The Limits to Collaboration

 

Collaborative Overload

Evidence is emerging that too much Collaboration can be a bad thing – Economist:

…the current Harvard Business Review(HBR) has a cover story on “collaborative overload”; and Cal Newport of Georgetown University has just brought out a book called “Deep Work: Rules for Focused Success in a Distracted World”.

A growing body of academic evidence demonstrates just how serious the problem is. Gloria Mark of the University of California, Irvine, discovered that interruptions, even short ones, increase the total time required to complete a task by a significant amount. A succession of studies have shown that multitasking reduces the quality of work as well as dragging it out. Sophie Leroy, formerly of the University of Minnesota (now at the University of Washington Bothell) has added an interesting twist to this argument: jumping rapidly from one task to another also reduces efficiency because of something she calls “attention residue”. The mind continues to think about the old task even as it jumps to a new one.

On the “Information Residue” issue, there were studies done many years ago on the impact of telephone interruption. They showed that it took several minutes to mentally “set down” what you had been doing, and several minutes to “set up” the tasks the telephone call is about, (or re-set what you were doing) before effective work could again begin. The more complex the task, the more the set up / set down time. Too many phone calls and the worker was effectively spinning in air and could do only the most menial tasks, any real “knowledge work” was out the question.

Another issue The Economist notes is the poor understanding of the costs of Collaboration, which also plays to the above point:

A second objection is that, whereas managers may notice the benefits of collaboration, they fail to measure its costs. Rob Cross and Peter Gray of the University of Virginia’s business school estimate that knowledge workers spend 70-85% of their time attending meetings (virtual or face-to-face), dealing with e-mail, talking on the phone or otherwise dealing with an avalanche of requests for input or advice. Many employees are spending so much time interacting that they have to do much of their work when they get home at night.

Problems

The HBR article referred to above notes X main problems emerging from modern Collaboration systems – I have split the issues into teh 3 main issues:

• Volume
• Victims of Virtue
• Vampires

Volume

How much time do people spend in meetings, on the phone, and responding to e-mails? At many companies the proportion hovers around 80%, leaving employees little time for all the critical work they must complete on their own. Performance suffers as they are buried under an avalanche of requests for input or advice, access to resources, or attendance at a meeting. They take assignments home, and soon, according to a large body of evidence on stress, burnout and turnover become real risks.

As one commenantor in The Economist points out, Collaboration kills itself by the Network Laws

“There is a simple reason why performance declines with collaboration. It is caused by an arithmetic progression: one person = 100% work (theoretically, and rounding the numbers for this example); two people = 90% work and 10% collaboration overhead; three people = 80% work and 20% overhead; four people 60% work and 40% overhead, and so on.
Why the sudden drop in work? Because when you have four people, you suddenly have six interactions to service (to prove it, draw a box with a person at each corner and draw all the lines between them). Above four, the number of interactions increase again.
By about 15 people, work has dropped to almost zero and collaboration overhead risen to almost 100% (all those meetings ‘the alternative to work”).

Victims of Virtue

What’s more, research done across more than 300 organizations shows that the distribution of collaborative work is often extremely lopsided. In most cases, 20% to 35% of value-added collaborations come from only 3% to 5% of employees. As people become known for being both capable and willing to help, they are drawn into projects and roles of growing importance. Their giving mindset and desire to help others quickly enhances their performance and reputation. As a recent study led by Ning Li, of the University of Iowa, shows, a single “extra miler”—an employee who frequently contributes beyond the scope of his or her role—can drive team performance more than all the other members combined.

But this “escalating citizenship,” as the University of Oklahoma professor Mark Bolino calls it, only further fuels the demands placed on top collaborators. We find that what starts as a virtuous cycle soon turns vicious. Soon helpful employees become institutional bottlenecks: Work doesn’t progress until they’ve weighed in. Worse, they are so overtaxed that they’re no longer personally effective. And more often than not, the volume and diversity of work they do to benefit others goes unnoticed, because the requests are coming from other units, varied offices, or even multiple companies. In fact, when we use network analysis to identify the strongest collaborators in organizations, leaders are typically surprised by at least half the names on their lists. Also see the Chart at the top of the post, these people start to burn out.

Vampires

Other people start to (ab)use highly collaborative people.

Instead of asking for specific informational or social resources—or better yet, searching in existing repositories such as reports or knowledge libraries—people ask for hands-on assistance they may not even need. An exchange that might have taken five minutes or less turns into a 30-minute calendar invite that strains personal resources on both sides of the request.

Consider a case study from a blue-chip professional services firm. When we helped the organization map the demands facing a group of its key employees, we found that the top collaborator—let’s call him Vernell—had 95 connections based on incoming requests. But only 18% of the requesters said they needed more personal access to him to achieve their business goals; the rest were content with the informational and social resources he was providing. The second most connected person was Sharon, with 89 people in her network, but her situation was markedly different, and more dangerous, because 40% of them wanted more time with her—a significantly greater draw on her personal resources.

Solutions?

From the HBR article, we can see 3 practical approaches (there are some impractical ones that we discuss later):

Redistribute the work

Any effort to increase your organization’s collaborative efficiency should start with an understanding of the existing supply and demand. Employee surveys, electronic communications tracking, and internal systems such as 360-degree feedback and CRM programs can provide valuable data on the volume, type, origin, and destination of requests, as can more in-depth network analyses and tools.

