Amid the extraordinary transformation of manufacturing over the past decade as the Fourth Industrial Revolution (4IR) advances through sector after sector, some manufacturers still face a challenge almost as old as manufacturing itself: how to achieve lasting productivity gains from labor-intensive operations.
The seemingly obvious solution is sometimes summarized as swapping labor for capital, in the form of automation. Despite the availability of ever-more-sophisticated machinery at ever-lower cost, in many situations a more attractive alternative is to use digital and analytics technologies to support people rather than supplant them.
Even today, labor-intensive sectors can include everything from toys, apparel, and jewelry to medical devices, consumer-electronic products, electrical goods, and automotive components. These sectors are a critical force in emerging economies, providing employment that reduces poverty and strengthens social stability.
Where labor is so instrumental in creating value, managing a workforce becomes a matter of strategic importance. Companies that perform better at hiring, retaining, and—most crucial of all—engaging their workers can build a substantial advantage over their competitors. To do so, they must overcome a host of challenges, some of which became even more vexing under COVID-19. Ensuring that workers feel safe must, of course, be the highest priority. But providing a protective environment may not be enough to persuade every worker to come into factories—particularly workers juggling responsibilities to care for children and families that may have been disproportionately affected by the pandemic.
Companies that perform better at hiring, retaining, and—most crucial of all—engaging their workers can build a substantial advantage over their competitors.
At the same time, spikes in demand have only intensified competition for workers in sectors where the clustering of operations makes it easier for people to change employers. Even when an employer can find and keep workers, changing skill requirements often create a mismatch between the capabilities the company needs and the ones its workforce can offer.
Nevertheless, a few companies have found a promising solution. New digital and analytic tools, applied diligently, can enable plants in labor-intensive industries to become best-in-class performers even as they continue to employ large numbers of people.
Common challenges for labor-intensive industries
Three longstanding issues have become especially critical for labor-intensive employers: unpredictable day-to-day attendance, above-average turnover, and mismatches in skills.
Who will make it to work today?
A major question facing many manufacturers in labor-intensive sectors is simply how many of their workers will come into the factory on a given day. In these industries and the communities where they are located, workers often face hurdles ranging from limited access to emergency family care to unreliable transportation. Unpredictable attendance limits the ability of managers to plan and execute production schedules, thereby reducing productivity, service levels, and profitability in factories.
These problems have been evident for some time. In 2017, for example, the McKinsey Global Institute found that women account for 39 percent of Mexico’s manufacturing workforce and 47 percent of China’s. Their earnings represent crucial household income, but childcare options may be limited or unaffordable, so that they can’t come into work when their children are sick or other family issues arise.
How long will workers stay with my company?
High turnover, often attributable to the “cluster effect,” is another fact of life in labor-intensive manufacturing. The many special industrial zones (or similar areas) around the world tend to attract companies in the same and related industries up and down the value chain. The benefits from such critical mass are well known: warehousing and logistics become more efficient, innovation tends to accelerate, and these zones start to attract more talent. But for at least some companies, there’s also a trade-off: close proximity makes it easy for sought-after workers to switch jobs.
A cluster’s geography may add further complications. In some countries, such as Mexico, industrial zones are located near adjacent foreign markets but far from many of the home country’s population centers. Under these circumstances, workers may be more likely to see their jobs as temporary—a way to save money before eventually returning home.
Will my workers have the right skills?
For manufacturers all over the world, hiring workers with the right skills is a challenge, but the challenge is accentuated in emerging markets, where a higher share of workers do not receive education beyond elementary school. This problem leaves manufacturers dedicating significant resources to onboarding and training—if they can even find enough applicants to fill open positions.
High turnover levels make investments in training more costly as well; even where it is less expensive than it is in mature markets, the outlays do add up. What’s more, managers may be reluctant to devote sufficient resources to training when a competitor might end up benefitting from it.
Over time, one medical-device player found that while personnel with previous knowledge could be trained in just three days, an inexperienced newcomer would require up to four weeks. This disparity can encourage employers to rely on expensive outsourcing or temporary labor sources—particularly in peak seasons—eroding some of the financial advantages of operating in a given location.
The analytics solution for productivity
Despite progress in channeling the enormous quantities of data that manufacturing now generates, only some manufacturers have fully succeeded in using these data to inform their decision-making processes. In labor-intensive sectors, the benefits have now become compelling, and that gives some players an opportunity to leap ahead of their peers.
A ready analogy involves ride sharing: today’s giants have shifted passenger logistics, traffic navigation, and the collection of money to central computer algorithms, so drivers can focus on driving. Indeed, they get virtually all of their instructions and feedback through their mobile phones, without human intervention—dramatically reducing training costs while also improving the quality and reliability of service.
