Evaluating productivity is deceptively simple: output over input equals productivity. In reality, however, how you measure input and output can be surprisingly important. The arrival of the industrial Internet of Things (IIOT) has not altered the fundamental problems. We may have more data but deciding which data to use is still the key question. I tapped Christer Bodell for his views on how our perspectives and expectations and changing even as realtime streams are enabling us to do more with less. Here is what he answered to my questions.
Q: How often does the concept of productivity feature in your discussions with customers?
"Quite often, which should not be surprising. Operational efficiency is at the heart of profitability, and every CEO is interested in improving productivity. Every few years, there is an initiative to re-assess the basis upon which productivity is measured. Although it can be tempting to call in experts like statisticians or economists or model-builders, I have found it is important to first ask managers what matters to the business, and ensure that they are involved in creating the index and models. A beautiful, accurate model that does not provide the answers the business needs will not be helpful. It is important to involve managers who know the business in creating indexes and models."
Q: How complicated has productivity reporting become in the age of big data?
"Actually, simplicity is still king. It is better to have a simple way of evaluating productivity that is understood and accepted by everyone than a beautiful, accurate model that nobody can understand, and therefore either does not use, or does not trust. Start simple and build up, rather than starting with a complex model and reducing down. But, we also need to remember that productivity goes beyond direct labour costs and product output. There are many possible inputs for measuring productivity: direct and indirect labour costs, materials, overheads, even capital costs and depreciation. The measurement of output can also be significantly more sophisticated than simple ‘products out’. For example, in some industries, a rapid turnaround time may be more important than keeping down costs, and outputs therefore need to include a time element, or possibly weighting of factors."
Q: How have you seen productivity indices evolve as more data has become available?
"There is no question that the options for creating meaningful indicators have improved. Multifactor indexes are arguably the more interesting of developments. Measuring several factors, whether or not combined into a single index or indicator, have become popular. Even a relatively simple multifactor assessment such as overall equipment effectiveness (OEE) includes assessments of machine downtime, and quality and quantity of output, and is therefore significantly more useful than a simple ‘outputs over inputs’ measure. Multifactor indexes also allow businesses to manage trade-offs. Multifactor indexes enable the business to avoid creating perverse incentives to particular behaviours, whether slow production, or reduction in quality. They also allow recognition of the quality–time–cost trade-offs that are inherent in any business, and let the business focus improvement attention where it is most needed: that is, on the element that is having the biggest impact."
Q: We talk about not comparing apples with pears, but how often are we guilty of that when measuring productivity?
"There are so many ways to measure productivity, and a wide range of values can be obtained, so it is up to managers to decide which one is most appropriate and fair under the circumstances. But yes, fair comparisons are indeed crucial. Just as important to remember is that a productivity index needs to provide useful information. In other words, it has to measure what matters. This is the reverse of the old adage that what matters is what is measured, where people concentrate on what is being measured at the expense of everything else. For a productivity index to be useful, it must measure what matters to the business, even if that is hard to do. Proxies are fine, but you must include what matters."
Q: What changes have you seen in how managers are using productivity reports?
"As with much of analytics, we are seeing the availability or realtime, or near realtime insights influencing day-to-day decisions. Indicators are not absolute, they simply highlight potential issues. You do not have to measure everything to be able to highlight potential problems and issues. Good use of the right indicators can enable managers to build up a picture of productivity across an organisation without having to measure everything. A productivity index is the beginning of the questions, not the end. For many this is a paradox - more data can lead to more questions- but for any modern manager it should be natural, and works as a good indicator of how fit the manager is for the role. It should be natural to consistently advance in analytics and reporting."
How IoT is re-defining productivity
We hosted a digital panel discussion on Twitter to explore how IoT is re-defining productivity. You can read a Storify collection of the discussion highlights here: #IoT and productivity: the adoption challenges.