Also, can one shift decision rights to more-appropriate people in the network? It may seem obvious that support staff or lower-level managers should be authorized to approve small capital expenditures, travel, and some HR activities, but in many organizations they aren’t. (Risk here though is too often responsibility is pushed down, but authority is not)

Structure for Collaboration Boundaries

Show the most active and overburdened helpers how to filter and prioritize requests; give them permission to say no (or to allocate only half the time requested); and encourage them to make an introduction to someone else when the request doesn’t draw on their own unique contributions. The latest version of the team-collaboration software Basecamp now offers a notification “snooze button” that encourages employees to set stronger boundaries around their incoming information flow. It’s also worth suggesting that when they do invest personal resources, it be in value-added activities that they find energizing rather than exhausting.

Also consider whether you can create a buffer against demands for collaboration. Many hospitals now assign each unit or floor a nurse preceptor, who has no patient care responsibilities and is therefore available to respond to requests as they emerge. The result, according to research that one of us (Adam Grant) conducted with David Hofmann and Zhike Lei, is fewer bottlenecks and quicker connections between nurses and the right experts. Other types of organizations might also benefit from designating “utility players”—which could lessen demand for the busiest employees

Measure and Reward the right things

We typically see an overlap of only about 50% between the top collaborative contributors in an organization and those employees deemed to be the top performers. As we’ve explained, many helpers underperform because they’re overwhelmed; that’s why managers should aim to redistribute work. But we also find that roughly 20% of organizational “stars” don’t help; they hit their numbers (and earn kudos for it) but don’t amplify the success of their colleagues. In these cases, as the former Goldman Sachs and GE chief learning officer Steve Kerr once wrote, leaders are hoping for A (collaboration) while rewarding B (individual achievement)

The Economist article is a bit more sceptical about how easy this is, however:

…collaboration is much easier to measure than “deep work”: any fool can record how many people post messages on Slack or speak up in meetings, whereas it can take years to discover whether somebody who is sitting alone in an office is producing a breakthrough or twiddling his thumbs.

Also, Managers “feel obliged to be seen to manage: left to their own devices they automatically fill everybody’s days with meetings and memos rather than letting them get on with their work”

The HBR article also mention two things which (to our minds at least) will not help a lot unless one is very careful:

(Don’t) Use more Collaboration Technology

HBR recommends using technology such as Slack and Salesforce.com’s Chatter, with their open discussion threads on various work topics; and Syndio and VoloMetrix , which help individuals assess networks and make informed decisions about collaborative activities. This seems like a retrograde step if the core problems noted above are not solved. Technologys is largely the reason for the problems to begin with.

(Don’t) Try new fangled Office Productivity schemes

As The Economist points out, the cult of collaboration has reached its apogee in the very arena where the value of uninterrupted concentration is at its height: knowledge work. Open-plan offices have become near-ubiquitous in knowledge-intensive companies. The HBR (and many organisational theorists) likes all this open plan, water cooler stuff. The Economist comments section punctures this however:

In my experience, open plan offices work well where there is a small team working on the same issue. Other than that they are a Dilbert zoo of oppression.

And….

[This] may explain another phenomena of the knowledge-intensive business: the rise of telecommuting. If you don’t go in to the open-plan office, you can actually concentrate on getting the job done.

And…

Well, the newest hype is flexible seating within open plan offices, assigning one working place to about 1.2 employees – strictly to foster collaboration, not as a cost cutting exercise. But indeed, this reduces the “collaboration curse” – because you do not know the people around you and thus have nothing to collaborate upon. So we are already further: we now combine the disadvantages of open plan offices without getting any of the promised benefits of collaboration…

And, cynically

“Collaborative” workspaces typically have a much higher density of employees per square meter. It’s cheaper to house employees and the company saves money. Collaboration is the cutesy label used to justify it to the employee base but the bottom line is that it’s cost driven.

 

What Works, What Doesn’t, What’s Next?

What works

• Collaboration – in moderation
• Careful curation of collaboration systems to optimise performance.
• Performance measure that encourage effective collaboration. This does not mean pushing people to be “more collaborative” as so many metrics today do
• Boundaries to protect the most collaborative people t keep their time sink to effective levels

What Doesn’t

• Collaboration unbound (just reproduces the email inbox problem, without email’s search and file tools)
• Use of new techniques and technologies without a lot of due care and thought about scale, scope and structure.
• Design and Office/Company organisation “fads” to foster Collaboration (especially if the real reason is cost reduction etc)
• Mathematically – too many people in the collaborative net, ensuring that so much tme is spent communicating rather than creating or working

What’s Next

• Increasing focus on monitoring internal collaboration systems
• Increased attempts at measuring and rewarding effective collaboration
• Reducing size/scope of collaboration networks as better metrics winnow down ineffective collaboration approaches- silos are dead, all hail the new silos
• Many collaboration system failures as the above optimisation requires restructuring to give people authority as well as responsibility, managers to cede control, and collaboration to be curated for impact (hard), not as a good in itself (easier)