Scrum has been around for a while, they say. The Scrum Guide holds the definition of Scrum, they say. The first, official version of the Scrum Guide was released in February 2010. So, how was Scrum defined before 2010 then? How did its definition evolve before and after 2010 and become the framework that we know today? What else happened along the road to the way that Scrum is defined and represented?
In the paper “Scrum: A Brief History of a Long-Lived Hype” I have described what changed to the definition and representation of Scrum over time, before and after the creation of the Scrum Guide. It shows how Scrum evolved into the framework that we know today since its first formal introduction in 1995. Because a touch of historical awareness is more than helpful in understanding Scrum and caring for the future of Scrum.
I looked for sources that are not just credible in terms of authorship but also offer regular enough check points. In the end, the sources I used for describing the evolutions of the definition of Scrum are:
The paper “SCRUM Software Development Process” by Ken Schwaber (1995, 1996)
The paper “SCRUM: An extension pattern language for hyperproductive software development” by Mike Beedle, Martine Devos, Yonat Sharon, Ken Schwaber and Jeff Sutherland (1999)
The book “Agile Software Development with Scrum” by Ken Schwaber and Mike Beedle (2002)
The book “Agile Project Management with Scrum” by Ken Schwaber (2004)
The book “The Enterprise and Scrum” by Ken Schwaber (2007)
“The Scrum Guide” by Ken Schwaber and Jeff Sutherland (2009, 2010)
“The Scrum Guide” by Ken Schwaber and Jeff Sutherland (2011, 2013, 2016, 2017)
“The Scrum Guide” by Ken Schwaber and Jeff Sutherland (2020)
For every source I have described the same three topics to show what Scrum consisted of at the time (regardless the different terms used), what the ‘definition’ of Scrum was at the time:
Roles, responsibilities, accountabilities
Controls, deliverables, artifacts
Phases, meetings, time-boxes, events
For every source I have included a graphical representation of Scrum or of a Sprint that was either taken from the source directly, either from an alternative source of the same period.
Finally, I have shared my thoughts and observations on the changes to the definition of Scrum for every source. Obviously, they represent what I deem noticeable. They hold no judgement, directly nor indirectly.
To complete the paper I have listed some important landmarks in the history of Scrum and included some personal musings on the topic of “Scrum and the Desire for Universal Truths” (and what the Scrum Guide was not created for).
I hope you will enjoy reading the paper. I hope it will help you grow a deeper understanding of Scrum. I hope it will help you shape your Scrum to get the most from it. I hope it will help you create better products with Scrum while humanizing your workplace.
Take care Gunther Verheyen independent Scrum Caretaker
In 2010, the principal co-creators of Scrum, Jeff Sutherland and Ken Schwaber, agreed on the first version of the Scrum Guide, thereby creating the official BOK (body of knowledge) for Scrum. A few small, functional updates have been released since then. There have been no drastic changes to the core definition of Scrum. The updates that were released reduced complexity of the Scrum Guide or introduced clearer terminology. Language and words do matter.
The desire for precision
I imagine the difficulty of every rephrasing involved in updating the Scrum Guide. The Scrum Guide holds the single definition of Scrum for countless practitioners across the world, all employing Scrum in unique environments and varying situations. I imagine the discipline of updating the Scrum Guide knowing that many people will not read it. Most of all, I imagine the difficulty of updating the Scrum Guide knowing that many will micro-dissect the texts to identify the tiniest of changes. Many over-think individual words upon the misassumption of traps, tricks and hidden messages. Many still look for methodological exactness and universal precision.
Regardless, the fact that Scrum has a clear and documented definition is priceless. The Scrum Guide provides clarity and purpose, albeit not the perfect precision that many seem to look for.
Scrum is a tool. Useless unless employed.
The Scrum Guide intentionally has no universally applicable, detailed instructions or tactics that in the end only work in specific circumstances. The Scrum Guide describes how the tool works, the rules and roles that apply, the behaviors that make the tool work; not the tactics to apply the rules. Scrum, by design, offers room and breathing space, inviting people to conceive an approach specific to their context upon the empirical foundation of Scrum. Scrum is a framework, not a traditional (IT) methodology. The Scrum Guide describes the framework, the minimal boundaries within which people deal with complex challenges and create complex solutions in complex circumstances.
All exact decisions within those boundaries depend on people, tools, technology, business domain, organizational environment, market situation, and many other aspects.
What is the availability of people? Their skills, experience? How well do teams gel? Do they work remotely, or co-located? How long have they been working together? How much multi-tasking does a team or team members have to do? How well are teams facilitated by the environment? What technology is being used? Which version? What dependencies are at play? What development practices does a team have in place? What tools and infrastructure are at the teams’ disposal? How long are the Sprints? How is the connection with product management? What is the competition doing?
