Engineering utilization, productivity and throughput are among the most important metrics for measuring R&D performance. If you want to improve your R&D capability, focus on these three metrics.
Productivity and utilization directly determine throughput, and throughput is the most important of all R&D performance metrics. It measures the rate at which an R&D team develops production-ready products. The higher the productivity and utilization, the higher the R&D throughput. Higher throughput means the R&D team churns out more products in a given period of time. That usually translates to revenue and profits— assuming the rest of the enterprise pulls its weight. Big assumption, right?
Every R&D organization should constantly be boosting throughput, and the quickest way is to increase utilization. Utilization measures the percentage of the engineering workforce's time spent on revenue-generating activities. It's the percentage of time it spends on developing products – those intended to generate revenue. Here's the key point though—only tasks directly contributing to a specific product's development qualify as revenue-generating. All others are non-revenue-generating, and time spent on them reduces R&D utilization. Although there are "gray" areas, these activities typically include attending trade shows and conferences, pre-and post-sales support, corporate improvement initiatives, paid time off, etc. In a nutshell, anything that diverts resources from revenue-generating activities reduces utilization.
According to PRTM, a top management consulting firm, the utilization level of best-in-class semiconductor companies is around 73 percent. In contrast, many organizations are in the 40 to 50 percent range. Show me workforce spending less than half its time on non-revenue generating activities and I'll show you a company destined for bankruptcy. (In the interest of disclosure, PRTM is a partner of Numetrics.)
Boosting utilization is a two part process, the first of which determines the percentage of time the R&D organization truly spends on revenue-generating activities. Management consultants accomplish this fairly quickly via interviews and surveying the R&D organization. After tallying the numbers, "resource leakage" immediately reveals itself, and this enables the executive team to take appropriate action.
Interestingly, improving utilization is usually much easier than increasing productivity, although the two go hand-in-hand. Companies can improve utilization rates overnight by simply changing policies—e.g. "no more trade show boondoggles." In contrast, increasing productivity requires meticulous dissection and study of the tasks underpinning product development. Both need to be done, but raising productivity often takes years, whereas utilization improvements can occur almost immediately.
Yes, vacation counts in the utilization percentage. (But it has a material impact on utilzation only when the amount taken deviates from the industry norm. For example, overly generous accrual policies can result in employees accruing excessive vacation days, which can have a noticeable impact on utilization figures.)
Wrt R&D managers and project managers, the relevant question is not the job title, but rather whether the specific activity of the individual (e.g. manager) is directly contributing to the creation of a product or service that is forecasted to generate revenue. For example, time spent by a project manager truly managing the project (that is going to result in a revenue-generating product) is time spent on revenue-generating activity.
Thanks for your comment.
I enjoy your articles very much. It's always interesting to see how one's own organization stacks up.
Does vacation count against the utilization percentage?
How do R&D managers or project manager count in the utilization?
Wouldn't you agree that keeping track of the revenue-generating vs. non-revenue R&D spending is important? If your R&D organization's Utilization is consistly low, it is unsustainable business model. Naturally, some development projects can take a very long time, but what if that amount of time crosses a threshold in which it is no longer sustainable -- and there are no metrics in place to flag it? Likewise, one of the reasons why best-in-class companies have Utilization rates of only 70% to 75% is because they are devoting resources to longterm research and longterm development projects, so the Utilization metric is useful to ensure that companies' Utilization levels remain within the target defined by their business strategies.
Thanks for your comment.
In my opinion such performance metrics may not apply to certain R & D organizations which have long term goals to achieve and cannot generate revenue in the near term. The organizations who work on futuristic technologies cannot have any intermediate deliverables which can be marketed as products.For example organizations working on new drugs for cancer or AIDs , or working on say quantum memories. If such organizations are forced to concentrate on some immediate product delivery then they will fail to develop new technologies and products.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.