Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Articles Why You Might Need an Island of Agility

Why You Might Need an Island of Agility

Key Takeaways

  • Organizational agility is highly dependent on how well the environment is cultivated for agility to grow and sustain.
  • Pathways to organizational agility are numerous, some evolutionary, some revolutionary.
  • Forming a culture bubble of agility is an evolutionary approach.
  • Following an isolation pattern is a temporary approach.
  • Creating an island of agility is a revolutionary and positively disruptive approach.
  • Start slowly. Expand carefully. Improve carefully.

It’s Thursday evening. There is a nice buzz in the Standup winebar off Liverpool street in London, England. “It’s so difficult to make any change in my workplace, “ Emily groaned, “every little thing we try has massive obstacles, and sometimes it feels like we’re going backward. I cannot even send out a survey because of rules. Even though we have the capability, I’m not allowed to collect flow metrics. Even my own transformation team doesn’t get the time to answer the test surveys and interview templates that I was hoping would help us assess the quality and quantity of executive agile behavior.” Emily tells Vanesa she feels like she’s in quicksand, and the more Emily struggles the more she feels like she sinks.

"What's more," replies Vanesa, "in my workplace, it's quite urgent that we finish our re-platforming. It's like everything other than the re-platforming is on hold. We have flow metrics coming out of our ears in my area, but the higher-ups are saying we don't have actionable metrics. We have a dashboard based on leadership, behavior, people dynamics, learning, how it feels, value validation, flow across the organization, and some countermeasures. We've been tracking this stuff for over a year, plus we have a narrative in lay-person language explaining each chart, like, for example, what's this, what is that, what pattern should we see, what's wrong/right here, why does it matter, and what can we do about it. Like, what the hell! People just aren't looking at them. It's like transparency is inconvenient. The latest I heard is the higher-ups want to use a different tool, and they're likely to lose all of the data collected. How convenient! I even heard a rumor about a big consulting firm coming to impose team-based agility on us. It’s so messed up, but at least they can see agility needs to work across value streams, not just within departments. It’s like they need to start over." 

Vanesa continues, “all people want to see is velocity and planned versus actual, even though none of the current work is comparable with work from the past.”

On the way to the bar to buy a round of drinks, upon overhearing, Lara asks Vanesa, “didn’t the pandemic change customer behavior? With all this re-platforming you mention, how has the organization managed to respond?” As Vanesa pauses for thought, Lara heads on to the bar. But Emily cuts straight in, “not only has digital-first made a difference, with everything else that is going on, but sustainability is also big on the consumer agenda, and all I see is green-washing. The consumer expects zero impact on the planet from consumption for no extra cost. And where I work we’re still producing shampoo in plastic bottles. I’m not convinced about the effectiveness of recycling; we need to avoid plastic.” 

Organizational change doesn’t happen overnight, but that doesn’t mean improving agility is impossible. Regardless of the agile approach, by creating an island of agility, we can set a course to agility while the rest of the organization catches up. The key to success is avoiding an island too small to have an impact, having a plan to grow the island, and adding islands to keep momentum.  

This article builds on ideas from others, namely culture bubble from Michael Sahota, isolation pattern from Heidi Helfand, and parallel organization from Craig Larman and Bas Vodde.

Vanesa remarks, “We have inter-dependencies everywhere. Sometimes I feel the organization needs to get metaphorically blown up. Maybe if it stays on this track, it will eventually implode from disruption.” Adam and William help Lara bring the drinks back from the bar. “What if we could create a brand new organization and have a door-person preventing set-in-their-old-ways folks from getting into the building? Only folks thinking in agility terms would get in”, Adam smirks, looking unsubtly at the no-nonsense unsmiling doorman nearby. “Do you mean Agile?” queries Emily in a louder voice as the crowd and music volume rise.

What is Agile?

Agile is referred to in the Manifesto for Agile Software Development, also known as the Agile Manifesto, while agile refers to the word in the (Collins) English dictionary). Organizations don’t need to do Agile to be agile. 

Simon Powers’ famous Agile Onion diagram sums up Agile for many people, easy to understand, challenging to do, and even more challenging to be. For those who look to the Agile Manifesto for insight, Agile is about values and principles. Others conflate Agile with specific methods, methodologies, frameworks, and strategies or see it as an umbrella for all of them. Regardless of how you approach it, without a fundamental shift in how an organization thinks and operates, a ceiling limits the ability to “be Agile.” 