Gathering the right data
Manufacturers have an opportunity that ride-share companies did not: data gathering is often built into production machinery. In labor-intensive industries, the important task is to implement mechanisms that measure output and yield at the most granular level. Companies use different tools and software; Exhibit 1 portrays a generalized example.
In a North American original design manufacturer’s digital-manufacturing excellence transformation, an electronic motion-measurement system calculated non-value-added (NVA) activities and rates for different tasks. Using motion sensors and geolocation to understand the workers’ positions in the plant, managers could easily generate spaghetti diagrams showing wasteful movement—they could actually see NVA. Similarly, an electronics-manufacturing player in Asia identified a potential productivity increase of 25 to 35 percent simply by eliminating the difference in unequal workloads between workers’ left and right hands.
Planners and schedulers can spend countless hours building the best production plans, but when employee counts vary at every shift, those plans must be readjusted in real time. Advanced-analytics algorithms can optimize the staffing of available employees, maximizing productivity and service levels by accommodating variables such as individual skills, available equipment, production targets, and shipping dates—all with minimal guidance from supervisors.
Dynamic scheduling bases production and staffing schedules on daily attendance, minimizing the impact of variability. After accounting for the actual number of workers in each shift, the system optimizes factory-floor staffing to improve utilization and productivity by using the individual skills and capabilities of employees to assign them to specific production process stations. That maximizes the profitability of plants and boosts the quality of their output.
At an electronics manufacturer on the US–Mexico border, for instance, annual turnover was more than 100 percent and daily attendance fluctuated by as much as 15 percent, no matter how much effort the plant leadership put into improving it. This unpredictability had multiple effects on production. For example, supervisors spent one to two hours a shift just assigning employees to stations, as they had to consider whether the operators available had the right training for the required stations and which products had priority.
In addition, on-time-delivery was low, because production supervisors who had only a limited view of demand guided the prioritization of production. Moreover, labor-cost absorption was below
60 percent, as the outcome of the assignment of labor to stations varied with the supervisors’ capabilities, and keeping track of more than 4,000 plant operators was extremely complicated.
As a result of all these problems, the plant implemented a dynamic scheduling system (Exhibit 2). When operators clocked in, an algorithm assigned them to a station by taking into account their skills and experience and the most pressing requirements for the shift, in line with monthly production targets. This step eliminated the one- to two-hour productivity loss—at a minimum, 12 percent—from the daily reassignment to stations and increased on-time-deliveries to 90 percent, from 50 percent. Labor-cost absorption rose to a much healthier 80 percent. Moreover, the tool could be deployed in real time, so that whenever operators became available, the system would recalculate their work assignments and suggest the next one. That gave managers the flexibility they needed to mobilize workers for stations that required specific skills or additional personnel right away to keep up with demand.
Narrowing skill gaps
For managers leery of training employees only to watch them decamp to competitors, analytics provides three responses that together bolster employee skills—and retention.
One way to boost returns on training investments is simply to help employees learn faster, so that
they can spend more of their time in value-added activities. Digital tools play a critical role in facilitating timely on-the-job capability building. At one manufacturer, for example, employees arriving at the beginning of their shifts would wait for up to two hours until the plant supervisor was available to coach them and provide them with a training module on a station. Deploying digital screens at every station made it easy to staff newcomers automatically to the least-challenging stations, where they could take multimedia training before initiating production. This step alone helped reduce average training time by about 70 percent.
A more advanced option helps complement the skills required of employees by filling gaps with artificial intelligence. Instead of requiring employees to take lengthy training courses, some companies have adopted advanced Internet of Things–based assistive technologies to guide operators through complex production processes. By keeping
track of the operators’ output, quality, and speed of production, managers can identify specific skill gaps and move operators to machines where they can complete on-the-job training via multimedia screens.
The final lever is to increase retention, in part by building better career paths. Many sites have several hundred or even thousands of employees per shift and only a few leaders, whose qualifications may have gaps. Productivity-tracking tools, such as those shown in Exhibit 1, make it easy to attribute performance to specific individuals. With this information, performance-based retention programs can be put in place for direct employees, increasing the retention of the top performers. Several companies on the Mexico–US border more than doubled average retention rates by following this approach.
Smoother operations can seem like an attractive opportunity in itself, but the economic impact is substantial as well. In a market in which cost, quality, and service levels are becoming more competitive, digital and analytics solutions can increase manufacturing throughput at the same direct and indirect labor cost, further increasing EBITDA. In one manufacturer’s case, a 5 percent increase in throughput generated a 45 percent increase in EBITDA (Exhibit 3).
Even before COVID-19, labor-intensive manufacturers could reap major benefits from adopting advanced analytics to address recurring pain points and challenges. The next normal offers even more opportunities as increasing competition and changing demand profiles exacerbate challenges while shaping a new landscape across the supply chain. This differentiator may well prove to be the critical factor determining competitive advantage for years to come.