On the Benefits of Co-Location

By Hannah Mellow on 11 July 2014

Seating arrangements in an office are known to shape our contact patterns: we tend to communicate most often with those sitting close to us. For this very reason, who sits next to whom is an important matter. Most companies decide to cluster business units in order to enhance team spirit and identity and to ease communication flows among team members. With for instance everyone in sales sitting together in close proximity, not only is team cohesion among the sales function increased, it also allows others to interact with sales more easily since everyone knows where sales has its home in the workplace.


While this sounds like a logical and straightforward process, however, the matter is not so easily settled. Many companies in our post-industrial society have multiple reporting lines and follow a so called matrix organisation, i.e. each member of staff has both a functional affiliation (e.g. advertising sales), but also a business line-based affiliation (e.g. media brand for which advertising is sold). With multiple affiliations clustering never works perfectly. So for instance in a Media business, would the advertising sales team sit together with the editorial team of a particular media brand to create a stronger brand affiliation, or would the sales team for one brand sit together with the sales teams for other brands to create a larger sales area? 

The diagram illustrates the two different propositions of co-location. We would speak of a mixed co-location approach if functions (sales, editorial, marketing, IT, research, etc.) are mixed up and share one area in the office to create a project- or brand-based identity. In contrast, creating larger mono-functional areas (i.e. a separate area for each function) would then follow a clustered co-location approach.

In one of our consultancy projects of a large advertising agency, we had the chance to observe the effects of both co-location approaches very closely. In advertising, two functions work closely together: the creatives and the account managers. Each client works with a group of account managers and creatives to deliver a campaign, so from the point of view of easing communication it might make sense to co-locate mixed teams by clients (and this is what many agencies in fact do). Yet the different working cultures and roles of the two functions may result in conflicts; also projects change quickly and some people may be involved in many different projects simultaneously, so ideal seating arrangements may be rather fluid and difficult to pin down.

In the particular case we have studied, the agency (for historic reasons) had two different units A and B, both delivering campaigns. Each of them had their own teams of account managers and creatives. With 46 account managers and 66 creatives, unit A was slightly larger than unit B (46 account managers and 48 creatives respectively). What differed between the two units was the seating arrangement: while unit A had separate areas for the two functions, i.e. followed a clustered approach, unit B co-located pairs of account managers and creatives, thus working with a mixed co-location approach. The two different ways of arranging the teams are clearly illustrated in the floor plan below. In both cases, the team areas span across floors.


Floor plan and allocation of space to Departments A and B

This difference in the spatial allocation of team areas was reflected in the structure of the relationship networks of the teams.

In unit B higher levels of unplanned face-to-face encounter occurred on a daily basis: among the 46 account managers and 48 creatives a total of 453 ties were reported, whereas the larger unit A with 46 account managers and 66 creatives only showed 391 connections of daily contact. Interestingly, the percentage of ties connecting the two functions, i.e. ties reaching across the boundaries of teams was 35% in unit B and only 25% in unit A. This means that the mixed co-location approach for unit B resulted in higher levels of contact overall and in a higher degree of face-to-face encounter between creatives and account managers.

Through our surveys, we also established the degree of usefulness between individuals. On the level of extremely useful contacts (defined as ability to provide people with resources and information to get their job done), 328 connections were reported for unit B and 309 for unit A (see network images below). While unit B as the smaller department still created higher levels of collaboration and cohesion, this time unit A was more successful in bringing teams together, since 38% of ties reached across team boundaries in unit A as compared to only 29% in unit B. This means that the mixed co-location strategy of unit B again increased the overall levels of usefulness in the unit, yet more extremely useful collaboration across the team/functional boundaries was achieved by clustering the functions and giving both account managers and creatives clearly allocated workspaces in the office as in the case of unit A. The table below summarises the results.

46 account managers, 48 creatives (94)
46 account managers, 66 creatives (112)
453 daily contacts
391 daily contacts
Ties connecting functions: 35%
Ties connecting functions: 25%
328 highly useful contacts
309 highly useful contacts
Usefulness ties connecting functions: 29%
Usefuless ties connecrting functions: 38%


What this shows is that the levels of contact are clearly spatialised: more contact between the creative and account managing functions occurs in the close spatial proximities of the mixed co-location approach. The results also suggest that patterns of collaboration, i.e. who finds whom particularly useful in their work is less a function of space and more influenced by work-related factors such as working cultures. The fact that creatives and account managers in unit A are more closely connected to each other than their counterparts in unit B may therefore be rooted in the nature of tasks, roles and work procedures (average size of project, average longevity of project, how often individuals mix up between clients in each unit, etc).

Coming back to the networks of usefulness and the impact of space, we can detect more differences between unit A and B.


Most notably, the network diagrams of high usefulness highlight that the account management team in unit A, which had dedicated team spaces showed higher cohesion – the majority of staff were connected in one large cluster. In contrast, the account management teams in unit B fell apart into various unconnected clusters. 

From this case of an advertising agency with two different business units we can conclude that mixing up creative and account management functions results in higher levels of contact between teams. Overall levels of usefulness were also higher in the case of mixing functions.

In summary, there is no simple answer to the dilemma of mixed co-location versus clustered co-location – each approach can have its advantages and creates different affordances for interaction, collaboration and team identities. Nevertheless, what we were able to show is the important effects co-location can have. By strategically choosing who to place next to whom in an office, connections can be fostered and important work processes can be supported.

Therefore, based on their work requirements and organisational cultures, a business needs to decide whether it is more important to create cohesion within a function (suggesting a clustered space planning) or alternatively, increase the levels of contact between functions (suggesting a mixed co-location approach). Or maybe the answer lies in being much less fixed and prescriptive about where people sit at all… But that is another story.

Topics: serendipity, data, design, evidence