The route to no-touch planning: Taking the human error out of
supply-chain planning
Slow,
manual supply-chain planning processes can be a thing of the past, with
machines taking on repetitive tasks that aren’t a good use of human capacity.
Supply-chain planning keeps
getting harder and more time-consuming, with the consumer goods sector as one
of the most extreme examples. The causes are familiar: Online retailing’s
endless shelf encourages consumers to be ever more demanding, yielding product portfolios that are ever more complex and lifecycles that are ever
shorter. Retailers continue to increase their service and delivery
requirements, with stiff financial penalties for non-compliance. On the flip side, more and
more real-time data are becoming available, with automation technology rapidly getting cheaper,
more capable, and easier to implement—raising the competitive bar for the
entire sector.
Traditional planning
processes and tools weren’t designed either to take advantage of technology’s
advances or to address the demands it creates. By and large, planning still
relies heavily on labor-intensive data aggregation and cleaning, manual
analysis, and personal judgment. Worse, with more customer and consumer demand
signals now available instantaneously, planners often feel compelled to keep
tweaking their plans, despite the weaknesses of existing planning systems and
processes. Well-intentioned adjustments end up creating more problems than they
solve, introducing even more errors and subconscious bias that can increase
costs and exacerbate service disruptions.
The time has come for a
new approach. Led by pioneering players in high tech and online retail, the
digitization of planning promises unprecedented levels of responsiveness,
agility, and speed. At one advanced-industry player, for example, an automated
inventory-planning system now automatically integrates and analyzes data from
more than a dozen different data sources. The ensuing reduction in inventory
levels allowed the company to release more than $75 million in free cash flow.
At a high-tech company, automating 95 percent of the order-to-ship process
(from order receipt through to transportation to the outbound shipping hub)
reduced end-to-end processing time by 60 percent, reducing variability and
making planning more consistent. And a major food and beverage player has
piloted predictive analytics to double the accuracy of its weekly demand
forecasts at the retail outlet level.
In this digital
environment, monthly or weekly planning cycles—driven not by the needs of the
business but by the capacity of planning teams—become a thing of the past.
Instead, no-touch, fully automated planning enables a continuous, seamless
closed-loop cycle of planning and re-planning, increasing accuracy and
efficiency both for the company and its customers.
What it will take
For automated planning
to work, however, machines will need to perform planning tasks at least as well
as a human: If they don’t, loss of trust or the need for time-intensive human
supervision will defeat the purpose of the exercise. For companies, this makes
activity mapping and segmentation crucial, so they can differentiate between
activities that can be fully automated and those that still require some level
of manual intervention.
Maximize today’s technologies, and
tomorrow’s.
Some activities, such
as the development of short-term demand forecasts for base stock-keeping units
(SKUs), can already be done completely automatically. The automation of other
tasks, such as mid- to long-term sales and operations planning (S&OP) or
supply- and demand-risk management, still requires development and
experimentation. In some cases, the required technologies or data resources are
not yet available, or are still too expensive or complex to apply economically.
Build the right foundation.
To fill these
technology gaps, companies will need the ability to experiment with innovative
processes and new solutions without disrupting their day-to-day operations.
That calls for a two-speed IT architecture, building a fast “test-and-learn”
environment (suitable for rapid prototyping and iterative development) on top
of the company’s current technology base. An agile development methodology with
weekly (or even daily) releases allows new approaches to be developed rapidly.
Once new solutions are refined and proven, they can be migrated to the
traditional architecture with a focus on repeatable and reliable service
delivery.
Automation is only part of the story.
To capture the full
potential of no-touch supply-chain planning, companies will also need to invest
in advanced analytics, machine-learning technologies, and process redesign,
while also adapting their organizational structures. Technology may produce
many “fractional FTE” savings, for example, as some parts of existing roles are
automated or eliminated. Turning those improvements into real cash savings or
re-investments will require a meaningful degree of role redesign.
The journey toward
automation needs a governing infrastructure that develops the right talent and
change- and performance-management practices. A proven approach relies on a
digital-planning center of excellence (CoE), whose function can range from
shaping the overall vision and direction to providing tactical program- and
performance-management assistance, such as through leadership
capability-building and vendor selection guidance. The CoE also provides
critical support for cross-functional collaboration, bringing disparate experts
together and aligning their work to the company’s digital strategy.
How to get there
A global
consumer-packaged-goods (CPG) player demonstrates the pattern followed by many
companies that have successfully digitized their planning processes:
Set the course.
The company’s leaders
started by establishing a bold aspiration that went beyond automation to
encompass digitization and analytics. Crucially, senior executives also drafted
a multi-year road map for reaching the goal, making the ambition far more
tangible to the entire organization.
Let the machine do the job.
Reexamining business
rules, systems, and tools throughout the company revealed the simple,
repeatable processes that were most amenable to digitization. In some cases,
existing enterprise resource planning (ERP) systems were enough to automate the
relevant tasks, others relied on basic robotic-process-automation (RPA)
software. Ultimately, more than 90 percent of planning tasks were automated,
dramatically reducing the number of manual interventions required while
providing decision support tools at the planner’s fingertips.
Infuse advanced analytics.
Identifying the biggest
paint points in the company’s supply-chain planning uncovered major
opportunities for advanced analytics. After collecting all available data—no
small task, but one made increasingly feasible by new technologies—the company
was then able to apply machine-learning algorithms to improve the accuracy and
granularity of its inventory-management, production-scheduling, and
demand-planning processes.
Think beyond S&OP and integrated business
planning.
As the new model is
fully implemented, reporting will happen on a continuous basis, eliminating the
need for monthly and weekly planning cycles and enabling faster, leaner, and
better decision making.
Hardwire the process into the business.
To make the entire
structure self-sustaining, the company revamped its governance structure,
basing it on an aligned and clearly defined set of truly cross-functional
performance indicators (and incentives) for all operations, commercial, and
finance functions.
This type of low or
no-touch planning process can dramatically increase service levels, while
reducing supply-chain costs and inventories to levels that most CPG players can
only dream of today. And it’s available now, not years in the future.
https://www.mckinsey.com/business-functions/operations/our-insights/the-route-to-no-touch-planning?cid=other-eml-alt-mip-mck-oth-1808&hlkid=e04678b9e95e4c41bd8ccd36d962316d&hctky=1627601&hdpid=9b98304e-1abf-4a88-9c65-aacb77e02b33
No comments:
Post a Comment