How Agile will expose your decision-making effectiveness

Article by Certus3 managing partners Simo Popovac and Michael Devlin.

A key goal for many organisations that go down the Agile path is to move faster.

The language of Agile supports this, packaging work into sprints and delivering new features and products on a regular cadence. So does the live experience: velocity is listed as the third-highest measure of success for individual Agile projects in the latest State of Agile report.

How fast you can go depends on a range of factors.

First, as Atlassian notes, velocity varies between teams. Each team estimates the amount of work it can complete in an iteration differently, and therefore works to a different pace. However, one would expect that pace to increase over time “as the team optimises relationships and the work process”, Atlassian says. There is a direct relationship between Team Performance and velocity.

Second, the higher the velocity, the better an organisation has to be at decision-making but making the right decisions.

In our experience, this is an area where many organisations still find they need some help. Artificial intelligence shows tremendous promise in this field because it is able to monitor across a vast array of complex scenarios thrown up in Agile projects and surface timely information and insights that help the business leaders overseeing these projects to adapt on-the-fly, make the right decisions at speed and keep to time.

Bad decisions still abound

Decision-making in Agile organisations is hard.

A survey by McKinsey in April found only 48 per cent of respondents agreed that “their organisations make decisions quickly”. Decisions taken at speed were not necessarily good; “just 37 per cent of respondents say their organisations’ decisions are both high in quality and velocity,” McKinsey found.

As if to highlight that, an earlier study, also by McKinsey, found 72 per cent of senior executives “thought bad strategic decisions either were about as frequent as good ones or were the prevailing norm in their organisation”.

That earlier study recognised the role that Agile organisational models could play in getting decision making “into the right hands” and being able to react to or anticipate shifts in the business environment faster.

Yet, adopting Agile by itself is not a guarantee that decision-making speed and processes will improve.

“In the digital age, good decision making entails taking more shots on goal and shortening iteration cycles. However, few decision makers are rewarded for such an approach,” a March survey says.

The success of a decision is still measured on the outcome it produces. How you arrive at that decision can be augmented and innovated on, and there is clearly room for that to occur.

A Swedish study on data-driven decision-making presented at the International Conference on Agile Software Development in late May shows the enormous promise of AI in this space.

While 79 per cent of respondents said data was “highly valued in today’s decision-making, a majority of the respondents agreed or strongly agreed that data should play an important role (71 per cent) and be highly valued (87 per cent) when making decisions in the future.

Bringing in artificial intelligence

In an Agile environment, governance is required to understand the metrics that indicate success in overall project terms and what actions need to be taken and when to get there. In that respect, information is power – the power to be successful.

Senior executives responsible for governing and assuring the success of Agile-driven transformation projects are rethinking how they get access to the right information at speed to help make good decisions.

Artificial intelligence (AI) is emerging as a key enabler. AI can assist people to access information that was previously inaccessible in a timeframe and format that enables sound and timely decision-making.

By making use of machine learning algorithms and expert systems, organisations can gather data from across a project and model it in new ways.

AI-based systems can also protect against internal bias and other factors which might weigh on the direction of decisions and results. Within Agile, you depend heavily on teams to accurately estimate how much work they can get through, and on people to provide assurance that things look correct. This is very prone to being influenced by organisational culture, politics and biases.

What is clear to date is that without AI, organisations and executives are far more limited in being able to measure and use the information for fast and accurate decision-making. Data-driven decision-making is the key to unlocking Agile success.

Why CTOs need to stop being overly-apologetic about Agile

Article by Certus3 managing partner Michael Devlin and Certus3 managing partner Simo Popovac.

Australia has more than its share of companies on organisation-wide Agile transformation journeys.

No one goes into these initiatives underestimating the complexity involved, but few companies emerge from the journey independently – without needing to call in reinforcements.

That’s because Agile at scale has its share of purists, models, misconceptions, and ambiguity. All are capable of sending an organisation-wide Agile transformation off the rails.

Many organisations are now dealing with the consequences of making the transition to Agile. How do I know if I’m doing Agile right? How can I develop maturity as the transformation progresses? These are just some of the questions we see Australian organisations asking themselves.

Some questioning of process and practice is healthy and is important to the iterative development of new ways of working inside of an organisation.

All too often, however, Agile projects in Australia are closely modelled on what has worked elsewhere. For many, it means adopting the so-called Spotify model for Agile organisational design – a way of arranging workers into cross-functional teams.

It’s often not the case that one size fits all. No one Agile playbook works.

Therefore, success is really about having the confidence to implement a flavour of Agile specific to your own organisational constraints and needs. It’s also about recognising and pushing back against other “needs” that might get bundled up in Agile transformations, but which do not necessarily fit with your own.

There’s no need for Agile purity

Agile ways of working do not exist only in a pure form.

Author Allan Kelly argues that Agility is a spectrum, with strict Waterfall at one end and Pure Agile at the other. The extremities are “sparsely populated”, Kelly says. Most companies land on the spectrum somewhere between these two extremes.

Oftentimes, companies are apologetic that the approach they’re taking is not Pure Agile. Arguably, however, those that strike the right balance are simply demonstrating a greater degree of maturity in their approach to Agile.

Agile maturity is often underestimated – or perhaps companies are overly critical of their own maturity, particularly where the end goal is defined in Pure Agile terms.

An annual ‘state of Agile’ survey last year found 84 per cent of organisations identify as being “at or below a ‘still maturing’ level when it comes to Agile.

Some organisations may be hard markers of their own Agile progress and performance. Positives for organisational and innovation culture are still possible without an end goal of Pure Agile.

But the high percentage also shows that not every company is as mature as you might think. Though it may seem competitors or early adopters have it together, many are still afflicted by the same core doubts about whether or not they are doing Agile right.

Of course, running in a hybrid fashion, potentially with different projects and parts of the organisation at different levels of Agile maturity simultaneously, comes with its own set of challenges. These include the extent to which it is possible to maintain a level of consistency, control and planning between these teams or programs of work. That may or may not be simpler than bringing an entire organisation up to the same bar of Agile maturity at once.

The decision is best determined by individual company circumstances.

Deconstructing org structure

Another maturity misconception is that people and programs of work naturally function better and faster in an Agile model.

Being performant is overtly promoted in Agile.

Agile relies on capable skilled individuals working in reasonably autonomous teams, often colocated with one another. Individuals and teams in Agile environments are seen as being well-connected, well-formed, and high-performing, with a greater ability to deal with change and to tolerate and respond to fluctuating levels of risk.

But high-performing teams exist outside of Pure Agile organisations as well. The same or similar results can be achieved in companies with a waterfall or hybrid organisational structures.

A related misconception is that self-governance reduces the need for external assurance. That is, a high-performing team does not need to track or measure its performance independently because of its inherent performance characteristics.

This notion is challenged, we would argue, by the spectrum of possible paths to agility and the lack of a one-size-fits-all approach to achieving success. Independent assurance is just that – a confirmation that things are, indeed, still on track.