Liberating Structures

Liberating Structures is a toolbox with over 40 moderation formats that ensure that everyone involved can contribute their ideas. Formats such as Troika Consulting for risk analysis and understanding requirements or TRIS for uncovering possible error scenarios are suitable for testers. Entry points are liberatingstructures.com, liberatingstructures.de and local user groups.

Key Takeaways

• Liberating Structures solves the hippopotamus problem: If the highest-paid person in the room speaks first, everyone nods and valuable ideas from the team are lost.
• In Troika Consulting, the person asking the question literally turns around so that the other two can discuss their problem without non-verbal interruption, forcing structured listening.
• Risk-based testing often fails in practice because Excel risk lists disappear in the network drive, while test-related risks such as unstable test environments do not appear in them at all.
• Liberating structures only work if there is a concrete problem behind them: trying out methods for the sake of trying them out is just as useless as retrospectives whose results nobody follows.

What Liberating Structures are

Liberating Structures are a toolbox for forms of moderation that aim to involve all participants in a discussion. The starting point is a simple assumption: everyone in the room has good ideas. What often prevents these ideas is the usual meeting culture, not a lack of substance.

Traditional moderation works by one person standing at the front and calling on individuals. Liberating Structures take this principle a step further. Instead of leaving the steering to a facilitator, they provide the group with a structure that forces participation. Hence the name: structured liberation.

The typical workshop effect that these methods avoid is well known. Half of the group sits around and says nothing. When the manager says something, all heads turn in their direction. There is a name for this, the HiPPO effect: highest paid person’s opinion. Everyone nods, the boss is right. It is precisely this trap that Liberating Structures are supposed to lead us out of.

Christian Kram uses these methods in his day-to-day work at a self-organized company without traditional bosses. Where no one ultimately decides by hierarchy, a communication framework is needed to bring the group to a targeted decision. This saves three quarters of an hour of talking around the actual question.

1-2-4-All: from the individual to the group

1-2-4-All is one of the core structures and a good way to start because the principle is in the name. First, everyone thinks alone for a minute. Then groups of two exchange ideas for two minutes. Two groups of two become a group of four, and at the end the whole group comes together.

The appeal lies in this gradual consolidation. Sitting out doesn’t work because everyone formulates their own thoughts at the beginning. Nobody gets through without having contributed.

In addition, something happens when the thoughts are brought together that does not happen in purely individual work. New ideas arise when two old ones meet. In groups of two or four, solutions emerge that no one had previously thought of on their own. Christian Kram uses the format when new topics are introduced, for example to gather expectations for new office space.

Troika Consulting: Sharpening test cases by listening

Troika Consulting collects structured feedback on a problem of its own, and the highlight is enforced silence. Three people form a group. One person asks a question, the others give advice, and the person asking the question is temporarily not allowed to speak.

The process is clearly timed. First, the person asking the question presents their problem in about one minute, for example: “How do I understand user story XY?” or “What else do we need to pay attention to?” Then the other two ask clarifying questions. The person asking the question then turns around and listens while the two discuss the problem behind their back and outline solutions. They are only allowed to give feedback after this timebox.

The format is suitable for testers to better penetrate the test basis and question requirements before writing test cases. Instead of formulating test cases alone in the room, you get impulses early on and check your own understanding with two other people. In fifteen minutes, you will have some initial, useful feedback.

Turning around is more than just symbolism. It takes the non-verbal component out of the game. The two advisors cannot see whether you are frowning and you cannot intervene immediately. Many people find it difficult to listen, especially when it comes to their own topic. This is exactly what the method trains.

Uncovering risks without Excel wallpaper

TRIZ uncovers risks by the group deliberately exaggerating and playing out the worst-case scenario. The key question is: What could go really wrong? Exaggeration is expressly permitted here.

One of the guiding principles of Liberating Structures is to tear up existing structures and make progress by failing forward. TRIZ does this in a playful way. In one team, the provocative question about a volcanic eruption brought one participant to a real point: the data center was located directly on the Rhine, and there had already been problems there.

After the exaggeration comes the sober comparison. Which of these bad scenarios have we actually encountered? The first major risk factors and a rough direction for planning emerge from this.

This hits a sore point in many test strategies. Risk-oriented testing is often propagated, but there is often a lack of accessibility. A risk analysis ends up as Excel wallpaper on a shared drive that is rarely looked at. The global risks are then listed there. The banal, local risks are missing: an unstable test environment, a narrow time window for testing. A dedicated risk manager doesn’t think about these.

The person who wears the shoe knows best where it pinches. Those who are affected afterwards usually know best where risks really lurk.

Christian Kram

Impromptu networking: building a shared understanding of quality

Impromptu Networking quickly creates different perspectives on a question by discussing the same question in alternating pairs. Everyone first thinks briefly about a topic, for example: “What is my understanding of quality?” or “What do we need to do to test this user story properly?”

This is followed by short timeboxes of two to four minutes in which two people exchange ideas. Then the partner changes again and again, iteratively. In this way, you gather the views of several colleagues one after the other.

The difference to a round robin is tangible. If you talk in a circle one after the other, you switch off by the fourth or fifth person at the latest. In a two-person discussion, the communication behavior remains different and you are involved. Christian Kram uses this format particularly with groups that don’t know each other yet, for example to ask why everyone is here in the first place.

This fits in with today’s agile environment, where quality is no longer just a matter for testers. Developers, product owners and UX designers work together on quality. An aligned understanding at the beginning is the basis for this.

Methods need a problem, not an end in itself

Liberating structures only work if they solve a specific problem. Using a method because it seems new and hip is useless. As soon as a team realizes that a real question is being answered, the skepticism will quickly fade.

Reservations are normal, especially in larger companies with established employees. “Now we’re doing a sit-down circle, what does he want from us?” Christian Kram goes ahead with such formats anyway, even at C-level, because Liberating Structures can be experienced rather than explained. Those who have participated once are often convinced.

The key is to have a concrete goal in the room: What should our test strategy look like? What obstacles do we expect to encounter in the next test phase? This means that those involved immediately realize that results are coming out. This does not only apply to these methods. Doing exploratory testing just to be able to post it afterwards is just as ineffective.

Many people know the same trap from retrospectives. They are carried out dutifully, then the results end up as an Excel spreadsheet on the drive or as a piece of paper in a drawer, and nothing happens. The next time, people rightly ask what the point is. Ideally, a structure ends with an idea or a direction that is then followed up. Without this consistency, every method has no consequences.

This is how the methods come together to form a process

The individual structures can be combined to form a sequence, a so-called string. Different methods pursue different goals, and this is exactly what makes the combination possible.

Some structures reveal things. Others help to scrutinize something. Still others are used to define a strategy or derive measures. A typical string leads from uncovering to questioning to concrete measures.

This rough classification helps with the selection. The following overview classifies the methods mentioned according to their purpose.

Where to start

The easiest way to get started is via the websites and local user groups. There is liberatingstructures.com online and liberatingstructures.de in German-speaking countries. There are now over 40 structures including instructions.

A selection tool on the German-language site that asks about the problem to be solved and suggests suitable structures is helpful. For those who find websites too old-fashioned: There is also a Liberating Structures app for Android and iOS, whereby the Android version is considered to be decent.

The most emphatic recommendation is for user groups and meetups. They exist in Hamburg, Frankfurt and East Westphalia, among other places, as well as groups such as The Liberators, which regularly organize workshops. You can try out the methods there without having to deliver a direct result in a professional context. This is exactly how the enthusiasm begins for many: go there once in the evening, join in and go out hooked.