Last year, McKinsey conducted a survey of a dozen global supply chain leaders across industry sectors to understand what changes could be expected in IT for supply chain planning over the next five years. One of the survey’s key findings was that 90 percent of respondents expected to overhaul their current planning IT.
In recent years, the COVID-19 pandemic and various destabilizing geopolitical events have made the need to strengthen global supply chains more urgent. Many companies have shifted from manual short-term solutions to longer-term solutions that incorporate advanced technologies, such as advanced prediction and optimization algorithms.
Such technologies are not limited to monolithic architectures and are often the result of advanced planning systems (APSs) that combine bespoke advanced-analytics (AA) models for increased functionality (see sidebar “What are advanced planning systems?”).
In a follow-up survey we conducted with 80 companies involved in digital-planning transformations that focused on the deployment of an APS, we found a wide range of ROI across companies, with the most successful companies achieving returns four times higher than the median.
This article focuses on the recipe that supply chain leaders can use to overcome challenges and achieve the highest impact during their transformations. This winning recipe is designed to support APS-centered digital-planning transformations and to make supply chains more resilient, effective, and efficient.
The five ingredients of a successful APS transformation
According to our survey respondents, more than 60 percent of supply chain–planning IT transformations take more time or money than expected or don’t achieve anticipated business outcomes. The winning recipe for a successful transformation requires integrating five ingredients to deliver at-scale impact (Exhibit 1). Much as a vehicle needs all of its elements, if one part of the system is missing, the vehicle won’t function as intended.
Linking these five ingredients will require fundamental shifts in the ways of working for organizations undergoing APS transformations. Making these shifts can help organizations achieve their business goals on time and under budget.
1. Processes and organization: How to defeat the attitude that ‘things have always been done this way’
A digital-planning transformation is not only an IT matter. It is first and foremost a redesign of planning processes, with a significantly higher degree of cross-functional integration. Simply laying new tech on top of old processes misses several opportunities for optimization, including offering value-added services to customers or driving profitability through the coordination of sales and operations.
Companies should understand “what good looks like” by getting a view of best-practice planning processes, which include demand, supply, logistics, and sales and operations planning—as well as the control tower, inventory management, scheduling, and collaboration with suppliers.
At the outset of the transformation, companies should invest the necessary amount of time into mapping every step of the process in conjunction with bottom-up impact analysis. This can help ensure that any newly implemented process supports the value drivers.
To address process and organization, companies should take the following actions (for an example, see sidebar “Example: Life sciences company”):
- Orchestrate processes and automate routine tasks, such as data updates, communication, and stakeholder input gathering.
- Design for critical executive decisions with readily available data and governance that facilitate alignment between commercial, logistics, production, and procurement.
- Enable comprehensive views of important risks and opportunities via root-cause analysis—for example, pegging orders to their raw materials.
- Embed process changes and improvements over several cycles, enabling incremental adoption of the new mindset and process.
- Create transparency for where new processes can be diverted by documenting process exceptions, such as when expediting orders creates supply chain inefficiencies.
- Monitor end-to-end supply chain performance and adherence with real-time dashboards to track system adoption and proper system decision making.
By the time the system is ready for transformation, companies should be ready to implement the designed process. Any divergence from the target design state should be monitored for both primary and secondary processes.
2. Data infrastructure and data management: How to overcome insufficient data readiness
Digital planning requires four system layers to integrate seamlessly: a system of innovation in which to use machine learning and other innovative solutions, APS as a core planning software (system of differentiation), a system of consolidation where multiple sources are harmonized and consolidated in the data lake, and a system of record where the company’s enterprise resource planning (ERP) systems typically have the majority of source data (Exhibit 2).
Automated data integration is challenging and should not be underestimated. What makes the difference in terms of impact is the timely availability and quality of the data with which the systems work. Even though most leaders are aware of this issue, we continue to see companies oversimplify the work, overestimate the capacity of their teams or their resources, and lack the necessary accountability for data preparation.
Treating data like a product can reduce the time and effort needed to implement new use cases by as much as 90 percent. On this point, APSs often have a well-defined data model, which is critical to scale the system. Today, companies are capable of preparing 70 percent of the necessary data tables in their own data lake well before the start of the actual transformation.
Typically, companies build a view or data tables that map to the APS vendor data model. Doing so can derisk future data pipeline issues. Data engineering inside the APS application is not easily accessible by users and can be considered a black box, and data changes upstream will likely cause data pipeline issues in the APS system. Companies therefore will need a hotline to an APS support organization to help them resolve the issues.
