Building effective teams in an age of difference

The other day I went to a very hands-on event about data. I didn’t expect it to be this way and I was out of my depth, scrabbling to keep up as my group undertook our main task: inputting on guidelines for data quality assurance.

As a UX and product person I had to fight the urge to check if I’d joined the wrong event. I stuck with it, re-reading the data guidelines over and over, and asking the occasional softball questions. 

Thanks for reading Without Constraints! Subscribe for free to receive new posts and support my work.

Did I waste my time? No. Not for the group and not for the room. Because of the Wisdom of Crowds.

Yes, I mention The Wisdom of Crowds by James Surowiecki. Often misread by your common-or-garden tech bro as involving throwing decision-making wide open, chucking things out on socials and seeing what the masses think, the book actually argues for something quite different.

What it says is: groups tend to make better and more accurate decisions than single experts.

And not just random groups made up of random people but groups made up of a diverse range of informed individuals.

Why diversity matters for teams

Diverse here doesn’t necessarily mean in the demographic sense, but rather that groups (and teams) should include people with different knowledge and insights. They should be diverse in terms of perspective and experience. They need to be diverse based upon their members’ specialism and access to information.

Time and time again, studies prove that the predictions and results of these informed and diverse groups outperform the predictions and results of solitary, highly specialised “experts”.

I won’t get into the studies that the author covers to demonstrate this, purely because their book does such a great job of explaining them. But I do love the book’s comparison of NASA today with the newly formed NASA of yester-year, whose new teams were so highly effective at solving terrifying problems (Apollo 13 anyone?) precisely because they were drawn from so many different and diverse walks of life.

But anyway. 

How did my presence, as a UX nerd, help the group at this event? And why does the diverse group perform better?

And what can we take from all this to designing products, building teams and running businesses?

1. If you build diverse teams they’ll be better able to consider and act on information

If a group (or team!) is made up of people whose experiences and access to information are all too similar (i.e. the team is quite homogeneous), then when evidence arises that goes against the group’s expectations, it is more likely to be discounted. 

With a more diversely built team, however, it’s more likely to be considered and factored into decision making. This in turn tends to lead to more accurate decision making.

So, business leaders, build your teams to include a range of skills and experience. Build teams that can bring different information to a challenge.

And take extra effort to do this particularly when it seems redundant to do so, when it seems ‘obvious'  that the work should belong in the remit of one particular skillset or job role. Because homogenous teams can do some very weird and risky things, as we’ll see next…

2. If you build diverse teams it will reduce the exposure to risk

This one surprised me.

Basically, homogeneous teams actually get more extreme in their decision making and predictions. Discussion itself amongst homogenous groups actually pushes the group towards adopting more extreme positions than the individuals within it would usually hold.

In other words, discussion in very same-y teams amplifies risk, as the group decides upon more risky courses of action. And this isn’t healthy for the business, the group, or the individuals within it.

And it’s not because of any sinister or selfish reasons, but rather because of the way that people subconsciously (and naturally) try to maintain their relative position amongst the people they’re surrounded by. For example, if an individual believes themselves to be more middle-of-the-road in their outlook on a particular issue, when the discussion in a homogenous group drifts toward taking on a more extreme position, that individual is more likely to shift the positions they support too, merely in order to retain their relative social position. They accept far more than what they’d usually view as acceptable were they not in this group.

And they do this just so long as they can still retain the feeling that they sit in the ‘middle’ of the group’s position. 

However by including a more diversely informed group of people, you create a buffer: the chances of this polarisation and the team adopting extreme, risky decision making is reduced.

3. If you build diverse teams, they’ll ask themselves more questions and better handle assumptions

Groups which are overly stuffed with experts from the same discipline don’t ask themselves as many genuinely open questions. Discussions in these groups - even though it still happens - tends towards reconfirming their original assumptions, and participants from minority disciplines feel less included to ask questions. Worse, they defer to the “experts”.

(Which is risky in itself, as research shows that all people have a natural tendency to be overconfident about their abilities. And worse, with experts this overconfidence is even more pronounced.)

Furthermore, homogeneously expert-stuffed teams can see their role as being to reach a decision as quickly and efficiently as possible. This is a mistake. Because by adopting this attitude, decisions tend to get made early - probably based on assumptions and the homogenous group’s past experience - and then any discussion that follows is primarily aimed at getting dissenters to agree.

What should be happening is carefully, structured discussion based on considering alternative evidence, which leads to informed decision making.

