The Next Organisation Chart
Earlier this year, a British Chamber of Commerce report revealed that most UK businesses use AI as a productivity tool. Only advanced adopters are considering redesigning their operations. What does that even mean?
Business management theory focuses on span of control. That is the number of people that one person can manage. Years ago, the CEO at Lloyds Banking Group told me about his rule of six. It was the system he learned in Spanish retail banking, long considered Europe’s most efficient.
The problem with span of control is that the larger a business becomes, the more layers are required to run it. Managers gather information from below and pass on decisions from above. The more layers that information must pass through, the slower it flows.
Using technology to redesign organisations is about replacing layers of human messaging with AI that understands how the business works. Two examples serve to illustrate this.
Block Behaviour
At the same time the BCC report came out, US payments company Block made headlines saying it was laying off 40% of its staff. Sceptics suggested this reflected cash flow issues rather than reorganisation. CEO Jack Dorsey then published a blog explaining what he was doing.
Dorsey has two requisites for redesigning a business. The first is a model of how the business works. The second is a rich flow of customer signals about what they expect. Everything in the business is recorded and analysed, from early-stage discussions through to progress with solutions.
By analysing patterns in customer behaviour, Block’s intelligence layer can adapt to individual customer needs. It spots seasonal slowdowns in corporate customers’ sales and suggests appropriate financial support. It sees when an individual has moved to another city and adjusts their products and promotions accordingly.
The intelligence layer coordinates the company and speeds up decision making and market reaction times. Humans focus on ethical issues, novel requirements and high-stakes situations.
In practice, this means there are three roles. The first is the specialists making decisions in their part of the business. The second is the cross-functional problem solvers who deliver strategic improvements. The final role is the player-coach, such as an executive who both codes and develops talent.
Dorsey asks what your organisation understands that others struggle to understand and whether your understanding gets deeper each day. That’s a way of asking whether you respond to hard-to-read customer signals or build based on managerial intuition. Firms doing the former will thrive. The latter will use AI to cut costs, while competitors with better organisational intelligence pull ahead.
Reimagining Human Resources
IBM has spent a decade rebuilding internal functions around AI. For example, its sales process used to be from salesperson to pricing to legal to finance to contract administration. Now AI drafts proposals, analyses contract clauses and suggests pricing. Billing is automated and humans handle exceptions.
In HR, IBM wanted lower spending without losing the capacity to grow. It replaced the helpline phone number and email address with AskHR. 21,000 frontline managers were told that this was their sole source of support. Internal satisfaction fell to -35 on a scale from +100 to -100.
Managers helped refine the system by using it every day. Automation offered immediate execution of repeatable tasks, such as employee transfers. Generative AI was introduced to coordinate the multiple autonomous agents carrying out those tasks. Access to human support improved.
AskHR now handles around 39,000 messages a day and customer satisfaction is at +76. IBM says if it did the transformation again it would start smaller and iterate. It also notes that culture is critical. Leadership must buy in and empower employees through skill-building and responding to constructive feedback.
Today, IBM’s HR executives spend their time on workforce planning, executive coaching, organisational design and difficult employee relations. AskHR automates more than 80 processes, answers 94% of common questions and has reduced support tickets by 75%. This helped to cut HR’s operating costs by 40%.
A Manageable Pace of Change
The BCC survey showed that 95% of companies using AI in the UK reported no impact on headcount. More significantly, only a fifth of the advanced adopters reported reductions. They did, however, report productivity improvements.
These gains came from using AI for workflow automation, operational decision making, forecasting and customer service systems. This led to new job roles and new skill requirements.
The survey cautions that most SMEs are at the early stages of adoption. AI is more likely to be an assistant than embedded into core operations. Retail, hospitality and personal services are least likely to have adopted.
These are industries where it is hardest to record everything that happens in the business and to determine a consistent course of action from every customer interaction.
AI will not transform the economy overnight. Businesses will reorganise one workflow at a time, at a pace that will be demanding but manageable.
Questions to Ask and Answer
Which of my business decisions are slowed down by unnecessary layers?
What knowledge do we repeatedly use but never capture?
What should my people focus on if AI handles the routine work?