To address data integration and architecture, companies should take the following actions (for an example, see sidebar “Example: Large global agricultural company”):
- Integrate, contextualize, and harmonize necessary data inside the company’s cloud. Build well in advance a view or data tables that map to the APS vendor data model. ERP systems will be the main source of information; however, many companies are running with different ERP systems, and some critical data will likely not reside in the ERP system.
- Set up a system and data architecture to support real-time decision making and optimize load times between local data updates and the APS user system.
- Institute the management of master data. Planning systems are built on master data, and even though old processes might not surface problems, the newer process will almost certainly suffer.
- Review transactional-data accuracy and timeliness. Transactional-data updates should be real time (at the inventory level, for example) and synchronized. In addition, companies should review the accuracy of data related to details such as arrival times and product weights.
3. Planning technologies: How to avoid paying for something that isn’t used to its full potential
Companies often select an APS vendor based on its industry experience or the number of installations the vendor has done. In a second stage, companies create a checklist of planning functionalities, such as pegging, scenario analysis, and debottlenecking. However, we recommend getting a deeper understanding of the quality of these functionalities and reviewing nonfunctional criteria—including ease of use, quality of delivery services, future flexibility, and risk—before selecting an APS vendor.
Because so many APS transformations are considered IT projects, many companies adopt off-the-shelf APS solutions. Instead, companies should assess how to build a two-level architecture that combines APS technology with AA solutions, such as advanced prediction and optimization algorithms.
Regarding AA solutions, a higher level of customization—and, in some cases, bespoke solutions—is needed. Our experience shows that approximately 50 percent of the value from APS transformations comes from the use of customized AA models in combination with APS systems. This allows for maximum flexibility and predictive accuracy and helps optimization models to best fit planning trade-offs and constraints.
To address planning technologies, companies should take the following actions (for an example, see sidebar “Example: Global metals producer”):
- Undergo a thorough vendor selection based on a set of functional and nonfunctional selection criteria, including industry use cases, installs, and product deep-dive sessions.
- Select the right systems integration partner, particularly one that can handle the complexity and customization necessary to fit the company’s needs.
- Get an unbiased, high-level view of the envisioned result and map it against the off-the-shelf APS solution. Align on the customization and list requirements that can’t be fulfilled.
- Create the mechanism to achieve impact during solution blueprinting and put extra focus on AA solutions. Because these are often the main value drivers, review applicability and customize as needed.
- Synchronize the rollout of the APS and the relevant AA solutions by region or business unit, thus maximizing value creation along the journey, potentially self-funding the entire transformation, and boosting its overall net present value.
- As early as the build phase, test that the technology and its various features are functioning correctly. Company product owners should be the final gate to declaring a feature as complete.
- Take the time to complete system integration testing, user acceptance testing, and hyper-care (the period immediately following a system change that requires elevated levels of support). This will make your system more robust in the long term.
4. Capabilities, mindset, and behaviors: How to avoid the return of old habits
Best-practice processes are typically designed at the start of a project, but once a company reaches the implementation stage of an APS transformation, the target processes are adjusted and people tend to fall back to old ways of working.
This usually occurs when organizations underestimate the importance of establishing the right internal capabilities to drive an enhanced supply chain or when companies have not established processes for adapting mindsets and behaviors to new ways of working.
To develop talent with more-complex skill profiles, companies can create two new types of capabilities: one with a combination of functional, technical, and leadership competencies to drive performance and another with technical competencies to build, maintain, and develop core AA-model archetypes.
Next to talent development and capability building, there are three other core elements that embed change into the organization: role modeling management (such as the presence of supply chain and IT directors in key project meetings), embedding change in formal mechanisms (such as disabling old systems), and providing intense support to adopt new ways of working (such as setting up a network of superusers).
To address capabilities, mindset, and behaviors, companies should take the following actions (for an example, see sidebar “Example: Semiconductor producer”):
- Set up a company-wide communications strategy across multiple channels. As part of change management, strive to create understanding and model best practices.
- Be a role model of top management by being present in crucial project governance meetings, leading company communication, and regularly discussing capabilities in executive meetings.
- Develop tailored capability-building programs focused on the specific skills needed. Focus on best practices around planning as well as leadership skills and specific APS software skills.
- Blend new competencies, such as data science or data engineering, with training in supply chain management or support from “translators” who blend functional and digital knowledge.
- Develop a performance management structure that rewards top performers and takes corrective action when performance drops below expectations.