What organizational agility looks like

An organization needs to foster an environment where agility can grow and persist. You can’t delegate agility; it’s a change for the whole organization to varying degrees.

Agility is not a team sport; it’s a company sport. - Klaus Leopold

Agility is about healthy interactions and moving together in a chosen direction while adapting when necessary. The prevalence of complex work makes a strong case for organizational agility. For complex work, what is unknown is often greater than what is known, so reliance on experts alone is risky; fresh, diverse thinking is often needed. 

It might be helpful to think of the components of organizational agility as a formula (although not literally).


Think of organizational agility as being about:

  • Dealing with the Situation at hand.  
  • Optics
  • Using guidance from a complexity Compass. We don’t need agility for everything, yet we need people who can spot emergence, capture it, and build upon it. Cynefin® is a valuable framework to guide desired behavior. 
  • Inspired, intentional, diverse, authentic, psychologically safe, trustful, and compassionate Interactions
  • Human-centric Adaptiveness that is informed by the customer, analytics, flow, and learning.
  • A little bit of Luck

Organizational agility is unsustainable without Fostering the right environment that includes a combination of:

  • Focusing on people, including deep listening to what frustrates them.
  • Discovering value.
  • Delivering value.
  • Decluttering. 
  • Organizing for value.
  • Changing the ecosystem of decision-making.
  • Decoupling leadership from the position of leader, as leadership is everywhere.

Organizational agility: what’s the urgency?

Making a transition to agility is difficult for large organizations because they often have culturally entrenched systems. How can we affect lasting agile change when cultural influences are slowing it to a snail’s pace or stopping it in its tracks? 

Some organizations want everything perfectly aligned first before making a linear transformation. But, authentic organic change doesn’t happen that way. 

As Leandro Herrero suggests revolutionaries don’t wait for everyone to be aligned before they begin their revolution. 

The reality is that imminent pressures make change urgent. For example:

  • Startups experience the stress of limited funding. 
  • A disrupted public company with massive debts after a market shock (such as war, pandemic, or stock market crash) struggles for market share and a better share price and must placate investors and lenders. 
  • A fast-moving consumer goods company (FMCG, e.g., shampoo) has not reinvented its approach to “pass on plastic.”
  • A slow-moving consumer goods company (e.g., automotive) has not reduced the carbon footprint while maintaining existing prices even though consumers demand it and are getting that elsewhere.
  • A failed “transformation initiative” drives unspoken demand for superficial change to at least show some “progress,” “because real change is too hard.”
  • A feature factory has no clue how, if at all, it is changing the behavior of customers or end-users in a positive or negative way. The word discovery is not well understood. The word experiment often triggers the thought of a lab rat rather than simply being a learning exercise. 

Sometimes there is so much delivery pressure in the system that execution bias dominates, resulting in the following: 


  • Accept delivery “carry-over” as normal.
  • Get insufficient time to discover the >50% of ideas that should never get built.
  • Don’t get time for learning necessary skills. 

The organization :

  • Tries to buy “agility in a box.”
  • Accumulates more and more “quick fixes,” where the cure is worse than the disease.
  • Loses customers without acquiring new customers.
  • Loses market share.

Sometimes, an organization needs a jumpstart to address these pressures and overcome the inertia to change. Dare we say it, sometimes we need to start a new company. The critical question is - is there a sense of urgency to do such a jumpstart?

Emily retorts, “that sounds interesting in theory, but wouldn’t the new organization get tied up in knots by the same complicatedness?” “True,” remarks Adam, “but have you heard of the strangler pattern? Martin Fowler made it a famous pattern. You can wrap your old systems and strangle them slowly but drive more traffic to new systems while always reconciling with the old systems until all traffic is moved across.” “Will it strangle my inbox too?” Vanesa sniggers.

“What about a culture bubble?” says Lara, in between sips. “Don’t bubbles burst?” says Adam dismissively and immediately corrects himself. “OK, let me be more positive; sometimes bubbles can last a while. Heck, I saw massive ones created by a street artist over at the riverside by London Bridge yesterday”.

It’s getting noisy in the bar. “I’m curious what the difference is between a culture bubble and isolation pattern” shouts William over the loud dance music. Good question…

Could a culture bubble be the solution?

Agile culture and leadership trainer and consultant Michael Sahota wrote about building culture bubbles to aid Agile adoption in 2013. A culture bubble is a pocket of an organization in which Agile can flourish temporarily to achieve a specific goal.  It’s an idea that has merit, but just as bubbles do, this concept eventually bursts due to its scale.

