When a new technology or way of working emerges, the first instinct of many organisations is to appoint a Chief X Officer, which provides the comfort of knowing somebody has ‘got this’ and also that there is ‘one throat to choke’ in terms of accountability. And of course, this role requires a safe pair of hands, which is prioritised over domain knowledge.
But there is nothing as anti-innovation as turning a new function or a technology into a C-Suite empire that tends to become risk-averse and budget-maximising. If the only way you can coordinate across functions is from the very top of the organisation, then you have much bigger problems than you realise.
Remember the Chief Electricity Officer, Chief Phone Officer and their ancestor the Chief Paper Officer? Of course not. The same will be true of the Chief AI Officer.
In a recent Financial Times piece on the topic, which suggested only 21% of firms plan to take this approach, Tata Consulting CTO Harrick Vin said:
“It isn’t about having one person and one role. Every function — whether sales, marketing, or software engineering is getting redefined leveraging AI.” Like innovation, AI is “everyone’s business. One person can’t be an expert.”
Don’t ask ‘who owns AI’. Instead, focus on developing new capabilities and services that the whole organisation can build on to create new experiences, efficiencies, products and value.
Until recently, the starting point for many large organisations was that anything more technological than a digital watch is owned and procured by IT, but this is not sustainable with multi-layered capabilities like AI that include components as varied as data, models, training, software, adoption and human supervision.
We are already seeing signs of budgets for AI initiatives shifting out of IT functions, which is good; but in conditions of uncertainty and rapid change, what we need is not pre-assigned budgets but internal investment funds and shared, agile processes for resource allocation based on discovery and experimentation, not the tired old trope of business cases that pretend they know the one sure-fire solution (and what it will cost) before they have even entered the game.
Lessons from Digital Transformation
A decade ago, we faced a similar challenge with organisations being structurally ill-equipped to deal with digital transformation, which includes org and process re-design, ways of working and management methods as well as technology. Our team found that the most successful solution was to convene a digital change leadership group comprising all stakeholders who might be involved in procuring technology to agree standards and shared architectures, and to make purchasing, development and deployment decisions that lead to shared capabilities the whole organisation can build on. This group would also typically support a network of change agents or guides whose brief was to help, teach and connect between local areas where new ways of working were being trialled.
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4421d116-7ff6-42db-8f8d-90dd3c5a171c.heic)
Image created by OpenAI’s ChatGPT
See here for a detailed exposition of this idea and slides from a talk about it:
Where digital technology can play the most important role, I think, is in automating, orchestrating and connecting the many service and processes that make up the value chain of an organisation to create a platform on which agile teams and people can be free to work in a more autonomous and creative way – a structure that is more machine-like and automated at the back-end to enable the organisation to be more human and adaptive at the front-end.
But the key change was shifting from a point solution mindset – “buy this, build that” – to a shared capability mindset, and this is where we first started using capability mapping to create common ground for people to work on together, which also had the benefit of uncovering a lot of duplication and orphaned systems that were no longer needed. Investment proposals were encouraged to demonstrate how they add or improve capabilities, or better still where they open up a new region of the map that was connected, but unexplored.
This is why we focus on capability mapping as a process to develop a shared understanding of business goals for technology between different stakeholder groups, and it is why capability maps are a useful touchstone artefact to keep cross-functional initiatives focused and on the same page.
Give the potential complexity of mapping capabilities and their relationships across a large organisation, this is now becoming a more distributed process, using simple bots that can pop up in team chats to use conversational information gathering in situ. So in the light of OpenAI’s GPT-4o announcement of multi-modal interaction, we could imagine the capability mapping bot to be in the room with the digital leadership group to answer questions such a ‘what do we have in predictive analytics?’ or ‘which teams are advanced enough in GenAI use to form a cross-functional task force to help others?’
CXO Perspectives
Whilst adding new CXO roles for every new focus area is not a good idea, foundational CXO roles overseeing finance, marketing, operations, people, etc., still have an important role to play in overseeing the business, and with the right approach, they can act as catalysts and connectors, rather than top-down managers.
All these roles are important stakeholders in the AI adoption process.
We need CFOs to take responsibility for their tech stack, and especially their data, which will be a key source for AI training and sense-making. And we need their grasp of numbers, trends and data to help guide analytics and business intelligence services that are becoming smarter and more automated. So many areas of book-keeping, treasury management and procurement will be made easy by automation and AI – see Gartner Predicts Half of Procurement Contract Management Will Be AI-Enabled by 2027, for example – that finance and related functions will have more bandwidth to do smart things with their data and models.
We need CHROs and CPOs to also step up when it comes to technology, experience design, and service design for employees. As with CFOs, in the recent past they could just buy or rent a big one-size-fits-all people platform and hope that it would be enough. But now we need them to really own the key people and performance data in their domain, ensure it is safely available internally, and also represent the interests of the people in the business when it comes to automation, new ways of working and upskilling to meet the needs of a changing business.
We also need COOs and CEOs to understand the art of the possible and encourage team and function leads to own and improve the digital workplace experience of their teams, and to experiment with centaur service teams, augmentation and supportive automation, because most of the learning and innovation is likely to emerge from very specific local needs.
Just as we have talked about conversational GenAI giving all of us the ability to code with a co-pilot, these technologies can also help leaders and managers to embrace their role as organisational architects, and as service and experience designers to create environments in which their people can thrive. If you can imagine it and describe it, then you can make it happen – as I wrote in our launch post for this academy:
The programmable organisation, in which small autonomous teams and groups using smart systems and structures can achieve more than entire divisions of a C20th corporation, might become a reality sooner than we think, especially in areas of high competition and dynamism.
FYI we are running customised workshops and immersions where leadership groups can explore these questions, and the 'aha moments' are amazing
Only Connect
There is a growing consensus in our field that the best template for a connected company – and a necessary prerequisite for successful deployment of AI – is a platform architecture that hosts common services (not just software) and makes them available for teams to build on.
Dion Hinchliffe shared one of his famous diagrams on the subject last week, and Boundaryless published a platform manifesto that distills their research into platform organisations and how they support entrepreneurial ecosystems.
A cross-functional digital leadership group focused on defining and designing the capabilities and services the organisation needs is both a smart way to manage your platform roadmap and also a good way to govern a platform organisation as it evolves.
What’s interesting is the lack of tools and language that non-technical functions have for this kind of management effort, beyond clumsy old notions of project management, initiatives and task forces.
If we are to bring stakeholders together and apply some rigour to organisational development to improve our readiness for AI and smart tech, I would imagine more businesses will start to use the language and concepts of tech teams (roadmaps, service definitions, interfaces, test-driven development, devops, etc) to run and improve non-technical functions as well, just as we have seen ‘business agile’ borrow from software development.
Next week, Cerys will share our digital leadership group playbook and example terms of reference, to help you try this powerful approach in your own organisation.