Artificial Intelligence (AI) is a high-stakes business priority, with companies spending $306 billion on AI applications in the past three years. Companies that scale AI across a business can achieve nearly triple the return from their investments. But too many companies aren’t achieving the value they expected.1
Scaling AI effectively for the long term will require the professionalization of the industry. Stakeholders—from practitioners to leaders across the private and public sector—must come together to distinguish clear roles and responsibilities for AI practitioners; demand the right level of education and training for said practitioners; define processes for developing, deploying and managing AI, and democratize AI literacy across the enterprise.
By formalizing AI as a trade with a shared set of norms and principles, companies will be poised to achieve more value from AI. They’ll be set up to create clear accountability which in turn helps avoid risks like bias, under delivering to clients, and other unintended consequences.
That’s why in professionalized fields such as medicine, construction and even food service, an inherent level of trust exists between customers and the businesses (or practitioners) that make up that industry. This trust is born out of standards that set expectations for everyone involved.
For example, you understand that architects, electricians and other construction professionals know how to build a house. They’ve had requisite training and understand their roles and responsibilities, safety standards and protocols to follow throughout the construction process. It’s unlikely that you’d trust a “citizen architect” to build your home in the same way that you wouldn’t visit a “citizen doctor” when you get sick.
Yet increasingly, companies are bolstering their core data science teams with “citizen data scientists” (or, people who create models using predictive analytics but whose roles are outside of the data science field), without providing them with necessary guardrails and standards to enable success. Even among trained and credentialed data scientists, there are varying degrees of standards. Beyond needing formalized and standardized training, organizations may find that these practitioners are working in siloes and can’t deliver on the promise of AI. Real value can only be realized when trained AI practitioners are working hand in hand with the business to accomplish their organization’s goals, and those interdisciplinary teams are guided by standards, rules and processes.
Only then will businesses be able to deliver the end product or service safely and predictably, thereby earning the trust of customers and raising standards for quality innovation and applications.
The benefits of professionalizing AI
Why professionalize AI now
Three out of four executives believe that if they don’t scale AI in the next five years, they risk going out of business.2 According to Accenture’s study of 1,500 C-suite leaders, companies that successfully scale AI (to achieve higher returns from AI investments) are employing tactics of professionalization. These “Strategic Scalers” are 1.5-2.5 times more likely to establish dedicated multidisciplinary teams, training, and clear lines of accountability. Professionalization, then, should be seen as a precursor to successfully scaling AI.
The COVID-19 pandemic has further sharpened the contrast between those who have professionalized and scaled their AI capabilities and those who have not. As businesses race to embrace new data and AI capabilities in an attempt to recover and return to sustainable growth, it will be important for these new scalers to professionalize in parallel.
How do companies professionalize AI?
Because AI technologies and use cases are advancing too rapidly for governments and regulators to implement basic industry reforms and standards, organizations need to take professionalization into their own hands. By following these steps to standardize professionals and processes, organizations can better set themselves up to scale AI and, in so doing, make the most of this quickly evolving technology.
Similarly, multidisciplinary teams of diverse perspectives, skills and approaches, must work together to innovate and deliver AI products or services. As Accenture research shows, 92% of Strategic Scalers leverage and embed multidisciplinary teams across the organization. And, 72% say their employees fully understand how AI applies to their roles3.
Strategic Scalers demonstrate the importance of distinguishing clear roles among multidisciplinary teams. They quickly stamp out redundant responsibilities and clarify individual remits. These teams, often headed by the chief AI, data or analytics officer, include data modelers, machine learning engineers and data quality specialists, to name a few. The mix and the ratio of roles is going to depend on the use cases you’re pursuing at the time and will vary from project to project. Tapping into partner knowledge and/or establishing a blueprint for how teams should operate will help this process become more turnkey over time. But one thing remains true across all projects -- you need to establish ownership and expectations from the start.
Case in point: At one factory, a pump was wearing down faster than expected, and nobody understood why. While an AI monitor indicated something was wrong, it couldn’t pinpoint the mechanical issue. The engineering team investigated further and discovered that the pump was leaking oil on the floor at night and losing lubrication. But a staff person simply mopped up the mess and replenished the oil each morning, which meant there was no visual indication of leaks or of the dropping oil level. Having an engineer on the team eventually ensured that the AI application considered all the necessary conditions to determine what was wrong.
Another example shows the importance of establishing responsibility within a team—the AI work of one team at an oil and gas company stalled for 10 weeks simply because they couldn’t decide who owned the project’s data.