Culture bubbles are people-centric and culture-centric. They are not process-centric, e.g., a SAFe Agile Release Train is not a culture bubble.

A culture bubble often still depends on the rest of the organization to get work done and relies on “adapters” to interface with the legacy organization to avoid an “antibody attack.” Further, without a plan to grow a bubble quickly, it’s starved of the oxygen it needs to live and thrive. 

Michael Sahota updated his guidance in 2018 and added culture bubble content to the 2021 paperback, Leading Beyond Change: A Practical Guide to Evolving Business Agility, that he co-authored with Audree Tara Sahota. It appears that Michael Sahota’s position on culture bubbles has softened from a temporary to a more long-term construct. A culture bubble develops metaphorically on a “cordoned off” or “reclaimed land” section of the mainland.

Could an isolation pattern help?

Heidi Helfand, the author of Dynamic Reteaming, writes about the isolation pattern idea for dealing with a crisis. People who demonstrate entrepreneurial behavior are invited by the leading change agent(s). Teams can choose how they work, the isolation team gets its own space, and other groups are told to leave the isolation team alone.

Examples of a crisis include needing to pivot the organization from failure, new product development, the discovery of innovations, and solving technical emergencies.

It's about forming a team to the side, freeing them of the organization's bureaucracy, and metaphorically having a sailboat sailing freely in the seas off the mainland. Visualization and synchronization patterns can help scale the idea, resulting in an organizational design emerging within the isolation of the flotilla of sailboats, if you will. 

The model is not without potential pitfalls, including infusing those within the flotilla with a sense of elitism. It can also limit "dogfooding." Finally, the folks working on the island might experience disappointment when they eventually re-merge with the mainland.

Helfand illustrates how the isolation pattern helped several organizations through several case examples in her book.  

“OK, they seem like good ideas, but what about this agile island idea?” asks Lara. “No, it’s an island of agility. Agile is dead. Long live agility”, says Emily in a sarcastic tone.

How about an island of agility? 

An island of agility is a team of teams working independently of the rest of the organization to the extent it can ignore guidance or direction from the c-suite. An island of agility is not a function/department operating agility in a silo. The island's scope needs to be significant enough to get the focus and commitment to succeed. Perhaps it's working on the organization's second or third most urgent thing. Maybe it's working on a pivot or new product development. In this case, it would not be entangled with current product development.

An island of agility exists permanently in a sea of less agile teams. It successfully demonstrates authentic agility to the point that it infects the waters and influences other teams, to take the voyage.

While changing an organization is difficult, affecting change among suppliers is virtually impossible. Teams on an island of agility extend membership to others within the organization and external suppliers as long as their managers do not interfere or have influence.

On an island of agility, we aim for:

  • Invitation and informed consent.
  • One owner of the product/service. 
  • People who eagerly volunteer rather than being “voluntold.”
  • Teams that self-design and devise effective ways of communication. 
  • People who pair, swarm, and mob on team tasks when it makes sense to get the work done.
  • Freedom for teams to use any approach that demonstrates authentic agility and the latest technical practices. 
  • Horizontal growth over hierarchical promotion.
  • Simplification over assumptions that scaling is always the best way forward.
  • Executives who ask questions that promote agility.
  • Design that involves the customer using a discover-to-deliver approach.
  • More opportunities for team members to work face-to-face with each other, customers, and end-users.
  • Respect for ways of working of other teams.
  • Focus on the risks we’re trying to manage and how to demonstrate compliance in more practical ways.

On an island of agility, we aim to eliminate:

  • Organizational entrenchment, e.g., organization designed for ease of management, not discovery/delivery of value.
  • An organizational divide, e.g., business and IT, marketing and supply chain.
  • Siloed organizational design that results in handovers and dependencies; e.g., handovers are not really handovers, resulting in “hurry up and wait.”
  • “Helping” people who did not ask for help as in a top-down “do it this way” type of change.
  • Co-ordination chaos, e.g., WaterScrumFall.
  • Tools that are not fit for purpose.
  • Project organization.
  • Project management office.

“I don’t want to get my hopes up. It seems like a pipe dream to me, but wouldn’t it be so wonderful! We seem to solve all the world’s problems over drinks. Let’s talk tomorrow morning again over coffee, and tease out this idea.” suggests Vanesa.

