’s automatically fed directly into the ERP system to create orders. In addition, the interface allows for the creation of forecasted future demand orders within the ERP system, which forms the basis for an inventory rundown.
Myth 2: I’ll Need to Hire a Team of Ph.Ds. and Quants to Implement
In years past, companies required a handful of “quants” (data analysts) that knew both the theory and how to code proprietary models and integrate them within ERP systems. Today, with the push towards the democratization of apps and technology, those tools that were once out of reach are baked into sophisticated engines.
A Ph.D. isn’t required to implement today’s optimization tool. Those models are already incorporated. They are proprietary and were built by data scientists. The implementation team spends time refining the inputs and the configuration parameters of the models to leverage the best regression possible and create the best digital twin of the supply chain.
The most key members of the project team are those in the business that understand how the business should flow and can determine if the output datasets are directionally correct or not. These business experts combine with team members who appreciate the data flow and underlying models to ensure that the optimization tool both receives and provides all the data needed for forecasting.
Myth 3: If I Can Just Get My Data Clean Enough, Then I’ll Start
A common excuse to not start a supply chain optimization implementation tool is that the data isn’t clean enough (“garbage in, garbage out”). The truth is that business is always in motion. Therefore, you’ll never have a perfectly pristine system full of the cleanest data imaginable.
Optimization tools assist users in uncovering holes in the master reference data. When the model runs with an objective function to minimize cost, the model works to push the product out through each of the configured routes or channels. Therefore, if master reference data or contracts are set up incorrectly (or missing), they’ll immediately be presented in exception reports. It’s important to leverage the power of linear regression modeling and learn to interpret the results of the runs. Take stock in knowing that there are infeasibilities that need to be addressed. In doing so, the ETRM system will be pushed to a higher standard of cleanliness.
In summary, optimization tools will find your issues long before they manifest and cause huge headaches at a month-end or quarter-end close. You just need to know how to interpret the results.
Myth 4: It Will Be Better to Implement Once Our Business Model Slows Down
There’s never a suitable time to interrupt the business to implement critical software. Given that an optimization tool will require change management, the strategy is to start small with a subset of the overall business and then scale it up as the business allows. Leverage reputable consulting firms with deep experience in the industry and the tool. Provide resources to perform signoffs at critical junctions on the design and then be present for the comprehensive testing.
As the team proves out components of the business, continue to tackle more. In doing so, the business will appreciate the extra time put into adjusting processes and shifting the mentality to being forward-focused.
Patrick Long is a director in Opportune LLP’s Process & Technology practice.