Tuesday, 22 October 2013

Planning for Productivity

The Holy Grail

For many years leading lights in the property and construction industry, such as Paul Morrell, have referred to the connection between office design and business performance (or productivity) as the Holy Grail. There is a view that the relationship is elusive and intangible, a myth even. That in itself is not a problem, but one consequence of believing that the impact of office design on productivity is not easily demonstrated, is that it is generally ignored. From a business perspective, ignoring the effect of your workplace facilities on your workforce’s performance is not just na├»ve but also irresponsible.

Incidentally, the story of the Holy Grail is a “monomyth” – a folk tale that is found throughout many cultures. The moral of most monmyths is that a hero ventures forth on a quest to seek a supernatural wonder but returns the wiser, even if they did not find what they went in search of. I think we have been on the quest for the link between office design and productivity for too long, but many of us are none the wiser.

Because of (wrongly) perceived measurement difficulty and lack of evidence, the potential impact of office design on business performance is generally excluded from the business case for workplace projects. The business case is simply weighted in terms of reducing property costs. This perpetuates the notion that property is a cost burden rather than a potentially lucrative return on investment for the business.

Furthermore, in completed projects the effect of the design on productivity is not thoroughly tested. So we may never fully understand whether the design had a beneficial or negative impact on business performance. So long as the new office design reduces property costs it will be deemed as successful, regardless of the consequences for the business. 

The key asset of any business is its people; they are also the most expensive component. There have been many discussions about ratio of staff to property costs – usually based on pie charts showing that salaries are around 85% of the total business cost over the life-cycle of an office. The implication of such charts is that relatively small changes to the design, construction and operation of office buildings can have a huge on-going impact on people and business. It is perfectly sound business sense to attempt to save money on the construction and operation of buildings but only if we are cognizant of the potential harm to business performance.

On a side note, I would argue that such charts are conservative as they ignore the revenue generated by the workforce which should be several times their cost. Anyhow, these cost ratios have been known and discussed for some time, since at least 1965 (by the National Bureau of Standards), but sadly the potential consequences of poor design are still pretty much ignored. The key question for the financial director and his/her colleagues is how to accurately measure or predict changes in productivity.

Predicting Productivity

This issue has been going on for some time; measuring productivity dates back circa 7,000 years to the Sumerians, who kept records of their workforce’s time and output compared to their costs (pay, food, lodging etc). Empirical research into the effect of the workplace on performance dates back to the early 20th century, which coincides with a growing interested in scientific management theory. In the early 1990s established professional bodies, such as ASHRAE and NEMA, debated and listed multiple means of assessing business performance. Furthermore, over the years, hundreds of academic studies have demonstrated a clear relationship between environmental conditions and office design with individual, team or business performance.

The problem is that the research reports a wide range of productivity gains; in our own literature review we found productivity gains ranged from 0.3% to 160%. Furthermore, the research is carried out in a wide range of work environments with different measures of performance of different work activities, which may not be directly relevant to office workers. The research therefore lacks credibility with the Financial Director looking for a single answer – the Holy Grail.

So we carried out a literature review of over 200 productivity research papers, and conducted a meta-analysis of 75 credible (robust) studies reporting 135 quantified effects on performance. The unique aspect of our review is that the reported percentage changes in performance were weighted according to the relevance of the research study to real offices and office workers. We consulted members of the Office Productivity Network to determine the relevance and weightings for the various environments that the research was carried out in, for example industry (35%), laboratories (40%), simulated offices (53%) and real offices (82%). They also helped develop weightings for the various performance metrics, for example self-assessed performance (48%), task performance (51%), absenteeism (67%) and business metric (68%). We used time utilisation data to weight the results by the time that the experimental activity would be carried out in an office, for example paper-based (8%), PC-work (24%), heads-down (32%) and general office work (64%).