Scrum does not have the false pretence nor ambition to offer universal truths, universal answers that apply in every single situation. Scrum essentially invites people to regularly match the actual state of their work against reality, in order to optimize flow and progress, while re-aligning and adapting to new circumstances, highly disruptive events and new insights. Scrum is a simple framework for complex product delivery.
The Scrum framework offers the simplicity needed to address complex challenges, without being simplistic about the unique and specific problems that teams and organizations face. Many people struggle with this deliberate imprecision when they try to improve their understanding of Scrum. They look for detailed instructions. They ask universal questions and demand exact and precise answers. How long should Sprint Planning be? And the other events? How much time does the Product Owner role take? Is the Scrum Master role a full-time job? Should a team be available full-time? How must we organize when the team is distributed? How much time of a Sprint should a team spend on testing? How many business analysts are needed in the team? How should we calculate utilization? How can we measure productivity? Should the Product Owner and the Scrum Master understand the technology? What if… ?
No document, no matter its length, can provide exact answers to all of these questions for all Scrum practitioners of all workplaces around the world. It requires skilled professionals (Scrum Masters, trainers, and otherwise) to build on the language and words provided by the Scrum Guide and explain intent and purpose. It requires courageous professionals to help teams, organizations, and students grow an understanding of how Scrum supports them in empirically dealing with the diverse, complex challenges they face in their real-life complexity. Such professionals understand that what works today might not work tomorrow. Exact instructions lead people astray, undermine their ability to think autonomously in terms of Scrum. It is highly disrespectful. Answers to the ‘what if’ questions will emerge from the use of Scrum, from the inspections performed in Scrum. Be prepared to learn and adapt.
My personal stance, as a trainer, a friend, a trusted advisor, a whatever in Scrum, is to facilitate people in understanding the purpose of the rules and the roles of Scrum. This is at the heart of my book, Scrum – A Pocket Guide, and all my actions of promoting and explaining Scrum. The only viable way forward for people is to devise their own solutions employing Scrum. No external instance, expert or otherwise, can or should do that for them. It may be tempting. It is certainly highly convenient. It might make a person appear knowledgeable. But it promises certainty where there isn’t. That is unprofessional. It ignores context and complexity. It ignores people’s intelligence and creativity.
Every case of Scrum is unique. There is no copy-paste. There are no universal truths in the complex novelty space. The definition of Scrum requires no methodological precision. There is no future in ignoring complexity.
-An inquiry into the expression of behaviors through Scrum-
Scrum is not a cookbook ‘process’ with detailed and exhaustive prescriptions for every imaginable situation. Scrum is a framework of principles, roles and rules that thrive on the people doing Scrum. The true potential of Scrum lies in the discovery and emergence of practices, tools and techniques and in optimizing them for each organization’s specific context. Scrum is very much about behavior, much more than it is about process.
It is worthwhile elaborating on the importance of people and behavior in Scrum. Because, indeed:
Scrum is very much about behavior, much more than it is about process.
Introduction: the simplicity of Scrum
Presumably there are many reasons why Scrum turned into the leading framework for Agile software development during the past decade. One of the reasons may be the simplicity of Scrum. Or, perhaps it is the opposite, the wide adoption of Scrum is more like a miracle given that same simplicity.
Yet, the simplicity of Scrum is ESSENTIAL. The simplicity of Scrum reminds us of the fact that the real complexity to be tackled in software development lies outside of the rules and roles of Scrum. The real complexity resides in the specific context within which Scrum is applied. In software development, ‘context’ starts and ends with people; what people do, don’t do, like, dislike, prefer, hate; how people jell, feel and behave. The rules and roles of Scrum help people tackle complexity. But no ‘process’ can replace or compensate the people aspect of software development.
The simplicity of Scrum creates openness. It is an open invitation for discovery and emergence. Yet, the simplicity of Scrum gives rise to many frowns, emotions, reactions, debates. It is experienced as enticing, provocative, offensive, powerful, inadequate, a mystery, impossible. Is the beauty of Scrum, expressed through this simplicity, therefore in the eye of the beholder only? Or is there more to Scrum than meets the eye?
In the end, more than the rules and roles of Scrum, people have the key to Scrum. Behavior is the key to unleashing the potential of Scrum. Scrum is a stance, too.
The eye of the beholder
The rules and roles of Scrum are described in the Scrum Guide. The Scrum Guide was created and is maintained by Jeff Sutherland and Ken Schwaber, co-creators of Scrum. It is the definite body of knowledge to Scrum.
