Stop T-shaping people

How to get from T-shape to Mastery

We come across a fair amount of worries concerning T-shaping. The to be T-shaped worry that they will be forced to shape into something they don’t want to be. Management, who would like T-shaped people, worry about resistance when T-shaping. We think they might be worrying about the wrong things.


When someone has enough skills and knowledge to be able to handle themselves in other domains than their own we call them T-shaped. In theory, it would make teamwork a lot easier and create a backup for when people are absent. However, most people don't like to just be able to handle themselves, to feel that they are constantly out of their T-shaped depth. It creates sub-par results that are not up to standards. It causes tensions between experts who have to check and correct each other's work. Resulting in management having to interfere more instead of less.

Communities of practice

Communities of practice or Chapters (cross-team structure of similar experts) are usually responsible for supplying the knowledge and skills necessary to enable T-shaping. However, they are often the most sceptical about it. From a CoP point of view, it does not make any sense to lose focus and waste time in forcefully creating generalists. They are right in saying that these generalists can not perform at their level of expertise. In fear of this dilution of quality, they feel the urge to create standards, domain-related definitions and carve out clear boundaries of role responsibilities. By acting like this we often see that instead of facilitating the spread of knowledge and skills CoPs or Chapters hinder it.

We have encountered test Chapters dictating strict definitions of ready, that grind the flow of value to a virtual hold. UX CoP claiming ownership of the user domain, blocking Design Thinking activities with the entire cross-functional team. Ops Chapters turning AWS (cloud computing) into an all singing and dancing starship Enterprise, seriously hampering any DevOps aspirations. These are just some examples of counter-responses to T-shaping.

Typically the people with a leading role in CoPs are the ones with lots of domain expertise and care for their subject matter. Unfortunately, they are often also the ones most opposed to T-shaping we have noticed. Because they fear a lack of quality control on the thing they hold most dear, they often challenge the possibility of sharing responsibilities with non-experts. So even if people are willing to T-shape, CoP leaders and groupthink will soon put an end to that. Anyone not aligning with the collective identity will experience a lot of resistance and might be outcast.

But we still would like to have cross-functional, self-organizing teams that can as independently as possible deliver value to the organization and our customers. If T-shaping people by injecting knowledge and skills is not the solution what can we do?


In Daniel H. Pink’s must-read book Drive, the concept of Mastery is described as the urge to get better and better at something that matters. Why do people spend hours and hours practising on a musical instrument? Why do we lock ourselves in escape rooms, train for marathons or solve Rubick’s cubes? Because we enjoy that feeling of getting good at something that we really want to grasp. We are in the zone once we are really focused on a skill we want to Master and get into a state of flow if things start to come together and continuously improve. Could the Mastery mindset be a better approach than just T-shaping people?

Goldilocks Tasks

If we want to use the Mastery approach we need new tasks at just the right level of challenge. If a task is too easy, people will get bored. If it’s too hard, they get anxious or frustrated. We want that sweet spot, where it’s is just challenging enough that we’re engaged and pushed to a slightly higher level but not so hard that we will give up. Tasks also need to be of the right size if it takes too long to master a craft we will lose motivation. Working at just the right level and size is called a goldilocks task.

Pairing with mastery

By pairing people, with a good mentor, on difficult stretch tasks we could increase the goldilocks factor significantly. The mentor could challenge us to search for the next level solution and would guide us with advice once we get stuck avoiding too much frustration or anxiety. For the mentor, new light might be shed on a common and mundane task by coaching a peer with another skillset and perspective than hers. Where T-shaping is rightfully met with resistance and scepticism, Mastery as a mindset is often welcomed because it respectfully starts from the individual's inner drive. Cops and Chapters usually also see the benefits as it gives them a chance to get their expertise and standards built into the system by pairing up with people that are far more receptive and interested in their message.

Theory X and Theory Y

The management paradigm shifts on the motivation of people as described in the theories created by Douglas McGregor is also essential here. Old school Theory X assumes that people are basically individualistic, lazy and need to be externally motivated to do work by rewards and punishment. Theory Y assumes that people are internally motivated and will work to better themselves and their surroundings and therefore don’t need to be punished or rewarded. Research shows that we get better results if we approach teams from a theory Y perspective instead of micro-managing individuals

Solve problems, don’t T-shape

So theory Y shows us that we genuinely like to work together and help each other to get the job done. Unfortunately, the workload is not always distributed evenly across our team. With the help of mentors and CoPs, we could use the Mastery drive of team members to solve these bottlenecks. At their own pace, driven by their own needs and facilitated by an organisation that can support these ad-hoc demands for mentoring and knowledge transfer. Help people to get better at the skills they need in their team at that moment, just enough and just in time. Don't just force them to gather random skills and knowledge just for the sake of T-shaping.