We believe that the revised figures are ones that are more likely to be accepted by financial directors when used in building a business case. They were actually used to inform a refurbishment business case at the Atomic Weapons Establishment and justify additional spend on the workplace fit-out. The predicted range of percentage change in performance due to the various individual (environmental) design factors are highlighted in the image below.

On first glance, the upper levels of change appear to be quite small (1.7% to 4.4%). However they must be put in context of the property to staff cost ratio – the percentage of overall business costs spent on salary and the potential revenue generated by the workforce. Increases in productivity (revenue) of just 5% will over the life of the building easily offset the building construction and operational costs.

Most designs include more than one factor, but the single design factors above cannot be simply added to determine the overall effect. A few laboratory studies exploring the effect of multiple variables indicate that a simple rule of thumb, based on the Law of Diminishing Returns, can be used for estimating the combined effect of multiple design factors:

PO = P1 + P2 + P3

Where:   PO = the overall percentage performance change;
              P1 = percentage performance change due to 1st environmental factor;
              P2 = percentage performance change due to 2nd environmental factor;
              P3 = percentage performance change due to 3rd environmental factor.

Back in the 1990s several worldwide respected productivity researchers, including myself, estimated that the productivity gains from environmental design factors are likely to be in the order of 5-15%. But there are, of course, many other factors that affect productivity and business performance. For example, what Herzberg calls the motivational factors such as reward and recognition etc, personal factors such as experience and training, business factors such as marketing and advertising, and extraneous factors such as the demand and economy. These motivational factors usually have a larger effect on performance and will overshadow those from environmental factors.
Generic Productivity Metrics

It seems to me that the productivity metrics that are easier to measure, more directly related to design, and less affected by extraneous factors are also less relevant to business performance. For example, measuring an individuals reading performance in response to changes in desk surface illuminance. In contrast measures of productivity more relevant to business performance, like annual revenue generated, are more affected by extraneous factors and less clearly related to design factors. But back to the Holy Grail, the question remains as to which productivity metric is best used in the business case and in demonstrating value post-project?

When recently conducting some research on Lawyer's offices I discovered The Lawyer journal carries out a survey of the top 200 UK law firms. They usually collate information on business metrics such number of employees, fee earning revenue, lawyer to secretary ratios etc. However, last year they also enquired about the space and offices that the law firms occupy. I was particularly impressed that they reported the revenue per sqft as a key metric and indicator of how the space was supporting the business. What is great about this metric is that it is an in-house established business metric that most companies will track and report on a monthly basis.

The survey results clearly show city lawyers Allen & Overy (A&O) is huge, with 1.7M sqft and £1,183M turnover, and 10 times larger than regional solicitors Minster Law, with 70K sqft and £104M turnover. The survey also indicates that Minster Law’s space not only costs one-tenth (£2K per person) of that of A&O (£19K per person) but the revenue generated per sqft at Minster law (£1,508 per sqft) is also more than twice that at A&O (£685 per sqft). The workspace at Minster Law is not only more efficiently used but per sqft it appears to be offering better support to the business.

Revenue per sqft still captures efficiency, i.e. if we can generate more revenue in less space, but its focus has shifted from cost to revenue, from saving to generating money. I would argue the profit, or margin, per sqft would be an even better productivity metric. Revenue per sqft (or per sqm in Europe) could be reported by building or department or P&L group for a more granular understanding of how our office facilities support business performance.

The reported revenue per sqft is akin to sales per sqft, the universal metric in retail. We have all been impressed with the new Apple Stores, with their radical redesign and enhanced user experience. In 2010 Apple’s retail sales, excluding online, jumped 70% compared to the overall retail industry’s sales growth of 4.5%. The average sales per sqft for who's in malls in the USA is $341 per sqft; Apple stores are 17 times higher than the average and twice as high as second place retailer Tiffany & Co. Apple doesn't have the highest sales per store, it comes around 10th in place, but Apple is clearly getting maximum benefits per sqft out of its retail space - their footfall is higher and their operating margins are higher. These standard retail metrics all clearly demonstrate how the new store design has had an impact on performance.