The rules and roles included in the Scrum Guide can be applied and followed as described, with no further inquiry into the why of these rules and roles. They can be regarded as ‘to be followed’ instructions, merely because the Scrum Guide prescribes them.
Blindly following the described roles and rules is for many in the industry of software development at least a great start to transform to Scrum. It helps. People, teams and organizations start, learn, and improve in creating and delivering software iterative-incrementally, with small steps of validated learning. However, sticking to, not transcending, such blind view and usage of Scrum is likely to turn Scrum into no more than yet another IT delivery process. As such it still leaves many holes, gaps, disconnects and waste. The rules and roles of Scrum, as described in the Scrum Guide, in themselves might not be enough to grasp the depth of Scrum and reap the full benefits of employing Scrum. The simplicity of Scrum may be somewhat deceptive.
It helps to dig deeper by:
Reflecting on the definition of Scrum included in that same Scrum Guide document, “A framework within which people can address complex adaptive problems, while productively and creatively delivering products of the highest possible value.” In the Scrum Guide this definition precedes the description of the rules and roles. This definition shows how Scrum is intended, i.e. an aid for the people employing Scrum, a way for people to structure and organize their own work. It sheds a different light on the subsequent roles and rules of the Scrum Guide. Yet, both are in the same document. Ultimately, the roles and rules described in the Scrum Guide can only be fully understood from the definition of Scrum and the clear intent expressed in that definition.
Reading the description of the roles and rules again, e.g. some time after having started with Scrum. It often leads to the discovery that the Scrum Guide describes behavior more than it has technical prescriptions. A typical focus of many processes is on ceremonial technicalities like meetings, deliverables, timings, phases; i.e. on what is expected from people. Scrum is a framework that gives people the room to organize their own work, yet provides boundaries as every healthy ecosystem needs.
Stepping back to a perspective that goes beyond the Scrum Guide document, the perspective offered in the Manifesto for Agile Software Development. The rules and roles described in the Scrum Guide, the behavior described, don’t just stand on their own. They are grounded in the values and principles expressed in the Agile Manifesto. Ultimately, the Scrum framework can only be fully comprehended when seen as an expression of these values and principles. Ultimately, the rules and roles of Scrum, the behavior described in the Scrum Guide, can only be fully understood in combination with the fundamental views expressed in the Agile Manifesto.
The intent and definition of Scrum, matched against the agile values and principles should ground and then drive the behavior expressed through Scrum.
The Stance of Scrum
Scrum has many facets. Scrum is a framework, not a methodology. The framework of Scrum is not just a set of technical prescriptions, but a recipe to deal with complex challenges. The rules and roles of Scrum support and complement, not replace, the intelligence and creativity of people. The framework of Scrum is an implementation of the values and principles of the Agile Manifesto. Scrum implements empiricism in software development.
The framework of Scrum thrives on implied principles, thinking and… behavior, on people taking a stance to product development through Scrum, a Scrum stance.
The definition of Scrum shows the way to the core of the stance typical to Scrum. If Scrum is an operating system for the values and principles expressed in the Agile Manifesto, the definition from the Scrum Guide shows the way to the kernel of the operating system.
The kernel is expressed as:
PEOPLE EMPLOY EMPIRICISM TO OPTIMIZE THE VALUE OF THEIR WORK.
People are respected for their intelligence, creativity and ability to organize their own work, to self-organize. People collaborate, thereby adhering to the values and principles of the Agile Manifesto and embodying the Scrum values of respect, focus, courage, openness, and commitment.
Empiricism serves to deal with the complexity typical to software development. In empiricism only reality and past results are accepted as certain. At a regular cadence outcomes and behaviors are transparently inspected for new and changed insights against set goals. These insights are used to adapt to observed reality.
The value of outcomes, work delivered to an ecosystem of creators, stakeholders and consumers, is constantly evaluated, optimized and maximized as a shared goal. Value comes in different shapes and appearances; satisfaction, money, improvement, credibility, risk. Optimizing value is very different from adhering to traditional development drivers like budget, tasks, scope, time, schedule.
In the end, Scrum has many appearances. In the end, Scrum -like all things agile- starts and ends with people. Scrum is a stance, too.
Scrum is a framework designed to help teams create and sustain complex software products in complex circumstances through the scientific method of empiricism. Scrum has replaced the industrial, plan-driven paradigm with well-considered experimentation to better deal with the complexity and unpredictability of modern software development. And although the use of Scrum is free, its roles, artifacts, events, and rules are immutable. Implementing only parts of Scrum is possible, the result however is not Scrum. By breaking its base design it is likely that problems are covered up, instead of being revealed.
