If only 7% of organisations are dynamic, we need to focus on creating more transformative structures rather than just sprinkle AI onto existing systems.


We are living through weird times in business and geo-politics, and it is hard to watch long-lived institutions, companies and even countries walk into existential threats because of the myopic, short-term strategic thinking of their custodians, and even harder to watch them walk away with their personal payoffs, ignoring the wreckage they leave behind.

But this is not just a bad actors problem – it is the result of poor institutional and organisational design that allows for (perhaps even incentivises?) management value capture, whilst suppressing the innovation and transformation needed for organisations to adapt and thrive.

Can we automate management?

Given the enormous cost of senior management compensation plus the occasional failed big bet on M&A activity designed to temporarily juice stock prices and bonuses, it is not surprising that some are asking whether the automation of jobs might also start at the top, rather than just the bottom, as this provocative piece from The Hustle outlines. This article speculates that machine learning + decision intelligence could potentially automate a significant proportion of the basic decision making and political compromises a CEO makes at the top table.

But of course, there are also some great examples of CEOs accelerating rather than hindering organisational transformation, such as Satya Nadella at Microsoft, whose most recent letter to shareholders once again talks about the need for the company – not just its products – to embrace constant transformation as we enter another paradigm shift with AI:

This next generation of AI will reshape every software category and every business, including our own. Forty-eight years after its founding, Microsoft remains a consequential company because time and time again – from PC/Server, to Web/Internet, to Cloud/Mobile – we have adapted to technological paradigm shifts. Today, we are doing so once again, as we lead this new era.

Bridging the Exponential Gap

The frustrating thing is that despite the huge cost of very visible management failures such as the German automotive sector’s collective capitulation to Tesla and Chinese electric vehicle innovators, or the banks and pension funds that had no game plan for high interest rates or yield curve inversion, we already have provably better models for creating and running dynamic organisations.

Josh Bersin’s latest research for Deloitte suggests only 7% of organisations are dynamic in the sense of continuously transforming to drive exponential outcomes. But this is not just a question of missing exponential opportunities – it is also about avoiding the threat of the exponential gap, as Azeem Azhar writes about in his book, where linear responses to exponential change mean that by the time a company or sector realises they are behind, it is often already too late to catch up.

At the business end of innovation and change, we have much more evolutionary and adaptive models, such as Rendanheyi, which can increase an organisation’s capacity to sense and respond to change or new opportunities. In HBR, Kaihan Krippendorff and Claudio Garcia recently asked Is Organizational Hierarchy Getting in the Way of Innovation?, showcasing the way that Haier’s model is having a positive impact even in the USA with GE Appliances, and highlighting four key elements of this organisational design model:

  1. Treat employees like intrapreneurs

  2. Work in smaller independent units (Micro Enterprises) rather than business units

  3. “Manage” MEs with decentralized marketplace structures rather than centralized authority

  4. Let MEs choose which support services (e.g. R&D, finance, IT) to work with

Process productivity: Collaboration + Automation

But what about the heavy layer of support services, back-office functions and the layered bureaucracy that sits underneath and requires so much manual management? We also have plenty of interventions and solutions for this area of the organisation that can remove a lot of cost from our businesses, whilst making them more agile and responsive to change.

There are two approaches that go hand-in-hand here: (i) collaboration around process integration; and, (ii) greater use of automation.

We have been patiently working to improve the collaborative potential of the workplace for longer than I care to mention, but the state of collaboration in the median organisation today is aptly summarised by this tweet:

We are seeing progress, of course, but it needs to speed up.

In the New Yorker, Cal Newport writes about the increasing value of collaboration software and its impact on productivity by using technology to improve workflow rather than just work, likening this to Henry Ford’s innovation in the production line, rather than just the car as a product:

What’s striking about this new generation of productivity software is not so much what it does, but what it doesn’t do. Until recently, most business applications focussed on providing faster and more powerful versions of the tools that knowledge workers were already using to accomplish their daily tasks – electronic spreadsheets were better than paper accounting ledgers, e-mail is better than fax machines. The new productivity services, by contrast, can’t be used to directly execute work. Their purpose is instead to help better organize these efforts.

This may not be news to any modern, digital organisation that takes this way of working for granted, but it is orders of magnitude more productive than those corporations still communicating and coordinating through the medium of nursery school PowerPoint pictures shared using the turn-based combat of passive-aggressive email.

And once you connect work processes better in this way, and inevitably uncover ways to optimise and organise them, then opportunities for automation are clearer and easier to address. We have a team development process we use, where we help teams think about their deliverables as services, identify how to standardise and shape demand, and then think about the service components and process design, and how these can be automated to free up more time for exception handling or continuous improvement. Collaboration and automation are not two competing approaches to process productivity – they can and should be used together.

If we want to create better workplaces, improve the experience of work and also reduce the crippling overheads associated with traditional hierarchical management and coordination, then this is a better path to take rather than just injecting AI magic straight into the existing system, which would be another example of replacing work tools rather than improving workflows or ways of working.

The reluctance of larger or more traditional organisations to embrace this challenge is partly why it is mostly tech companies that are able to make the lower-cost, higher-satisfaction model of hybrid and remote working succeed, because they have the systems in place to coordinate work online, such as Dropbox who operate a 90% remote / 10% off-site together mode of operations:

“Since our transition to becoming a Virtual First company, staff retention is at an all-time high and 70% of recently hired employees cite Virtual First as the main reason they applied for a role at Dropbox.”