I consulted a senior retail analyst at O2, previously at Sainsbury and M&S, and he explained how the sales per sqft (yes sqft not sqm) are continuously monitored across all of their stores. If some intervention is made, such as a shop redesign, new branding or a regional advertising campaign, he can monitor the impact on the sales per sqft. If the numbers show no change in performance then they know the intervention failed. They can also monitor how long the impact of the intervention lasts. I also recently learn that in some shopping centres the retailers may be paying a "turnover rent", this is a rent that is proportional to the sales figures. The shopping centre developers will have a good handle on the weekly revenue of each shop, know how they are all performing and know the effect of any design changes or campaigns.

My friendly O2 analyst considers this kind of data and continuous monitoring fundamental to running a business. In workplace it's rare that we measure the impact of our intervention on the business, and when we do they are usually a one-off measure with little context, making it extremely difficult to interpret the result. When we do bother to measure the relationship between design and productivity then we attempt to measure the wrong thing. We focus on the minutiae and try to measure individual performance such as time to complete a task. But we need to focus on good objective and existing business metrics like revenue per sqft. It is akin to trying to measure the waggle dance of the worker bee rather than how much honey is produced by the hive.

So in conclusion, there are lots of ways of measuring productivity and lots of studies demonstrating the relationship between design and business performance. When presenting the business case for workplace projects to the Board, we should focus on the potential effect of design decisions on business performance, rather than just report cost savings. We must also measure the impact of our workplace designs on business performance; measurement difficulty is not an excuse to ignore the impact. Rather than attempt to measure productivity using specific and complex metrics, we should use widely available business metrics which are readily recognised by the business and the Board – revenue per sqft is the obvious one. We should also monitor business performance over time and across locations, so that the result of interventions such as office design can easily be tracked.

This blog formed the basis of my presentation at the annual Workplace Trends conference. See me present the paper here.


  1. Guys I know that we live in metric world this side of the Atlantic but just following the Lawyer's lead and pandering to the agents.

  2. Nigel good thoughts. Here's a true scenario. Two like for like organisations. The first operated out of a big coverted warehouse (really poor space) but had a great leader and a great set of first class managers. The other was in fancy City offices with a poor leader and ineffective / average managers. Which one was the more productive? I would also add, which set of staff were the happier? I can't reveal the identities. Suffice to say they both exist today. I am not sure if there are any prizes for the answers.

  3. Nick
    I take your point and I do mention that there are other contributing factors that affect our performance. Our world is that of property and our remit is to ensure the workplace supports productivity and facilitates motivation and good leadership etc.

  4. Nigel, many thanks for writing this up. I'm drawn to the "factors affecting performance" diagram and agree that Nick has a point. It's a huge question, but unless we have a sense of the relative wieghts of each of these factors, it is hard to argue that the workplace inputs are significnt compared to other ways in which an organisation might spend its resources, such as better management or giving everyone an iPad. This is why I think we keep getting faced with management-by-accountancy questio that try to reduce everything to ROI.

  5. Hello Nigel,

    Interesting that you created some heat as well as light with your presentation. There are some good comments here already. As you say productivity need not be that elusive a variable. The problem comes when it is conflated with cost cutting. These two elements are discreet and correlate negatively. It is this that business seems to find hard to accept and yet the data are overwhelming..

    Many workplaces are abysmal. However, design is but a small part of this (as alluded to by Nick). The key in all the productivity research we have seen, conducted and consulted is helping people to realize their own identity at work.

    Evidence is pretty clear; space (and its design) needs to reflect the people that work within it, management need to trust their colleagues and autonomy should be maximized. Sounds easy but, as an example, how many office ‘hives’ are going to switch off the monitoring machines, stop stamping corporate identity across their staff's environment and relax the restrictions of flexible space? There are no gains in reinventing bad methodology..

    You are doing great work here. Keep going.


  6. Many thanks for all the comments on LinkedIn: http://linkd.in/1ceJAYR