So, groups that are diverse and which include a broader range of informed participants are actually more critical and perform better. Perspectives or outside information that would benefit the overall decision making tend to be valued more, whilst natural tendencies to minimise and ignore new evidence are tempered. 

So go you! If you’re part of a group and you feel that you’re not the expert, take heart! Asking questions is the vital role that you need to play in order to help the group remain critical and productive. 

And go me too. My role at the data event, asking low ball questions, had value after all.

Okay. You’ve established a diverse team. What next?

If you’re building a well performing team, you need to ensure there are processes to impartially collect, aggregate and act on information

It’s no use building a diversely informed team if power and responsibility are still centralised within that team.

For example, if final decision making is made by one person in the group then that centralisation of decision making will tend to undo all the gains from having a diverse group. 

“No decision-making system is going to guarantee corporate success…. But the more power you give a single individual in the face of complexity and uncertainty, the more likely it is that bad decisions will be made.”

Or, if the way that the group functions tends to minimise or filter out new and divergent information, then this will also reduce the impact of the group. A classic example of this is when team members are posed questions which are worded in such a way that it’s implied there is a ‘correct’ answer, or that any responses counter to ‘the norm’ will be unwelcome or an effort to consider. (Simple examples being “Are we all agreed that X is the best approach?”, “What are the downsides to not doing Y?” etc)

Yet if there is a mechanism to collect and “aggregate” information without judgement, then the group will be more able to act and make effective decisions.

Simple ways to collect aggregate and act

There are some very common ways to help ensure aggregation, and they’re not magical and they won’t surprise anyone. However in the context of the above hopefully you’ll see why they are especially important.

1. Ensure discussions come first, then lead into decision making steps.
2. Make sure that everyone has a turn to speak.
3. Decisions should be made by voting (blind voting os even better)
4. Aim for consent rather than consensus.

More structural ways to increase collection and aggregation 

Less simple ways that help facilitate aggregation exist too. As an added win they also help build healthy work cultures. The following are some top level ideas - each worth a world of guides and posts on how to implement them - but I hope they get you thinking.

1. Increase decentralisation.

Just as removing decision-making from the hands of single individuals within a team will increase the chances that good decisions are made, so too removing the centralisation of power and decision making from single teams will help harness diversity of thought, bring more information into the decision making process, and protect against bias. In this scenario teams should be able to work independently in a way that avoids duplication of effort, and be trusted with power.

Toyota's Production System, the classic example of Lean manufacturing, whose teams of frontline factory workers can solve problems independently without referring things to higher ups - and even have the power to halt production altogether - is a great example of decentralisation of power and decision making.

2. Increase the flow of information between teams.

But it’s not enough to have more teams doing the same thing. You need to increase the overlap, the interchange of information and evidence between teams. So consider:

• How might your teams better communicate and share information with one another?
• How might they overlap, perhaps, with some members sitting across teams to facilitate the sharing of relevant information?
• And what will count as relevant? How might you ensure that you find the right balance between sharing only that which is relevant and informing, and sharing all the information and thus increasing duplication of work.

3. Reduce hierarchies and status, which restrict the flow of information.

How many people need to be involved vertically for your team to be able to make a decision? How might this be reduced? 

You might not be ready for a flat organisation structure just yet, but there’s still much you can do.

For instance, user-centered design has already identified that people closest to the problem often have the best insights, and so it aims to get end users involved in product design decisions. So when it comes to your teams and your hierarchies:

• How might you remove the distance between top levels and the problem faced by users?
• How might you empower and build trust in project teams so that upper levels of power aren’t needed to sign-off decisions? (Tip: get power holding stakeholders involved at the beginning, give them the equal voting powers as the rest of the team, and get them to approve / back the team’s forthcoming efforts at the beginning. I.e. get them to express their trust for the team to continue. Then their involvement can be tailored off through the project).

And when it comes to product design itself, how might you involve people who work with your customers on the frontline, no matter their official job title? It’s true that the realms of product, UX and design have all the sexy job titles - but don’t the people working hard in customer support or staffing the phone lines speak to countless numbers of users every day? They may not working sexy digital or product roles, but they can channel your users’ voices like no-one else can.

These last questions are all deserving of a deep dive and I wish you well in finding out more. If I see anything particularly relevant in my web travels I’ll add it in the comments.

Good luck building effective, empowered and diverse teams!

Thanks for reading Without Constraints! Subscribe for free to receive new posts and support my work.