The exponential growth of the digital economy means that leaders who don’t develop a digital mindset will soon no longer be able to lead their organisations effectively. Leaders and those who wish to thrive in organisations need to upgrade their skills and become digitally literate or they will get left behind.
That’s the view of Tsedal Neeley, professor of business administration at Harvard Business School, an expert on virtual working and co-author of The Digital Mindset, what it really takes to thrive in the age of data, algorithms and AI.
Neeley says concern about machines replacing humans is missing the point. “Humans with digital knowledge will replace humans without digital knowledge,” she tells The Irish Times.
“You can no longer survive with low literacy levels about digital technology. You need to understand the language of digital in terms of how it impacts on strategy, operating models, employees, retention and recruitment, stakeholders, products and services – and if you don’t understand it, you will not be able to participate in the digital transformation that is now taking place.”
The days of C-suite leaders drafting in digital experts or recruiting high-powered IT directors to their boards to compensate for their own lack of knowledge appear to over in this analysis. The good news however, according to Neeley, is that acquiring digital literacy is not as difficult as many imagine.
Drawing an analogy with learning a foreign language, she says it is not necessary to be a master or expert, but it is vital to be literate enough to be a meaningful participant in conversations.
A non-native English speaker, she notes, needs to acquire a vocabulary of about 12,000 words to be considered a master of the language. However, with as little as 3,500 words, people can generally comprehend and communicate effectively in a workplace setting.
Minimum threshold
Broadly speaking, people need to follow a “30 per cent rule”, which is about the minimum threshold you need to understand and take advantage of the digital threads woven into the fabric of our lives now.
Digital literacy does not mean you need to master coding or become a data scientist, but you do need to understand what computer programmers and data scientists do, how to make use of A/B tests, how to interpret statistical models and how to get an AI-based chatbot to do what you need it to do.
Neeley says this level of knowledge can be acquired without undertaking a conventional degree-level programme. The important point is getting a baseline on your existing knowledge and finding short programmes or modules to fill the knowledge gaps.
Harvard Business School, by way of example, runs a nine-month course called Harvard Business Analytics. In that time, managers learn how to code and acquire a basic knowledge on statistics and what AI and machine learning can do. It’s just one way in which managers can get up to speed quickly.
Learning to code isn’t essential, Neeley underlines, but it is no harm either. She cites the example of Japanese company Rakuten, a leading global 5G operator. In 2019, it required its entire workforce to learn how to code and gave them six months to do so. Every person in the organisation should now understand the idea of data and the importance of algorithms and statistics, she observes.
This is not a one-off process, however, but one that involves a shift in mindset.
“It’s about things like understanding how to do collaboration differently in a digital world, how to think about data and security and how to make decisions around data, and finally how to think about change at a time of rapid transformation that requires a continuous learning loop to order to continue to innovate and make good decisions.”
One of the ways that getting a better understanding of digital informs better decision-making is that leaders gain a better appreciation not only about the many benefits digital offers, but also about its limitations and deficits and the ways in which humans should best interact with it.
Artificial intelligence trumps humans in lots of areas and part of developing a digital mindset is acknowledging that machines are better than humans at making certain predictions and doing specific tasks. Healthcare provides some great examples.
Potential cancers
Researchers at Seoul National University Hospital and College of Medicine developed an AI algorithm called Deep Learning-based Automatic Detection (DLAD) to analyse chest radiographs and detect abnormal cell growth such as potential cancers. In a four-year study the hospital found that the AI was able to dramatically reduce the number of overlooked lung cancers on chest radiographs, without a proportional increase in the number of follow-up chest CT examinations.
In another case, Google Health creates a machine learning algorithm to identify metastatic breast-cancer tumours from lymph node biopsies. Its unique advantage was its ability to identify suspicious regions indistinguishable to the human eye. LYRA was tested on two datasets and correctly distinguished between cancerous and non-cancerous results in 99 per cent of cases.
The success of the AI creates issues for physicians. As the authors observe in the book: “It can feel threatening to have a machine contradict your diagnoses. That’s where the idea that machines aren’t humans is important. It feels threatening but code doesn’t make threats. It’s just a tool for us to harness.”
Reinforce existing prejudices
The possibility of bias is another issue the authors caution against. A well-intentioned model builder can reinforce existing prejudices if they don’t consider the wider picture.
Consider the example of the city of Boston, which used what appeared to be a well-designed app called StreetBump to solve the city’s persistent pothole problem. The app records accelerometer data from smartphones as a resident drives, producing data that a car just hit a pothole. However, since the residents with smartphones tended to be of higher income, it was generally identifying potholes in more affluent areas.
When Boston’s Office of New Urban Mechanics discovered the data problem, the model was quickly adapted to represent the entire city and the results changed significantly.
Several authoritative studies have also revealed inherent gender and racial biases within facial recognition software or have exaggerated the level of crime in less affluent areas, based on human intervention about the data inputted.
“Police presence in low-income neighbourhoods results in more data points which then puts them in contact with more low-income individuals, which creates more police records, which then have a higher likelihood of accumulating into the chronic offender score.”
Having a digital mindset also recognises that reworking older technological investments is vital, a process described as “investing in technical debt”.
“It’s like home renovations. We’d much rather install new countertops and appliances than spend money on updating the plumbing or electrical. But if we keep spending our money on the fun stuff and don’t invest in the maintenance, eventually pipes will break and circuits will short and we’ll have to take on debt to fix the infrastructure emergency.”
Continuous improvement
Another very tangible benefit from achieving digital literacy is how it facilitates experimentation, a vital way to extract value from data and to support continuous improvement and learning (see panel below).
A digital mindset also involves recognising the issues around collaboration in the workplace. As someone whose work has involved many years studying and writing about the interface between in-person and remote working, Neeley observes that Covid has broken down divisions between office and remote-based workers, especially given the much wider adoption of tools such as Zoom, Teams and Slack, among others.
“What gives me confidence today is that the level of empathy that managers have with remote versus non-remote team members has never been higher than it is now. There’s a lot still to be figured out but managers are better equipped and communication and conversations are taken place on a much more level playing field.”
How experimentation helps drive a digital mindset
Know why: Begin with a testable hypothesis and a clear rationale for how and why an experiment should be done. Then create a learning agenda by outlining an experiment’s key questions, steps and how to evaluate its outcome.
Overcome barriers: Remove organisational roadblocks by making resources and data available to teams across departments and reward teams that experiment. Digital experimentation can lead to increased revenue, cost-cutting, innovation and employee satisfaction.
Value failure: Establish psychological safety by framing experiments as learning opportunities. If an experiment fails, think about the lessons learned and how those lessons can be used to inform future experiments.
Recognise your data assets: The hundreds of millions of data points you have as employees and customers use their digital tools can be turned into digital footprints that can form the basis of strong experiments.
The Digital Mindset, what it really takes to thrive in the age of data, algorithms and AI, by Paul Leonardi and Tsedal Neely, is published by Harvard Business Review Press.
