Operational Excellence: Productivity in assembly operations

 Joe Haviv 180

 Joe Haviv

Boards and CEOs are increasingly called to respond to ever-changing challenges, whether through industry disruptions, competitive pressures, or the quest for optimal profitability. However, traditional approaches aimed at conquering labour productivity are falling short of targets. Similarly, automation, though an important lever, can’t deliver its ultimate promises. How then should a VP of operations respond to such new environments and assert the crucial relevance of manufacturing? What are the hidden jewels? Where and how should resources be directed? What follows is a roadmap to new performance thresholds.

Ask a manufacturing executive for the plant capacity and you will get a measure of output per unit of time. Though correct, this fails to recognise the multiple interdependencies amongst manufacturing processes and the critical output constraints. A better approach is to focus on defining clear production value streams, identifying constraints within each stream, optimising the current configuration and rebalancing the production process to yield the fastest throughput times. 

In the real world, two questions are often raised: (1) how to define proper value streams and (2) how to optimise the current configuration. Defining value streams serves both communication and analytic purposes: the most effective way is to limit the number to a manageable subset (somewhere between three and 10) by determining value stream ends at specifically identifiable output points, whether a subsystem, a rough casting, or when specific value-added features have been incorporated. But optimising the current configuration can’t be as prescriptive, as it is crucial to precisely quantify constraints and attack them systematically. Accelerating throughput time might require streamlining of set-up times, new changeover procedures or creating adequate buffer stocks.


In most manufacturing environments materials represent 35 to 50 percent of total product cost. Yet materials optimisation efforts tend to be at best sporadic and often lack rigour. Setting aside the obvious approach of rebidding for materials, overlooked aspects include specification adjustments to use common materials, rigorous certification and quality inspection practices, standardisation of components, changes in chemical compositions, and kanban arrangements with suppliers. Though not glamorous, the maths of targeting materials improvement are compelling.

The proverbial “pot of gold”. Long forgotten if not ignored, warranty repairs and costs are an extraordinarily rich source of data and improvement potential. Pareto analyses of claims can yield substantial insights into ineffective assembly operations, unwarranted failure modes, product design faults and component deficiencies. Once the data is compiled, cross functional teams can attack specific improvement areas.


Long viewed exclusively as a quality indicator, first pass yield is one of the strongest indicators of the improvement potential in manufacturing operations. It embeds all deficiencies, whether through scrap, rework or excess intermediary work-in-process. Above all, it quantifies the limits to a faci-lity’s output, and can provide direct insights into improving throughput time and output.

Nothing builds success more readily than early wins and galvanising the organisation in the search for more. Operational interventions can be readily expanded to other functional areas, maintaining the mantra of “speed based competition”. As this process unfolds, new areas can be explored: product design, quoting processes, prototyping and customer service. Speed-based competition is no longer a luxury but a necessity. Charting a course and a mandate for senior operating executives is not only compulsory, but can successfully reposition a company for long-term sustainable competitive advantage.


A common scene in a manufacturing war-room: tasked with new headcount reductions plant executives struggle with where to go next. Traditional approaches have been exhausted: automation, quick changeovers, cutbacks by edict and productivity training have been implemented with strong success, yet the Board is requesting more actions. The quest for new thresholds has to necessarily seek new approaches and solutions, sometimes dusting off proven old techniques.

Managing a labour intensive manufacturing environment often overlooks a very simple truth: labour is tied to individual contribution, and the management approach needs to fully reflect such reality. Rather than aggregate or average measures which most often lead to amorphous conclusions, leaders need to embrace the fact that regular, consistent measurement of pre-established performance expectations at the individual level are the key to success. Above all, analytics need to be tied to clear consequences for meeting expectations (i.e. rewards) or failing to do so (coaching or more severe interventions). This level of granular analysis is critical to reshaping shop floor performance and culture.

Productivity measures are almost entirely focused on productive time on the shop floor. Overlooked in such effectiveness oriented processes is “lost time”. Especially in large scale operations, the actual start and end time of production can vary widely from scheduled shift times. Similarly, scheduled work breaks, often triggered by audible signals, can falsely suggest that pause times are respected on the shop floor. Over the length of an eight-hour shift, it is not unusual to identify lost time levels of an hour or more: that’s a 15 percent time recapture that should not be forgotten. Though such gaps are easily measurable, corrective actions need to properly incorporate cultural and behavioural changes within work crews.

Never underestimate the importance of segmenting complex manufacturing processes into specific value streams. By adapting concepts from cell manufacturing, value streams can be further subdivided into shorter, clearly identifiable process sequences that yield a specific output (i.e. closed loop). This new unit of measure can be fully examined to redefine task sequences, component progression and tooling, thereby leading to new job contents, responsibilities and accountabilities. Even well-designed processes can benefit rapidly through such activities, unearthing force of habit shortcomings, and providing compelling evidence through simple pilot programmes.

Boooring! Indeed they might be so, but – when properly utilised – few techniques can provide as swift a change in analytics, behaviour and root cause identification. Success hinges on a few critical factors: (1) record as few variables as needed, typically time and output measures, resisting the temptation to cover the waterfront; (2) enforce strong discipline in properly coding causal factors for deviations; and (3) routinely conduct Pareto analyses to identify immediate corrections. Above all, proper leadership involvement can ensure that goal boards are a live tool, rather than a burdensome reporting device.

Tasked with capturing savings, manufacturing executives should take the unusual step of examining the complexities introduced on the shop floor by product line proliferation. Changeover and set up times, customisation requirements, and incremental processes should be examined and quantified, especially for low volume stock keeping units. Seldom have such analyses originated within the marketing organisation, and most commonly the hidden cost of labour is not an explicit factor in deciding product line breadth. Though a clear “overstepping of bounds” by manufacturing executives, the analysis can become a strong catalyst for cross-functional decision making.

Protostar Partners is a New York-based private equity mid-market buyout firm and an early pioneer in the direct secondaries sector.

Joe Haviv is the managing member of Protostar Partners and has over 20 years of private equity investment experience across a broad range of industries. He also spent 12 years at McKinsey and Company focusing on operational and strategic excellence.