This post is adapted from Chapter 1 of Sooner Safer Happier by Jonathan Smart, with Zsolt Berend, Myles Ogilvie, and Simon Rohrer.
Do you want to do or are you currently doing an Agile, Lean, or DevOps Transformation? If so, my best advice is:
Instead, focus on the outcomes you want to achieve. Then you will achieve agility.
Focus on: Better Value Sooner Safer Happier.
This is the number one lesson I’ve learned after almost thirty years as an agile and lean practitioner delivering business value through software in the Age of Digital, from leading Ways of Working at a large, old, global, regulated organization to working with many large firms across different industry sectors. Together as a team-of-teams, as servant leaders, we experimented, learned, and pivoted.
Agile, Lean, and DevOps are not the goal. An organization can score highly on a “How Agile Are We?” test (or worse, “How Much Are We Rigidly Complying to a Specific Agile Framework?” test, or “How Many Scrum Teams Do We Have?” test) without producing better business outcomes. I’ve seen it happen time and time again. The wrong thing can be produced more quickly. Teams can become feature factories, a self-fulfilling prophecy of backlog replenishment with a focus on “More output!” rather than a focus on better outcomes. In addition, Agile can be viewed as an IT-only thing, no more than a local optimization, an agile bubble in a sea of traditional approaches. Or teams can exhibit cargo cult behaviors, with new labels and rituals but with the same old behaviors as before.
Agile, Lean, DevOps, design thinking, systems thinking, Theory of Constraints, and so on are all proverbial tools in a toolbox that organizations can employ to achieve desired outcomes. They are bodies of knowledge, years of wisdom acquired in the field of organized human endeavor, articulated as principles and practices. As we’ve seen, they are suited to specific contexts—contexts that are the new normal in the Age of Digital—as venerable old firms (the “horses” of a previous age rather than new digital “unicorns”) move on from ways of working that are more than a hundred years old, originating from two technological revolutions ago in the late 1800s.
Every organization is unique and is a complex adaptive system. Culture change is emergent. So the interventions chosen need to be applied uniquely in context. There is no cookie-cutter, one-size-fits-all approach. There is no silver bullet, no snake oil, no panacea. To know whether the bodies of knowledge, the principles and practices you’re using, are having the desired impact, you need to know what your desired outcomes are and keep your eye on that ball. What job are you using the bodies of knowledge for? What result do you want to produce?
At every organization I’ve worked for or with, those desired outcomes can be articulated as Better Value Sooner Safer Happier (BVSSH).
What Is Better Value Sooner Safer Happier?
So what is Better Value Sooner Safer Happier? What do the terms represent and how are they measured? An important point to note is that they are not only IT outcomes and measures. They apply across organizations, anywhere work is being done to deliver value. They’re about a collective “our business,” not an us-and-them “the business,” irrespective of job role. In the Age of Digital every company is a software company directly or indirectly, and there are few cases where value delivery does not in some way involve Information Technology.
Better is quality. For example, for a software product “better” could mean fewer production incidents, a faster mean time to recovery, and improved static code analysis measures. For internal audit, “better” could be less rework of internal reports. For an operational area of an organization, such as processing payments, transactions, or loan applications, “better” could be a lower error rate. The lower the “failure demand,” the lower the cost of keeping the lights on and the greater the percentage of the budget that can be spent on new value-adding activities. Quality should be built in, rather than inspected in later.
Value is in the eye of the beholder. It is unique and it is articulated via quarterly business outcome (also known as Objective & Key Results or OKRs). It’s why you’re in business. “Value” could be market share, revenue, units sold, P&L, margin, diversity, carbon emissions, app downloads, minutes streamed, subscribers, and so on. Value should cover the perspective of the consumer and producer.
Business outcomes are hypotheses, as we’re in the emergent domain. They are nested, with a lineage up to longer-term, organization
wide strategic outcome hypotheses (yearly and multi-year). There is fast feedback with daily releases of value into the hands of customers to test the hypotheses. The value measures are the KRs in OKR with leading and lagging measures. Daily, weekly, monthly nested cadences enable pivoting based on fast learning. Typically there is a monthly cadence on the quarterly business outcomes to inspect and adapt. With daily releases of value, it is possible to have daily feedback on multi-year strategic hypotheses. See Chapter 5 for more on this.
Sooner is flow, which is at the heart of agile and lean. It’s about optimizing for fast and efficient flow of safe value with respect for people. There are three key measures that can be aggregated up to the organization level or disaggregated down to the team level:
- Flow efficiency is the percentage of time that work is actively being worked on during its elapsed end-to-end lead time, as opposed to waiting to be worked on. It is one of the most important measures, yet it is rare to find an organization that knows its flow efficiency for knowledge work. For most large service-based organizations, in my experience, flow efficiency is typically 10% or lower. This means that work is waiting at least 90% of the time. This is where significant gains can be made. Focus on where the work isn’t, not where the work is. Focus on the work, not the worker. The wait time is usually caused by impediments to flow, such as role-based or time-zone handoffs or multiple committee review steps, leading to work being queued. A high wait time is also caused by organizations attempting to do too much work in parallel. The more cars on the road, the slower they go. Identify and alleviate the impediments to flow and limit concurrent work in progress.