The purpose of Scrum is to help people inspect & adapt, to provide transparency to the work being undertaken, to know reality to base decisions on, to adjust, to adapt, to change, to gain flexibility. The rules, principles and roles of the framework, as described in the Scrum Guide, serve this purpose. These are the rules of the game of Scrum, the base setup to have in place.
But we distinguish principles from techniques, the what from the how, the rules of the game from tactics to play the game. It’s like in all games and sports, all players and teams play upon the same rules, yet some teams seem more successful at playing than other teams. It depends on many factors, and not all are equally controllable by the teams themselves. One factor are the tactics used by the team to play.
There are many tactics to use within Scrum. Good tactics serve the purpose of Scrum. Good tactics re-enforce the Scrum values, not undercut them.
I’ve previously described how practices from eXtreme Programming provide great tactics, ways to play Scrum, even when not mandatory from the Scrum perspective.
Here are some more illustrations on the distinction between the purpose of Scrum and tactics for that purpose:
The Daily Scrum questions
The Scrum Guide suggests that in the Daily Scrum meeting every player of the Team answers 3 questions (Done? Planned? Impediments?).
But the players might just formally answer the questions, limit it to a personal status update, and talk to the walls or to the Scrum Board. They just make sure that they -well- answer the 3 questions. Because the Scrum Guide tells them to, or some smart coach or Scrum Master or manager.
But is the team seeing Scrum as a methodology? Or use Scrum as a framework for discovery? It doesn’t help much if they don’t talk to each other. It doesn’t help much if they don’t surface the information to optimize their collaborative work plan for the next 24 hours against the Sprint Goal. Maybe they use the meeting only as a reporting obligation in which they make sure all their microtasks are logged, to cover against possible blame. They miss the opportunity to gain insight in the real situation, to inspect it and to adapt upon it.
Maybe Scrum should describe only ‘what’ is expected from the Daily Scrum in the time-box of 15 minutes. Although, actually, the Guide already does so by saying: “The Daily Scrum is a 15-minute time-boxed event for the Development Team to synchronize activities and create a plan for the next 24 hours.” We could eliminate or rephrase how this is exactly to be achieved, in turning the 3 questions into a good, yet non-mandatory, tactic.
The Definition of Done
The Definition of Done was rightfully given a lot more attention in the latest version of the Scrum Guide. However, some people wonder why it isn’t an official artefact. I did so too, even launching the suggestion to make it one. However, I got to see that the Definition of Done is an aid to transparency, and I overcame my urge to turn it into an artefact.
Here’s how the Definition of Done helps playing Scrum:
Scrum assumes an inspection of a working product Increment at the end of every Sprint, at the Sprint Review. The Definition of Done gives transparency to that inspection by unveiling what’s the work that has been done upon the Increment, what criteria were met.
The Definition of Done guides the team at the Sprint Planning to pull work into the Sprint, as the work needed to get to “Done” is likely to influence the amount of Product Backlog that can be pulled into a Sprint, as well as the estimates for that work -if the team uses any-.
The Definition of Done helps the team create and manage their work plan for the Sprint in getting things done, and not end up with a pile of nearly-done work at the end of a Sprint.
So, what Scrum requires is transparency. The Definition of Done is a way to provide that transparency at several levels and several occasions. It’s a great tactic that stresses the importance of seizing the opportunity to ship.
Product Backlog Grooming
Product Backlog grooming is an activity in which the Development Team and the Product Owner look at Product Backlog currently sorted for one of the next Sprints. Certainty that the included items actually are going to be implemented is growing. So teams might want to unveil dependencies or help a Product Owner understand what is useful to know from a development perspective. Grooming increases the chances to pull the work in more easily when it is presented at the next Sprint Planning.
Product Backlog grooming is not a mandatory (time-boxed) Scrum event. The ambition of Scrum is to remain simple, yet sufficient. The ambition of Scrum is to help people and teams discover additional practices that may or may not be appropriate in their specific context. Product Backlog grooming is an activity that seems to help many teams to smoothen their Sprints, and certainly limit turbulence in the first days of a Sprint. Other teams however cope without it, and perceive it as optional or even overhead if it was mandatory from the Scrum framework.
Product Backlog grooming is a great activity within a Sprint, a good tactic to collaboratively manage Product Backlog. Some can do without however.
Sprint Planning Part 1+2
Sprint Planning meeting is an opportunity to inspect the actual state of Product Backlog and identify what work is most useful and feasible at the actual point in time, with the option of pouring the work into a work plan for the Sprint.
