Economic Stages of Competitive Agility

If we explore operational excellence “transformation” using our economic definition of agility: responsiveness to signals in a market with imperfect information and imperfect competition we see six major phases. You will note that some ideological camps – perhaps even the one you are currently involved with – tend to promise, but frequently fall short, of any realistic definition of Agile Utopia.

Stage 1 – Big Bang Waterfall

Big Bang Waterfall is the post-apocalyptic, dystopian, late-adopter, core-competence-turned-mass-layoff nonsense that every marketer, sales person, or consultant uses to sell the need for Agile (with a capital-marketing-“A”) utopia. Although serious blunders have taken place by late adopters, even contractors have gained enough operational best practices to assist in limiting the damage done by “Big Bang” – after all, extremely long-term contracts must be extremely defensible or they are poorly enforceable. This would be overhead few development consultancies could afford. Instead, like your average apartment in the city, everyone agrees that a commitment horizon of greater than one year with no clause for early cancellation is foolhardy and unrealistic.

1. Reliance on top-down internal signals
2. No concept of holding costs for information and knowledge
3. Unknown economies of scale are assumed infinite
4. No control over Work-in-Progress

Stage Gate Control

This is the real “waterfall” – in the real world – that you are likely to see. As it turns out, it is socioeconomically defensible: if a company clearly understands its strategic position, understands the necessary tradeoffs of that strategic position, and that position itself limits its digital product presence, it may make sense to fully outsource digital product design and development. Likewise, if and only if that position makes it economically advantageous to take a risk-intolerant late-adopter position for new technology, it makes sense to invest only in short bursts to keep up with the minimum expectation of that industry. It would even make sense to copy proven market demand and the value offering of proven winners in digital because innovation is very often a winner-takes-all competition. Unfortunately, the majority of companies we see that behave this way do not do it because of strong strategic focus. Many of them may have built their PMO on this foundation and did not notice the industry change around them.

1. Reliance on top-down internal signals, with additional “resonators” added
2. Increased overhead makes holding costs for information and knowledge apparent
3. Unknown economies of scale are assumed infinite even though batch size is limited
4. Project duration limits provide control over Work-in-Progress

Cost Center Agility (Agile XP, Kanban, Scrum)

This is the agility the sales reps deliver after promising utopia. I will not name any names. The tool, the process, or the framework can be very lucrative when sold at the right price, while a handful of true believers can reinforce the value proposition on behalf of the sales team for a lifetime. From our economic view, the real benefit of all such systems originates with two shifts: 1 – The gradual shortening of planning horizons until they are realistic, suited to the volatility of their market. 2 – The visibility and progressive restriction of work-in-progress, controlling previously ignored holding costs for information.

1. Tempo of internal signaling is increased
2. Sequence or prioritization can manipulate holding cost for information and knowledge
3. Stable teams can collect enough data to reveal actual scale economies
4. Time-boxed incremental effort provides control over Work-in-Progress

Continuous Delivery (CI/CD, ATDD, DevOps)

Continuous Delivery is focused on automation of manual activities. These activities were economically appropriate when short-run optimum batch size and long-run were equal (i.e. one project building something that lasts a long useful life). In Cost Center Agility, it became apparent that the economics of short-run manual transactions like testing were very different from the long-run economics, which justify automation. Throughout this process, transaction costs for testing internal and external signals of value must be minimized. The most important shift that occurs due to this reduction of transaction costs is an increasing definition of quality and an increasing check for signals. It is the first time in the pursuit of agility that the organization begins to seriously and methodically consider the possibility “we might be wrong.”

1. Internal signaling is formalized and shifts toward instantaneous
2. Decentralized control diminishes transaction cost considerations
3.  Holding cost for information and knowledge is primarily within the planning and design portion of the value stream, making product marketing behavior patterns from Big Bang Waterfall unfit
4. Product operational data is aggregated, allowing multi-fractal pattern analysis
5. Canary analysis, A/B deployment, and automated rollout remove Work-in-Progress pressures

Hypothesis Driven

Once we move from the assumption we, as rationalists, can create and deliver against a “perfect” solution plan, and work instead from the assumption that uncertainty makes it necessary to validate continuously whether or not our assumptions are correct, we can then ramp up our attention and responsiveness to market signals directly. It is not sufficient to listen to complaints and work “very, very hard” to please people. We must be relentlessly scientific and maintain strategic focus at all times.

1. Internal signaling is consistent, reliable, and part of organization self-identity
2. External market signaling replaces extensive planning because product marketing cannot maintain the same pace as technical delivery
3. Holding cost for information and knowledge is diminished through direct market responsiveness
4. Aggregated multi-fractal pattern analysis now combines marketing and operational
5. Distributed control and single-piece flow reveals and removes value stream inefficiencies

Culture of Innovation

The Culture of Innovation, frequently promised as part of “agile utopia” is really not necessary for most businesses. This is because, unless you have established a strong strategic position that necessitates continuous innovation – that is, unless you are a technology company –the risk of such novelty is unjustified. Most organizations are wise to encourage “innovation” as a benefit to employees while maintaining tight control over administrative context and strategic fit. Especially for a mature publicly-traded company, this typically implies spinning off the new business unit because it no longer fits well with the historic risk profile of its stocks.

The Limits of Agile as Operational Effectiveness

Most organizations end up “stuck” in stages three and four. Without a clear of understanding the economics of operational effectiveness, this is the source of years of frustration for the consultants and coaches working diligently to encourage best practices that have diminishing marginal utility.

The standard of Operational Excellence for the majority of companies will likely fall in the Continuous Delivery stage, occasionally flirting with Hypothesis-Driven design-development (also called Lean UX by some). The Culture of Innovation, when looked at closely, is actually quite extreme. As mentioned above, we really would not want this level of agility in most of lives or in most of the economy. There is an element of controlled gambling because the economics of this stage rely upon asymmetric payoff and induced demand, creating or expanding demand for something no one had asked for. It requires such an intimate knowledge of the market that a company can go above and beyond extreme responsiveness and attempt to predict or even invent non-existent future demand.

None of these phases (or their ideological camps) are intrinsically correct or incorrect for a company’s digital product delivery. Instead, the validity of the philosophies, processes, and tools at each phase depend on the economics of interaction with the market. While none of these “phases” are intrinsically or ideologically correct, the firm that pursues a differentiation strategy dependent on superiority in digital product innovation as a competitive advantage will fail without guiding the economics of responsiveness to market signals. If a company chooses to pursue innovation as a competitive strategy, it will go through these phases to get there.