- Lead time is time to market, the time from starting work on an item of value to getting it into the hands of a customer. Reducing lead time enables faster feedback, quicker learning, reduced risk, earlier monetization, and the ability to pivot sooner to maximize outcomes. Lead time is a distribution—typically a Weibull distribution, a type of continuous probability distribution—that resembles a normal distribution skewed to the left and with a long tail. The recommended measure is the 85th percentile lead time and its change over time.
- Throughput is a count of items of value delivered into the hands of a customer in a given time period. As lead time comes down, throughput should go up. If it doesn’t, then flow has an upstream impediment. Ideally, throughput should not increase directly in line with reduction in lead time. Instead, some of the time gained from reducing lead time should be used for innovation, time with customers, and continuing to improve the system of work, further alleviating impediments to flow. We want to maximize outcomes with minimal output.
Note that the word “faster” does not appear here. “Faster” can have negative connotations. A “feature factory” can work fast, churning out features that no one wants, working harder rather than smarter.
Safer means continuous compliance, agile not fragile, a topic we cover in Chapter 6. It is about not making the news headlines due to leaking customer data. Safer is Information Security, cyber, data privacy, General Data Protection Regulation, know-your-client, anti-money laundering, fraud, and so on. It is Governance, Risk, and Compliance (GRC). Safer is speed and control, not choosing one at the expense of the other. Safer is cultural, with a continuous conversation on risk.
Happier is happier colleagues, customers, citizens, and climate. Improving ways of working is not at any human, societal, or climatic cost. It is about a more humane, engaging way of working, with multidisciplinary, empowered teams centered around the customer. Happier is working smarter not harder; it is improving the system of work and removing impediments. Happier is obsessing about customer satisfaction (which will lead to revenue, rather than a primary focus on short-term financial measures). Happier is also about social and climatic responsibility.
Together, Better Value Sooner Safer Happier balance each other. If Sooner is achieved by working people harder or cutting corners, the result will be a reduction in Better and Happier.
BVSSH contains two sets of outcomes. Better Sooner Safer Happier are the how outcomes. They measure the improvement in the system of work. Value is the what, the business outcome hypotheses that the system of work produces and that I discuss in Chapter 5. The two sets of outcomes form a virtuous circle. Improvements in the how leads to improvement in the what due to faster feedback, the ability to pivot, higher quality, and more engaged colleagues and customers.
Note that just as I don’t mention “faster,” I also never use the word “cheaper.” A lesson learned by organizations adopting lean principles and practices in Japan is that “cheaper” is an antipattern. It will create a headwind. People don’t want to work themselves or a colleague out of a job. It is not a motivating call to action. Cheaper has negative connotations on quality and happiness.
Also, a focus on reducing visible costs often increases hidden costs via a reduction in flow efficiency. There is a hidden cost to cost-cutting. For example, introducing more handoffs, communication paths, time-zone challenges, differing incentivization, and so on all reduce flow efficiency. It increases the time that work is waiting. This reduces throughput and makes lead time longer. The system of work becomes less efficient. Learning and pivoting is slower. The company spends less, but it’s also doing less and has made the system of work less effective. It’s a double whammy on the ability to generate value. The organization has throttled back, both with a step change down in value production and a reduction in the gradient of adding value over time, due to reduced flow efficiency. This reduces income, which puts further challenges on cost.
Improving ways of working for product development is about “value-tivity.” We want to optimize for value and time to learning. Outcomes over output. We want to maximize outcomes in the shortest time and with the least output. We want to maximize the value curve, cut the tail, and pivot to the next value curve. Typically, a focus on “cheaper” has the opposite effect, making time to value and time to learning longer.
Instead of cheapness, focus on Better Value Sooner Safer Happier outcomes and improve the system of work. As lead time reduces and throughput increases, striving for the highest value in the shortest time, with greater agility, the “income” in cost-income ratio should improve, all other things being equal and compared to maintaining the status quo.
If an organization doesn’t have the runway to improve first, or macro drivers (such as a global pandemic), to change the business fundamentals and force a need to spend less and do less, my advice is to pay very close attention to the system of work. Don’t increase the hidden costs with a reduction in flow efficiency and a longer time to value and learning. Don’t prioritize cutting costs at the expense of flow. The result will be increased hidden costs. Do have a focus on throughput accounting as well as traditional cost accounting.
Now that you have a deeper understanding of Better Value Sooner Safer Happier, and you are ready to focus on outcomes rather than Agile, we will look at two of the most important, most fundamental, antipatterns. They generate a significant headwind. They are detrimental, as they do not apply an agile mindset to agility. These antipatterns (as with all antipatterns) are approaches that more often than not reduce the likelihood of achieving desired outcomes. They make a hard job harder.