Islands of agility aren’t a tropical dream; they exist

The island of agility concept is based on the LeSS Guide of Parallel Organization  (3rd LeSS book, by Craig Larman and Bas Vodde, page 74), which features case studies, including the huge LeSS Huge at BMW Group — Autonomous Driving and JP Morgan Chase

Stopping at one island won't allow agility to gain traction organizationally – you will need to build momentum using an archipelago of agility. Over time, the archipelago can take on more and more work from the mainland.

Adopting authentic agility with large-scale frameworks such as LeSS Huge, huge LeSS Huge, and Flight Levels can help avoid the islands' isolation and foster collaboration across products and value streams. 

As a bridge to the archipelago, other teams can proceed with experiments and a slower evolutionary process towards agility. The archipelago can quickly gain the advantages of agility while supporting the larger organization to make the cultural shift necessary for wholesale change. This is one approach toward broad and deep agility, which you can learn more about in this discussion: See "broad and deep agility."

Over coffee the following morning, concerns grow that the islands and archipelago would reconnect with the mainland and that they would influence each other negatively. Adam starts, "the mainland is optimized for efficiency of the current product set, and the archipelago would be optimized for adaptiveness and effectiveness of potential value validation. We'd want some of the mainland to stay as is, and we'd not want the archipelago to get polluted with efficiency tunnel vision." Emily agrees and continues, "yeah, a friend of mine witnessed unraveling after a leadership change. I wonder how the archipelago could avoid this unraveling. It would be such a waste."

How to avoid the culture unraveling in the archipelago of agility

While it might be intuitive for an island of agility to reattach with the mainland organization once it has achieved a level of maturity, doing so risks regressing due to the dominant organizational culture on the mainland. Instead, work should slowly migrate from the mainland to the archipelago of agility, but not necessarily all work.

Lara has an epiphany “that seems like a strangler pattern for the organization, except the mainland would not get entirely strangled. Some things don’t need to be agile.” “Spot on, it must be good coffee,” William jokes.

In a context where the mainland does not have an environment where agility can grow, there are several reasons it can pollute the culture of the archipelago:

  • Measurement
    • The mainland is more likely to use measures that hinder agility and drive even the best people to behave in counterproductive ways. As the author of The Goal, Eli Goldratt, said, “Tell me how you measure me, and I’ll tell you how I behave.”
  • NoNos - people who will always say no regardless of the suggestion
    • The archipelago of agility filters out NoNos through volunteering and transparency. 
    • The mainland’s tolerance for objections to agility almost always allows uninspired and uninspiring leadership to flourish.
  • Lip service on the mainland - saying one thing and knowingly doing something else
    • Lip service undercuts change.
    • Sincere, constructive vocal resistance is always better than lip service.
    • Lip service threatens authenticity, humanity, leadership, future revenues, and engagement. 
  • Organization Design
    • The mainland is designed for ease of management.
    • The archipelago is designed to ease iterative and incremental problem space work, discovery, delivery, and value validation. It also needs to “drink its own champagne” and do operations work.
    • The mainland relies on coordinators. The archipelago relies on self-management.
  • Competitive culture
    • The mainland continues to promote internal competition over collaboration.
    • Individuals compete to be in the top right corner of a matrix grid on the mainland.
    • The more things change, the more they stay the same.
  • The mainland’s sincere opposition to the change slows down its pace.
    • A publicly-traded company is less likely to leap to authentic agility – there are quarterly results to deliver and expectations to meet.

“OK, how do we start,” Emily, Lara, Vanesa, and Adam say in the exact same second. They are inspired.

Where to start?

Agree on an inspiring purpose

  • Consider an elevator pitch or business problem statement when agreeing on a purpose.
  • Consider societal value.
  • Consider an almost-impossible product/service/process perfection vision supported by a direction of travel with multiple time horizons, but embrace the gathering of evidence for data-informed decision making.
  • Ideally, the purpose is co-created and continually adapted, so everyone on the island is inspiring and inspired.

