Open plan v private offices déjà vu

Another round of articles was recently published (mostly by Inc.) on why private offices are better than open plan, and consequently suggest we should eschew open plan design.  Despite the open plan office being described as a management fad by one journalist, the adoption of open plan dates back to the 1950s* and the occurrence of such negative articles has been ongoing ever since then. 

However, every five years or so, such articles seem to converge and attract the attention of the mass press. It seems that academics and journalists have a particular dislike for open plan. But open plan is clearly not a “fad” and I suspect it is here to stay, so it is more useful to focus on how to resolve any demonstrated issues with open plan rather than simply say it should be banned. 

The points I raised in my blog in 2013 (also published by Workplace Insight) in defence of good open plan design still apply today. My main issue at the time was misinterpretation of the research by the press in order to demonise all, rather than poor, open plan design. In terms of the physical elements, poor open plan tends to be high density with no screens or semi-partitioning and minimal non-desk work settings. 

Myself and others within the workplace community, e.g. Neil Usher a.k.a. workessence, are currently frustrated with the mass press view that office design is a simple dichotomy of open plan or private offices – it is not. To illustrate the point, and aid an informed discussion, Neil has created a taxonomy of 13 office typologies. One dimension of office design is how it ranges from the fully enclosed to fully open plan. From my own experience, the solutions and scenarios at extreme ends of a range tend not to be the best. So, a fully open plan environment, with no (semi-)private spaces, or a workspace of just private offices are unlikely to suit the needs of the dynamic modern organisation. 

I recently conducted and presented a literature review of open plan schools, a highly controversial subject in education circles (I aim to publish my report next month). In brief, the research shows that fully enclosed schools do not support modern pedagogic student-centred teaching practices, and neither do the fully open plan ones, due to disruption when conducting contrasting activities in the same space. The researchers acknowledge that school design is on a range (Fisher & Dovey propose five types) and the schools which are mid-range and more successful are termed Innovative learning Environments (ILEs). The best ILEs facilitate the dominant approach to teaching in the school, are well-managed and well-designed with open areas but also easy access to semi-partitioned spaces. 

I think it is time that we design Innovative Working Environments (IWEs). The design of such spaces starts by understanding the full range of activities in the space, the current and desired working practices and the existing management style; it continues by incorporating the daily operation of the space and encouraging the preferred behaviours within it. As mentioned previously, high density desking (often implemented as the cheaper solution) is unlikely to support the majority of organisations, but there are nevertheless exceptions such as call centres and trading floors. 

The recent run of articles criticising open plan are mostly a consequence of recent research published by Bernstein & Turban in the Harvard Business Review (HBR). The authors found that when moving to open plan “Contrary to common belief, the volume of face-to-face interaction decreased significantly (approx. 70%) … with an associated increase in electronic interaction. In short, rather than prompting increasingly vibrant face-to-face collaboration, open architecture appeared to trigger a natural human response to socially withdraw from officemates and interact instead over email and IM.” To evaluate interaction, the researchers had their participants wear a sociometric badge which included a microphone, infrared sensor, accelerometer and location tracker (see image). 

In psychology, when conducting research it is common best practice to minimise invasiveness and avoid experimental elements that may influence the results, that is to create a “blind” study. For example, the mere presence of the researchers may have affected the results in the famous Hawthorne Study. It is unlikely that the cumbersome sociometric badge worn by the participants of the HBR study did not affect the results, especially if their colleagues learned there was a microphone attached. 

My second point is that the researchers measured interaction and, despite the title of their paper, not collaboration. In a previous report I explained that collaboration is when two or more people come together and produce something that they could not have created alone. A precursor to collaboration is trust and a precursor to that is social interaction. So, generally all interaction is good but there is a point when interaction becomes a distraction and, for productivity, should be reduced. A fellow workplace strategist, Sam Sahni, pointed out that the researchers reported an average of 5.8 hours of interaction per day prior to the move to open plan, which was reduced to 1.7 hours per day post-move. For most organisations, having their employees engaged in interaction for approximately 75% of their working day is not conducive to productivity – did they really only have circa 2 hours per day to complete focussed work? It could be argued that the new open plan design allowed the occupants to work with less distraction and be more productive. The increase in electronic interaction may have been because the participants learned to respect their colleagues' privacy and reduce distraction and interruption in their new environment. 

Another fellow workplace strategist, Andrew Mawson, pointed out the that environments studied were not fully described – no plans or images were provided. We know that in one study the organisation “decided to use the latest open office workstation products to completely transform the wall-bounded workspaces in its headquarters so that one entire floor was open, transparent and boundaryless” and in the second they moved to “assigned seats in an open office design, with large rooms of desks and monitors and no dividers between people’s desks.” I shared early workessence’s taxonomy of open plan design solutions, and I also explained how there is good and poor open plan design. It is not clear from the HBR paper whether the occupants were moved to well- or poorly-designed space. That there were no desk dividers and large boundaryless areas, perhaps with no shared facilities to help break up the space, indicates a poorly designed space. There is also no mention of the design and transition process, for example was there a change management programme to help prepare the staff for their new working environment. 

Previous research has indeed shown that it is more difficult to provide privacy and control over noise and temperature in open plan environments, but survey data such as the Leesman Index and case studies also show clear positive benefits. In particular, agile working environments, where the staff have more empowerment over where and when they work, are more successful. Personally, I can’t see the benefit of providing a building in which the workforce are placed in rows of isolated boxes – you may as well send them home. Let’s monitor all working environments (more POE please) and take up the challenge of improving them, moving from poor open plan to better Innovative Working Environments. 

* The concept of the open plan offices has been around since the turn of the 20th Century, and Frank Lloyd Wright’s Johnson Wax Headquarters is a fine example of early open plan design. But the uptake was more prevalent in the 1950s-60s with the development of the landscaped office or Bürolandschaft.