Algorithmic Leaders

As mentioned in my previous article, ‘algorithmic risk management’, several key technological trends have converged to drive a step change in software innovation which presents an opportunity to evolve risk management to help improve project outcomes. Whilst the engineering and construction sector has been lagging other industries, it’s finally beginning to wake up to the potential value of big data and artificial intelligence (AI), especially in the form of machine learning. Reports of new use cases are beginning to appear more frequently and adoption appears to be on the rise.

Recognising the potential value is one thing but understanding how to exploit the technology and then to drive organisational change is something else. This is where organisational leadership is vital especially when this new knowledge, expertise or capability has not traditionally existed within organisations. This is applicable to not only client-side organisations wanting to use technology but also software developers.

Business leaders within construction and engineering companies are accountable for developing a vision, strategy, building a strong team and nurturing the right culture to exploit opportunities. They don’t necessarily need to understand all the technical details but the process of making ‘big’ decisions and taking risks comes from the accumulation of their own experiences usually in areas they are familiar with, the paths they have travelled or the people (experts) they surround themselves with. This helps them to assess the level of risk and reward associated with potential new ventures.

It’s not implausible to suggest some (potentially many) current leaders within these companies are not familiar with or have not been exposed to AI because of the nature of their core businesses. It’s also unlikely that their traditional recruitment strategies have included data analysts, data scientists or programmers which might have contributed to the lack of diversity in roles and people within their organisations to either offer advice or drive innovation internally. Another reason may be a generational issue in that some leaders have not grown up with the exposure to rapidly evolving technology that the younger generation have become accustomed to and are more open to adopting and experimenting with.

“Every organization faces the problem of how to identify the people who are most likely to lead your teams through growing complexity, uncertainty, and change. Such individuals may have a very different profile from those who have succeeded in the past, as well as from those who are succeeding in the present”. - Harvard Business Review

Perhaps for the above reasons client-side leadership has either not moved with the times or simply failed to establish or nurture the right culture to promote technological innovation from within the business, especially within a risk management context. At the other end of the spectrum, leadership across industries where technology and data have been prioritised and are more closely aligned to their core business strategy have promoted and embedded a culture of technological creativity and innovation at their very core (i.e. finance, retail and entertainment). Within these industries the data has been in more plentiful supply, more structured and codified in a way that lends itself more easily to exploiting developments in AI. The recruitment strategy in these industries also has a long history of attracting the type of technical experts mentioned above. Whilst it is improving, this is certainly less mature in the construction and engineering sector which may have constrained progress.

It is somewhat more surprising that risk software developers, where technology is at the heart of their business, have also been slow to lead the application of advanced project data analytics and AI to evolve their products. It may be that traditionally when software companies looked to develop their products, most of the time and energy was spent exploring the product’s technical features by looking on the ‘inside’. Less time was spent looking ‘outside’ to understand how external clients and the technological context was evolving and attempting to extrapolate ahead. The lack of competition across the relatively small number of risk software vendors may have also stifled innovation.

Enterprise risk management (ERM) system developers may have also taken the lead from long established risk management processes promoted by professional project or risk management organisations. In the past, beyond intellectual exercises, there has been no significant innovation promoted by these organisations in leveraging the power of risk data or risk analysis. This influence on software developers may be evidenced by how similar the most popular ERM systems appear to be and are overly focused on the traditional form of a risk register. In the main they still appear to be glorified databases with limited functionality beyond data entry and basic reporting.

In addition, risk software supporting more sophisticated forms of quantitative risk analysis (QRA) haven’t really evolved significantly over the last decade. There are alternatives which have incrementally improved aspects of functionality (e.g. integrated cost and schedule risk analysis), streamlined user interfaces and more automation, however, given the level and pace of technological disruption AI has driven in other industries, progress in this area pales by comparison.

On a positive note there is more recent evidence to suggest professional bodies are beginning to recognise the value in techniques like QRA and a dialogue finally appears to be opening up on the potential application of big data and AI. However, one could argue they are tentatively following rather than leading the thought leadership piece when compared to other types of leaders described later. There does appear to be a greater level of diversity slowly growing within these organisations and institutions with people from different cultures, roles, industries and perspectives which might explain them being more receptive to change.

Incremental evolution isn’t necessarily a bad thing but it’s just that the focus has been on current customers and stakeholders and existing ways of doing things rather than looking beyond the future horizon. Assumptions we make today about products, markets and behaviours may not be true in our rapidly evolving future.

Mike Walsh, widely regarded as a ‘futurist’ explains in his book, the ‘Algorithmic Leader’, that to be a successful leader in this new era requires a different approach, a different set of skills and a different way of thinking. We live in the present moment but when attempting to forecast and predict the future, we tend to project forward from what we’re already familiar with.

 “An algorithmic leader is someone who has successfully adapted their decision making, management style, and creative output to the complexities of the machine age” - Mike Walsh (Futurist)

Innovation in project risk software has recently been driven by a small but growing number of ‘algorithmic leaders’ who think computationally and immerse themselves in advanced technical subjects (AI) but who also live between the intersection of technology and business. Some of these leaders, their businesses and products will be explored in more detail in future articles but the following provides an introduction.