- Combine specific trainings based on learning material from the APS vendor with in-person classroom trainings for a set of superusers.
- Facilitate change management in which process improvements will be made over several cycles, enabling incremental adoption of the new mindset and process.
Finally, this type of transformation can have a twofold impact on performance: in addition to the overall improvement of systems and ways of working, the investment in people could generate a substantial return because of increased motivation. Shaping and building solutions that help employees make more effective business decisions in a more efficient way will enable employees to work smarter—rather than working harder to manually run routine tasks and failing to address complex economic optimization trade-offs in planning.
5. Integrated transformation management: How to avoid project delays and working in silos
Whether a tech-related transformation is for the supply chain or another business function, such transformations are notoriously difficult to get right.
For this reason, value assurance has emerged as a solution for ensuring on-time, on-budget, and on-vision delivery as well as supporting rapid value capture and long-term sustainable impact. Leaders can focus on value-led transformations by implementing guardianship around three distinct, equally important, and mutually reinforcing pillars of responsibility: design, delivery, and value.
- Design. Creating a blueprint for success is crucial to effectively managing ingredients of success, such as process and organization, data infrastructure and management, planning technologies, and integrated transformation.
- Delivery. Ensuring that system and organizational readiness ultimately stacks up to initial design blueprints, delivery also helps resolve tough design choices regarding customization versus standard system design, enabling rapid end-to-end implementation of the solution across the value chain.
- Value. Driving successful change management can help establish a value-focused transformation office to ensure cross-functional stakeholder alignment and prioritize actions.
With these points in mind, a necessary component of value assurance is building a multidisciplinary team. When it comes to managing the deployment of a wave itself—for example, planning the blueprint, build, testing, or hyper-care for a certain geography and business unit—leaders should set up a strong hands-on team, led by someone who can steer a diverse group of people.
Typically, an advanced planning transformation requires collaboration among at least three to four different companies—for instance, a planning software company, a systems integrator, an operating-model consultancy, and a middleware company. It is important that together these companies bring the necessary expertise on all different layers of a digital-planning system. This requires a diverse team set up with more than ten different roles, including industry experts, supply chain planners, solution architects, integration architects, solution configurators, data lake engineers, source system IT, middleware experts, and others.
Several factors make this level of collaboration challenging. To begin, it requires a change in mindsets among buyers and suppliers that may be used to more transactional or even adversarial relationships. And most collaborative efforts need intensive, cross-functional involvement from all sides—a marked change to the normal working methods at many companies.
To address integrated transformation management, companies should take the following actions (for an example, see sidebar “Example: A metal company”):
- Set up a project management and project steering committee headed by both IT and the business. Final decision makers, such as the COO and chief technological officer (CTO), can lead the steering committee, removing roadblocks and providing clear direction; project managers can drive top-quality content because of their deep supply chain and ERP expertise.
- Take the time to become one team and carve out time to have fun. Have an in-person kickoff to align on roles and responsibilities, deliverables of blueprinting, and ways of working.
- Establish a project management cadence that involves all parties. Key stakeholders of all companies and departments should be present for daily project check-ins, weekly progress reviews, and monthly steering committees.
- Create clearly defined roles and responsibilities. Summarize team roles and their descriptions on one page and the deliverables per role and project phase on another. Buy-in from senior leaders of each company involved can also help hold everyone accountable.
- Use one project management tool that captures the entire workload—including user requirements, user acceptance testing (UAT) or system-integration-testing cases, issues, and change requests—and projects along the implementation timeline. Track progress rigorously and reprioritize when needed.
- Tailor the meeting cadence to what is needed per phase of the project—for example, blueprinting meetings can vary week by week, while testing meetings can be held daily at predefined times and with a consistent agenda. That said, some phases, such as blueprinting and testing, benefit from face-to-face interaction, while others, such as building, are more remote.
Improving cross-functional engagement is a leadership issue. Organizations with the most successful collaboration programs often use a formal approach to managing cross-functional teams, with clearly defined roles and responsibilities. This is necessary because a formal approach not only helps ensure that the business does not see the transformation as an IT project but also helps the APS vendor feel accountable for the outcome in terms of its impact on performance and not merely its execution.
Old playbooks for transforming supply chain systems are no longer useful, so companies need to think APS transformations through across a broad range of areas. To succeed, it is vital that companies look beyond IT to the transformation process itself: the desired data pipeline, which APS system works best for them, what cultural changes the organization needs, and how collaboration across organizations will work now and in the future.