A possible approach


First of all, we need to find mentors for people that are in need of new skills or knowledge. Not driven by the mentors need to spread their knowledge but based on the need of the novice. For this, we need mentors to be servant leaders, the mindset that foremost leaders exist to serve the people instead of the other way around. To enable this in our corporate system we could introduce an apprenticeship model where we have a natural connection and commitment from mentor to apprentice and vice versa. Position it as an additional support structure and not along reporting lines to create a safe and open environment. Facilitate self-selection ceremonial events where we emphasize servant leadership and create publicly committed bonds. Create mastery triads of apprentices and mentors for diversity, availability and stability. Let multiple triads of mentors and apprentices connect with each other to create an informal embedded network of Mastery. We will need to facilitate this continuously due to organisational change and changing circumstances. People coming and going, new domains being explored, constantly-evolving knowledge and skills needs are but a few reasons why this should be an organically growing and evolving system.

Why a network?

We could organize in traditional CoPs or Chapters but often networks are far more effective in modern companies. As we don't want to optimize for mass T-shaping but do need an ad-hoc availability of Mastery support, the slow and large CoP knowledge transfer process is often not a good fit. Novices are far more in need of more personal connections for a quick chat, a short question or a brief glance over their shoulder. Remote working and co-creation tools have only made this all the more apparent. CoP meetings, although still attended, are often lacking focus and active participation, interaction is minimal if not forced in smaller breakouts with a limited number of people. We often have to resort to a maximum of 3 persons per break-out that we take back and forth to a central room, de facto creating a network on the fly. We don't want to suggest that we don't need Chapters or CoPs anymore, they still play their role in setting standards, solving systemic problems and creating communities among other things. The goldilocks zone needed for Mastery however is better served with an informal stable network.

Why Triads?

Research shows that people don’t like to shut single persons out from a group, they tend to stay together or from subgroups. As triads are undevolveable, if it would break up into smaller parts we would be left with one person on their own, triads are the smallest stable group. This leads to more durable networks as described in Simmel’s hypothesis.

Often decisions are to be made between two opposing opinions. If we have a 50–50 split in a group this leads to stalemates and delays. Due to the inherent power inequality in triads, this is not possible and we see faster decision making, which is a big advantage in our VUCA world. An alternative point of view is often present and because everybody is on an equal footing it is more often expressed which leads to better decisions.

The resilience is higher, if one member is absent the others can still fulfil their function especially if we have a certain redundancy or overlap in knowledge and skills. Finally a controversial point, triads are less intimate than the gossipy duo, which could lead to more focus and drive, in a professional setting this might not be undesirable.


Now that we have our apprentice support system we need a problem to solve to embark on purposeful Mastery. If we work together as a team of experts problems arise when there is an unbalance in the flow of work through the team. Resulting in a pile of work mounting for certain individuals while others will run idle. This often leads to people looking for more work to do, creating even more unbalance. We can use a Kanban to visualize this unbalancedness.

Visualize bottlenecks.

There is usually only one bottleneck in the system. There is a misconception that finding bottlenecks is hard. If you use a Kanban properly like in the video, sooner or later you will experience empty “Ready” columns just after your bottleneck where the work is pilling up. This would be your trigger to start optimizing for flow by helping out the bottleneck.

The Daily

Assuming you are doing something like a Daily standup this is the typical time the team will discover that someone has an empty “Ready” column and no items to work on. The next thing to do is to discuss if we can exploit the bottleneck, by taking any additional work of the plate of the people that are working on the items in the bottleneck column, so that they can have full focus on only these items. By helping people with their additional work we are exposing the rest of the team to new challenges and skills. Just by pairing or even shadowing the bottleneck, team members will pick up a tremendous amount of knowledge and always spot improvement possibilities. This is a massive natural T-shaping opportunity.

Even flow

Next, we should try to even the flow upstream to the pace of the bottleneck, because all additional inventory of items that need to be processed is just waste. Especially in a cross-functional team where the whole team should be involved in upstream activities, creating unwanted distractions for the bottleneck. We should do this by trying to diminish the work needed to be done by the bottleneck, which is typically done by improving the quality of its input ergo the output of upstream activities. Again a T-shape opportunity for all parties involved. So in your Daily, you would have your empty column people ask your upstream people if they can pair on improving the quality of their work.


Now if all this seems counterintuitive, you are right. We have not experienced very many teams that start doing this by themselves. Bothering upstream people to improve their quality or sitting idle while observing the bottleneck is usually way out of peoples comfort zones. However, exactly these kinds of counterintuitive behaviours have been extremely beneficial for both T-shaping and creating flow. What you could do is try these technics out in simulation. Okaloa Flowlab is one way of simulating the importance of WIP limits to create flow. In LEAT®, our framework to experience and learn through safe simulations, we have created a special module that brings out the aspects of flow through a typical cross-functional team as mentioned above. Book a free LEAT mini experience and come and see for yourself.

Natural born Agilista, Lean product thinker, design journey maker.