The first type of algorithmic leaders are Chief Executive Officers (CEOs) and Chief Technical Officers (CTOs) who have founded AI based software companies which are beginning to excite and inspire the construction and engineering sector. Whilst at first glance many would regard them as pure software developers, their origin story is from the world of project management. They have been motivated to overcome the challenges, obstacles and frustrations they faced when working on past projects. This obsession and passion drove them to innovate. This adds to their authenticity and credentials as they have lived the same issues that their potential future customers are currently facing or might face.

These leaders were convinced there was a better way to avoid repeating historic levels of poor project performance and one consistent theme binds them together. They all recognised that new technology could play a vital part in improving how data could be used to support more informed and efficient decision making. They looked beyond the present and reimagined what future customers would need to better manage project risk and reduce uncertainty in this new technological landscape and then set about experimenting, developing and delivering on their vision. In effect these pioneers realised the best way to predict the future was to create it!

Whilst they have trail blazed the industry, have first mover advantage and have already experienced commercial success by building a customer base, other potential rivals or alternatives are now quickly emerging on the scene which may begin to crowd the market over time. Already we’re beginning to see similar marketing claims and value propositions but at the same time ‘one size may not fit all’ so it’s more than likely we’ll see variation in product offerings.

Historically, it was easier for the industry to understand the unique selling point of a risk software product, what software services were being offered and how it compared to alternatives. There were less significant technical differences to understand. However, with the growing number of AI tools coming onto the market it’s not easy for those unfamiliar with the technology to navigate the options and have confidence in picking the right solution. Even experienced risk practitioners may not have had the opportunity yet to learn about these new tools and understand how they work to help reduce project risk.

A key challenge for AI based risk software vendors will be in explaining the so called ‘black box’ technology and benefits sufficiently well to both technical and non-technical audiences to give them confidence to invest or at the very least trial their products. They need to work hard to help shape how customers perceive what ‘value’ looks like from a risk software product and the focus and value shouldn't be seen through a single lens. If the leadership in client-side organisations think strategically these tools may offer new opportunities to become levers to drive broader organisational change, nudge behaviours and ultimately improve efficiency and project performance.

In order to do this well they will need to be educated appropriately but also have access to suitable expertise to offer sound advice. This may mean recruiting a more diverse set of people who speak and understand the language of this new technology such as data scientists but given the high demand for this skill set it may not come easy, especially in an industry where these types of people would not normally gravitate toward or be familiar with. The alternative some organisations have resorted to is to bring in external consultants to help lead and expedite change.

A different and special kind (my favourite) of algorithmic leader has emerged from cyber space! The Project Data Analytics online community have been influential in the AI thought leadership space. The collective group is energetic, creative, innovative and are not only leading the charge but unlike the professional organisations and institutions, they offer a more hands on practical demonstration of how new technology can help solve problems within a project and risk management context. They also publicly (freely) share and promote their approach and learning online and are not locked away behind commercial fees or subscription models.

The group has amassed 7,500 members across many industries and roles now which is an incredible intellectual resource and network to influence change. It has grown organically using the power of social media and online forums to learn, share and connect with people interested in understanding what can be achieved for the benefit of all. They regularly host ‘hackathons’ in a competitive format where a set of real life project-oriented problem statements are defined, ideas are developed, experimentation occurs and initial solutions developed within literally a few days. The business world is changing fast and this group was born in response to the need to adapt and evolve quickly.

Traditional forms of risk management education and training offered by professional organisations have not been responsive enough to keep pace with technological disruption as syllabuses still remain rigid. In addition, the lag time between textbook and application of theoretical knowledge is woefully long if not ever achieved. I do question the value of the current educational system and whether it's time for a more fundamental rethink. I will come back to this point in a future article. The approach taken by the Project Data Analytics community showcases a refreshing way of solving business problems with the power of the group collective. Bringing together a diverse spectrum of people ranging from data scientists, business analysts through to the project management community promotes creative thinking and innovation as evidenced by their achievements to date.

There are many other individual and group based algorithmic leaders out there but an important observation is that they are now collectively beginning to ‘disrupt’ not just the traditional risk software market but fundamentally how organisations think about risk, manage risk, create more efficient processes, make better and quicker decisions, communicate and collaborate. There's a lot to chew on there but I do believe we are at a turning point where we have a great opportunity to improve the way we manage project risks using technology but in a mindful and practical way rather than rushing in head first or jumping on the first bandwagon!

I'm really excited about writing my next article which will be my first attempt at diving deeper into understanding and appreciating an algorithmic leader. I'll be exploring an AI software company which seeks to improve the way in which projects forecast outcomes to manage risk and reduce uncertainty.

___

After writing this article I took a moment to reflect upon what leadership actually means to me and began by searching for a definition. Many were very academic and generic, however the following is one that struck a chord:

"Leadership is a process of social influence, which maximises the efforts of others, towards the achievement of goals" - Kevin Kruse (Forbes)

By writing these articles my aim is to explore and understand where opportunities exist to use new technology to maximise the effectiveness and efficiency of the way we make data driven decisions on projects. I believe my "social influence" is the way in which I create content to share knowledge and new ideas with as many people as I can and hopefully stimulate discussion and constructive debate in a positive and thoughtful way. Ultimately I want to help other risk practitioners understand and learn how they can help their project stakeholders reduce risk and uncertainty to improve project performance and outcomes.

As always I welcome your thoughts, comments or opinions and ask that we remain respectful of differing perspectives.