The empiricism of Scrum assumes following for the Sprint Planning:
IN: Product Backlog
OUT: Sprint Goal & Sprint Backlog
Currently the Scrum Guide prescribes the flow of the Sprint Planning meeting:
Part 1: Product Owner explains highest sorted Product Backlog items. The Team discusses, evaluates and pulls in items. A Sprint Goal is crafted.
Part 2: the Team decomposes, discusses and designs the work.
It is however possible to respect the IN and OUT of the Sprint Planning meeting upon a different flow of the meeting. Maybe Scrum can suffice by describing only ‘what’ is expected from the Sprint Planning in its time-box without prescribing ‘how’ the Scrum Team should achieve this. Organizing it in a part 1 + part 2 might be a great tactic, but it’s probably not the only one.
Scrum has no exhaustive and formal prescriptions on how to design and plan the work, actions and behavior of all players involved in product development against time, let alone how such designs and plans would have to be documented, approved, stored, etc. Scrum has no rules for upfront predictions of document types and (intermediate) deliverables to be produced or the time at which they should be produced. Instead of installing and thriving on hand-overs, toll gates and control meetings like IT methodologies typically do, Scrum helps eliminating them as a major source of delays and waste.
‘Methodologies’ are typically composed of stringent and mandatory sequences of processes and procedures that implement predefined algorithms. As such, methodologies tend to rule out the creativity, autonomy and thinking of people with components like phases, tasks, must-do practices, techniques and tools. As long as the methodology is being followed everyone feels safe, because they are formally covered, even in the absence of working results or proven progress. Methodologies depend on high degrees of predictability, otherwise the preset algorithms fail. Complex problems, for which Scrum was designed, are more unpredictable than they are predictable. Today’s world has more complex challenges than simple cases.
Scrum is not a methodology. Scrum is even the opposite of such big collections of interwoven mandatory components and instructions. Scrum implements the scientific method of empiricism, the process of inspecting in order to adapt at regular intervals, not aspiring to try to predict what the adaptations will be. Scrum replaces a programmed algorithmic approach with a heuristic one, with respect for people and thriving on the self-organizing capabilities of people to deal with unpredictability and address complex challenges.
Is Scrum a process?
If Scrum is a process, it is certainly not a repeatable process. That might be a challenge to explain, because the term ‘process’ typically invokes a sense of algorithmically predictable steps, repeatable actions and enforceable top-down control; the sort of expectations that in our world are typical for a… methodology.
Scrum is not a commanding process. If referred to as a ‘process’, then Scrum is a servant process. What works best for all involved players, their working process, emerges from the use of Scrum. The players discover the work required to close the gap between an inspected intermediate result and an envisioned outcome. Scrum is a process that helps surface the real (daily) process, structures and a way of working that are continuously adapted to the actual context and current circumstances. Therefore we call Scrum a… framework.
Scrum is a framework
Scrum describes the roles and rules upon given principles that help and facilitate people in a low-prescriptive way, that help people create a framework for regular inspections and adaptations. The Scrum Guide holds the definitive description of the base rules of the game. The prescriptions are minimal, but every single one of them addresses dysfunctions that were (and are) common in (software) product development.
Over the 2+ decades of Scrum, the rules of Scrum, as captured in the Scrum Guide, have gradually evolved, with small functional updates and releases. The prescriptions of Scrum, what needs to be in place to have the full benefits of Scrum, become more and more focused on emphasizing ‘what’ is expected in developing complex products over instructing ‘how’ to do it.
A good illustration of such an evolution is the elimination of burndown charts from the Scrum framework as mandatory (a ‘how’ of managing progress). This obligation however has been replaced by the explicit expectation that progress on the mandatory Scrum artefacts, the Product Backlog and the Sprint Backlog, is visualized (the ‘what’). The form or format of the visualization is no longer prescribed, thereby turning burndown charts into a non-mandatory, but still good practice; a good way to play the game and suitable in many situations.
Yes, it’s Scrum if Product Backlog and Sprint Backlog exist and if a visualization of their progress is available, accessible and clear. This may be a burn-down chart with open effort. It may also be a burn-up chart in value. It may be a Cumulative Flow Diagram. It may be as simple as a Scrum board.
The Scrum framework leaves options for different tactics to play the game, ways that can at any time be adopted to the context and circumstances. Scrum, as a framework, can wrap many practices. When applied well, the overall system will still be recognisably… Scrum.
The Scrum core values give direction to the actions, the behavior and the additions to the framework. Upon that core, in a ScrumAnd way of thinking many opportunities emerge. Have a look at some illustrations of ScrumAnd.