Educate people who demonstrate curiosity

  • Seek an executive who lives and breathes authentic agility, who talks last, and gets people’s attention in a positive way
  • The executive must be coachable and light on ego – they will demonstrate ↑authenticity, ↑self-awareness, ↑courage, ↑go-see, ↑speaking-truth-to-power, ↑alignment of what they say privately and publicly, ↓backlogs, ↑prioritization, ↑bottleneck-awareness, ↑focus, ↑flow, ↑embracing uncertainty, ↑serving teams, ↑coaching at all levels, ↑fixing problems beyond the influence of teams, ↑enabling organization design towards the direction of travel.
  • Consider the Vanguard Method, Nexus, LeSS, and Flight Levels, or another authentic approach, even if it means going back to values and principles but catering to people who need some rules.
  • Live and breathe a “discover-to-deliver” approach (hat-tip to the book “Discover to Deliver: Agile Product Planning and Analysis“ by Ellen Gottesdiener Mary Gorman). Avoid feature factories and other pathologies.
  • Get the most credible change agents. Filter out anyone encouraging “change theatre” or with detrimental hidden agendas. Prune change agents who claim they did X number of transformations because we can never achieve agility in perfection; we’re never done.
  • Embrace change agents who act based on values and principles. Embrace change agents who can catalyze authentic agility.
  • Encourage change agent humility - they need to learn the business domain fast. 
  • Look for curious volunteers with a can-do attitude and avoid rockstars. Everyone will pair, swarm, or mob to get the job done.
  • Educate volunteers on authentic approaches to agility.
  • Let volunteers decide whether they want in or out once they’re educated - it’s about informed content. 
  • Use self-designing team workshops to reduce asynchronous dependencies, empower people, and turn the power pyramid upside down.
  • Design a team alliance or working agreement to agree on norms.
  • Publish a team API so people know how to interact with each other effectively. 

Demonstrate systemic agility

  • Do not set up the island/archipelago as a supplier to the mainland - it’s the wrong relationship.
  • Implement a horizontal growth path - recognize people for the number of deep skills they are practicing and for which they can coach others with high-rated peer satisfaction.
  • Implement shorter funding cycles than the mainland (mentioned by Joe Justice on the Xagility podcast) - consider Beyond Budgeting or Throughput Accounting.
  • Consider Obeya in a humane way (mentioned by Jim Benson on the Xagility podcast)
  • Research Indi Young’s work on deep listening to understand the problem space.
  • Implement the latest technical practices; get skilled coaches to go from team to team for months at a time.
  • If you must measure, measure the how not the what (Yves Morieux Ted Talk) towards the purpose (inspiration from John Seddon). For example, consider the “switching cost” (inspiration from Craig Larman) that comes without having to cancel/hold other work when changing from one idea to another. Avoid making comparisons. Avoid imposition. Avoid singling people out. Avoid singling teams out.

Improve team member experience

  • Ensure no double-jobbing between the mainland and an island. Fence off the island.
  • Strive for stable teams or embrace dynamic reteaming.
  • Foster OpenSpace Technology and Liberating Structures and the meeting spaces to support them (no boring meeting rooms).
  • Implement soundproofing and audio filters.
  • Implement slack time regarding it as normal.
  • Design with the customer.
  • Use flow and probabilistic forecasting to reset expectations regularly.
  • Listen to the people - consider SenseMaker to pick up the mood.
  • Continual education, coaching, and walking around for the executive.
  • Encourage constructive dissent – monitor closely for after-effects.

Expand carefully

  • Expanding too quickly leads to shallow change and inadvertently taking on NoNos.
  • Maintain the same careful onboard approach for every batch of new people who join.
  • Consider flight levels to glue the islands of the archipelago together.
  • Keep the organization flat.
  • Keep one P&L statement.

Improve regularly

  • There is no change team. There are (perhaps rotating) team/group representatives who agree and deliver improvements. Managers, if they exist, help to improve the system.
  • Improvements come from retrospectives and continual marginal improvements. To avoid “should” conversations, consider including improvement delivery within the retrospectives.
  • Consider LeSS Huge or “huge LeSS Huge” (it’s a thing) to build improvement into the rhythm.
  • Develop and improve an “anti-handbook handbook” that is short and balanced.

“Hmmm, that looks hard”, says Adam. Lara, Vanesa, and Emily smiled at Adam. It’s a telepathic smile. Adam continues, “ok, I get it; authentic, sustainable agility can’t be bought in a box. It’s an approach; there are other approaches. But this one seems to offer hope for organizations struggling to adapt quickly enough.”


Attainment of authentic, sustainable agility is not for the faint-hearted. Even the best, most enthusiastic people struggle in a flawed system. Sometimes the system is beyond repair. But a sense of urgency is needed. If you experience a crisis, don’t waste it. Try an island of agility.

About the Author

Rate this Article