Anyone interested in how businesses can start making better use of the explosion of data that’s available for capture, analysis, and insights, has probably heard about the data skills crisis.
Professional, experienced experts in machine learning and other advanced forms of data analytics are in short supply. In the UK this year, government researchers found that almost half of businesses were struggling to recruit for roles that require data skills. Often, talent is snapped up by big tech or finance as soon as it hits the jobs market. This makes it difficult for smaller companies, or those in industries where adoption of AI and analytics is less mature, to make a play for their slice of the $15.7 trillion that AI could add to the global economy by 2030.
It's important to note that "data skills" don't just mean hard analytical skills. As well as number-crunchers, businesses need people who understand how to take insights and act on them, as well as those who can specialize in interpreting insights in ways that anyone can understand (these are known as “data communicators”).
Luckily, businesses that want to capitalize on the data available to them but don't necessarily have a need to hire in-house PhD-level machine learning experts to build bespoke infrastructure have a number of options open to them. Broadly, they can be broken down into two strategies – developing in-house capabilities and outsourcing the solutions you need.
Developing in-house data skills
This year, data scientist was once again ranked by recruitment consultancy Glass Door as the best job in America, with an average salary of $108,000. Scanning the vacancies, there are opportunities in every industry, with Starbucks, Verizon, Pinterest, Barclays, and Toyota among those currently hiring.
So recruiting is certainly an option that can be considered, with the provision that you will be competing against some of the biggest and most successful companies in the world! As well as competing on salary, you’ll have to be certain you can offer the culture, lifestyle, opportunity, and bonus benefits that come with working for companies like those.
Another option is upskilling an existing workforce. This is increasingly viable thanks to the growing number of “ready-made” data science solutions coming onto the market. If tools that do the job you need are readily available, challenges are likely to revolve around fostering a culture of data-driven discovery and literacy among existing teams. In retail, for example, companies are training shop floor staff to work with analytical tools that let them offer better service to customers when they walk through the door. This might mean having access to analytic insights on what customers have previously bought, thanks to loyalty programs. In media organizations, journalists and producers are increasingly being trained to use statistics and metrics to create content that matches their audience’s interests. Researchers and lawyers can be introduced to AI and analytical tools that cut down the “leg work” of ingesting large amounts of documents and information that is often a necessary part of their jobs.
Outsourcing your data expertise
Once again, this is a large and fast-growing market. Plenty of service providers offer analytics services, either through an entirely outsourced model or on a consultancy basis. Companies like Google, Amazon, and IBM offer a full suite of services in this area, often alongside the cloud infrastructure services that are needed to put them to work. There are also any number of specialized data consultancies, often focused on providing solutions within specific industries. One thing to be careful of is that, generally speaking, this industry is largely unregulated, so if you're considering working with someone, you should always ask for examples of work they have previously carried out and what sort of results they've achieved. Consultants and third-party service providers might be business-focused, where they will first look at the business problems you're trying to solve and recommend potential solutions. On the other hand, they might simply specialize in a specific technology or methodology and will expect you to have worked out whether their approach is the right fit for your problem.
Another option that might be worth considering is crowdsourcing. Kaggle is an online platform that calls itself the "world's largest data science community." Organizations use it to host competitions where they set problems, and members of the site compete to provide the best solutions. It’s been used by companies including Google and Netflix to tap into the potential of millions of “armchair data scientists” who either enjoy solving problems in their spare time (for a reward, of course) or may even be looking for work in the field of analytics. During the Covid-19 pandemic, Kaggle has even used by the US government to crowdsource insights to help track the spread of the virus. It might not work for everyone, but if you have a problem that catches the imagination of the community, it could be very effective!
Why not let machines themselves do the hard work?
Computers are great at crunching numbers, work incredibly quickly compared to humans, and in the era of AI and machine learning, are becoming capable of learning and coming up with their own solutions to problems. Maybe they can work out the best way for your business to use data by themselves?
This is the thinking behind the AutoML (automated machine learning) family of applications, designed to (within reason) enable just about anybody to build, train and operate AI projects. Providers including DataRobot and Alteryx are among those that offer services promising to automate the entire process of data-driven discovery. There is also a growing “no-code” and “low-code” movement in AI, promoting the development of software that can create algorithms from input, including drag-and-drop interfaces or even spoken language. Solutions such as these have a great deal of potential when it comes to helping organizations overcome the challenges posed by the AI skills crisis.
Building data competencies within your organization is just one of the subjects covered in-depth in the new edition of my book ‘Data Strategy: How To Profit From A World Of Big Data, Analytics And Artificial Intelligence.