Digital transformation has been on the minds of the C-Suite for nearly two decades. In the late 90s, the dotcom boom offered a guiding light for businesses, new and old: the future of business is digital, ‘get on board or get left behind’, was the message.
And while that bubble soon burst with catastrophic consequences, the message remained. It’s hard to argue it was wrong. The computer is our tool of choice, mobiles mean we can be ‘always on’, and digital touches every aspect of our personal and professional lives.
But businesses are still way behind where they need to be. Forrester’s report The Sorry State Of Digital Transformation In 2018 – a survey of 1,600 business and IT decision-makers in North American and European enterprises – threw up some surprising stats. 45% of companies haven’t made any investments in SaaS, 56% of firms are transforming but their level of investment and scope of transformation are still mostly small, and only 17% are investing in artificial intelligence.
AI transformation, however, should be on everyone’s lips. It represents the biggest opportunity for competitive advantage, for productivity and for future growth.
AI right now
The AI revolution, of course, isn’t one of dramatic change: no-one has turned up work to meet their new robot colleague. The changes have been iterative and subtle with increasingly adaptable algorithms capable of learning from vast data sets and making decisions based on learned experience. In this definition, there are pockets of exciting innovation, research and development happening already.
Online supermarket and delivery service Ocado, for instance, is using AI to detect fraud in online grocery purchases. The system identifies orders that are delivered but not paid for and analyses data from past orders to determine whether they are the result of malicious intent. The supermarket claims AI has improved its precision in detecting fraud by a factor of 15.
You may like: Can AI destroy bias and improve diversity in recruitment?
Aside from fraud, the retailer is also looking at AI to automate management of the thousands of customer service-related emails they receive, and is even replacing its barcode scanning with smart vision in its warehouses, which it hopes will streamline its warehouse and delivery processes.
Japanese insurance firm Fukoku Mutual Life Insurance has gone a step further, replacing nearly 30 percent of its payment assessment department’s human staff with IBM’s Watson Explorer AI. They estimate this will increase productivity by 30 percent and save around 140 million yen (£977,000) a year in salaries – although it costs £1.4 million to set up and £105 million a year to maintain, so time will tell whether that’s a good idea.
On a smaller scale, Virgin Holidays has implemented AI for marketing purposes – aimed at improving open rates for their newsletter by analysing the performance of previous subject lines and tailoring new ones accordingly. Saul Lopez, Customer Lifecycle Lead at Virgin Holidays, says the change has led to a two percent increase in open rates. It might not seem like much, but for a company that sends around 22 million emails a year, this represents a significant bump and, according to Saul, “that uplift has generated several million pounds.”
So far, so promising; but AI is riddled with issues and challenges – not least debugging. As Gordon Cooper, product marketing manager for Synopsys‘ Embedded Vision Processor, told Semiconductor Engineering: “If you’re training a network, the attraction is that you can make it faster and more accurate. Once you train a network and something goes wrong, there is only a certain amount you can trace back to a line of code. Now it becomes a trickier problem to debug, and it’s one that can’t necessarily be avoided ahead of time.”
Nevertheless, with research groups like Accenture estimating that AI has the potential to boost rates of profitability by an average of 38 percent by 2035, AI is still worth the investment. But how do you make sure your business is thinking AI first?
Thinking AI-first
AI is a tool. A (very) smart one, but a tool nonetheless. Its purpose in business is to solve problems, and it can help with everything from productivity to promotional campaigns. But it isn’t an overnight solution, it takes work to implement. So how do you create a business that thinks AI-first?
– Education – There is a lot written about AI in the press, and the things that make headlines are always around robots coming to steal our jobs. Of course, companies like Fukoku Mutual Life Insurance has made that leap, but for the most part, AI is designed to make our working lives easier. From top to bottom, all staff within a business need some demystification of what AI is, and what it definitely isn’t. When people can see the potential benefits (to productivity for the C-Level and to day-to-day working lives for managers down), getting staff educated and excited is key. AI is an ongoing investment: without buy-in from the top table, an AI transformation programme is a non-starter.
– Culture – It’s not solely about getting people interested, there needs to be a change of mindset in staff too. In some respects, AI is a leap of faith for companies. You may invest in third parties, products or people, but you can’t be 100% certain of the outcome. Creating an AI-first culture demands a declaration of intent from the C-Suite and to get that, they need to be convinced of the validity and opportunity of AI as a growth driver and a business benefit. The rest of the business, too, need to make AI a part of their everyday thinking. If they have process problems more than once, the culture should be to report that as an area of potential improvement.
– Data – Central to the success of any artificial intelligence programme is data. Companies need to get to grips with the sheer amount of data they hold, and structure it in a uniform way. AI is only as good as the data it uses, and if you can’t provide quality data, it won’t work. Part of the culture of the business should be understanding of the importance of data and an agreement on how this data should be collected and stored. If you need an AI in two years, you’ll need to start collecting data now.
Much like AI itself, a programme of AI transformation is not an overnight change in processes, it’s more an ongoing evolution, powered by the right culture and the right education. And introducing AI into a business doesn’t have to be disruptive.
Steps to AI transformation
Ocado’s public-facing business is a grocery shopping site, but Ocado.com is only one arm of the Ocado Group. Also part of the business are Ocado Technology and Ocado Engineering. The business knew that developing IP would give it a competitive advantage. It recently sold its unique, automated picking and packing system to US retail giant Kruger in a deal that saw share prices rocket by 44%.
Of course, AI and robotics are its two biggest focus areas right now – which is why they’re innovating so quickly. They’re not the only ones. Tesco Labs was set up to experiment, of course, but also to build “a culture of innovation across the business.”
“Live demonstrations of how AI could impact the business are invaluable.”
Option one for instituting an AI-first business, then, is to set up an in-house innovation lab – a space for experts to be creative, to experiment, to inspire. Culture is undoubtedly an important part of change, but live demonstrations of how AI could impact the business are invaluable. AI is a unique specialism, however: to get the most out of an AI lab, you’ll likely need to hire in.
This option has long-term business benefits, of course: creating IP is a huge asset to the value of any business. Having an in-house expert is almost the only way to do that. As Uber’s Head of Machine Learning, Danny Lange, told Computerworld UK: “We have really had machine learning for a while but it is something that can be really hard for software engineers to get. So we have created machine learning-as-a-service inside the company as a cloud service.”
The other option is partnerships. While Ocado and Tesco have set up their own initiatives, M&S has partnered with Microsoft to bring artificial intelligence into stores. Rolls-Royce, meanwhile, has combined both of these, creating an ‘innovation factory’ in house, but inviting third parties like the Alan Turing Institute to bring their specialist expertise to the table.
The final option, of course, is simply to buy in AI capability. US-based business management software company Domo announced its machine learning capability in March 2017. The AI (which they named Mr. Roboto for the early-80s pop fans out there) uses machine learning algorithms and predictive analytics to power business insights, recommendations and alerts for business decision makers across the enterprise.
Mr. Roboto plugs straight into the business cloud, and helps business users get faster access to the benefits of machine learning, AI and predictive analytics without having to hire a legion of data scientists to implement those technologies into solutions and apps.
For all the headlines it generates, we’re only beginning to see the first fruits of the benefits of AI. As these systems get smarter, as data becomes more ubiquitous and powerful and as society changes, so businesses that invest early in transformation will see the greatest benefits. AI transformation, like digital transformation, isn’t a project to complete, it’s an ongoing process of discovery. And any good journey starts with a first step: businesses have to decide whether to step forward into AI, or march on the spot and be overtaken.