Companies may struggle to professionalize for a range of reasons, including the inability to address skills gaps. Even if they can pinpoint the skills they are lacking, it can be challenging to find and hire enough people with the required experience to fill that gap at speed. One solution is to partner with (or even acquire!) a professionalized firm with not only the right skill sets, but also teams that have proven, production-oriented methods. Accenture, for example, has acquired several professionalized software and service companies in the last year to reinforce its position in the market.4
2. Demand education and AI training.
As organizations invest more in their AI and data capabilities, employees understand the growing influence of these technologies on their companies and careers. But despite their best efforts, many of these employees will not have the right training and qualifications to work effectively with AI.
It’s important for organizations to establish education and training requirements for their AI practitioners. Data scientists have varying qualifications, and not all have sufficient training in mathematics or computer science for AI projects. Even an employee with a Ph.D. might have studied a narrow field that isn’t relevant to a particular company’s needs. On the other hand, companies that can scale AI successfully make sure they have people with the right mix of skills and qualifications. Of Strategic Scalers, for example, 70% say their employees have formal training around data and AI5.
To establish an effective professionalized workforce, it’s up to companies to assess which skills they need, their workforce skills gaps and the qualifications of their talent and match them to the appropriate roles. While one way to address the skills gaps may be to acquire or work with a professionalized firm, another approach is to establish an academic partnership. Companies can partner with research and academic institutions to reskill employees or strengthen their future talent pipelines.
For example, Accenture is building a strategic relationship with UC Berkeley’s Institute for Data Science (BIDS) to advance the field of data science. The program is designed to create opportunities for researchers, students and Accenture’s Applied Intelligence practice to work together to explore complex problems facing society and to learn from each other while doing so6.
When recruiting new talent, companies often use technical screening to assess whether an applicant has the required level of knowledge to fulfill a role. To create greater accountability and confidence among AI practitioners, organizations should implement regular assessment points throughout an employee’s career to test their knowledge and to continue their technical education. Like construction workers or medical professionals who must renew their certifications as techniques change and theories evolve, companies should test and re-test the professional competency of AI practitioners to make sure they are upholding rigorous standards and trust, as well as providing them with the training they need to evolve those standards.
To enable a consistent approach to training, companies should create clear career paths for their AI practitioners. Each career level should have established pre-requistes such as coursework and training to help build necessary skills and proficiencies. These pre-requisites should be shaped by a combination of leaders across technology, data and human resources and even outside counsel from leading academic partners. This transparency and consistency will provide clear educational expectations for anyone working on AI projects – from data architects, to test developers to machine learning engineers. The added benefits of establishing career paths are better talent retention, employee development and a market-leading, professionalized practice.
3. Define AI processes.
While some argue that formalized processes and governance could stifle innovation, our research has shown the opposite. Companies that govern innovation extensively over time said they expected to double revenue growth in five years.7
In professionalized industries, there’s a standard approach to testing and benchmarking during the creation (or optimization) of products and services. Similarly, whether a company is making smart devices or building a data science model to improve the online retail experience, establishing systems and processes to support the development of the AI product or solution allows people to innovate in a predictable and efficient way.
When someone is sick, the patient passes through a number of nurses, doctors and other specialists to diagnose the problem and recommend the safest treatment. Once companies have distinguished clear roles for their AI teams, they should follow the example of the medical profession—establishing defined processes that formalize the development, deployment and management of AI solutions. These should inform how people work together, how they choose technologies to support production of the AI solution and how they interact with those technologies. For example, once a data science group creates a new algorithm, an organization with a professionalized approach to AI would establish a system to test the algorithm to ensure it does what it’s supposed to do in a safe, predictable and consistent way.
One telecommunications company, for example, has mapped out an automation and AI organization structure with clear roles and responsibilities. Its project management office is responsible for specific tasks, including continuous improvements, financial reporting, follow-up and target setting. The company’s working model provides clarity to stakeholders developing AI solutions, from proof of concept to minimum viable product to product solution.
4. Democratize AI literacy across the organization.
While there’s certainly growing interest from leaders to invest in AI technologies, true professionalization will result in (and rely on) AI literacy across an entire organization.
Organizations owe it to their employees and to their bottom lines to provide some form of AI education. We found that 62% of workers believe that AI will have a positive impact on their jobs. And 67% of employees say it’s important to develop skills to work with intelligent machines.8 It’s clear employees recognize the impact AI could have on their jobs and are keen to learn more, and organizations have an opportunity—and Accenture would argue, a responsibility—to enable it.
To start, companies should define the minimum level of AI knowledge they require from their employees. Helping the entire workforce understand what AI is, how it impacts their jobs and how it benefits the company are part of building confidence in AI and driving adoption and usage.
Continuing our earlier analogy, each employee in a hospital plays an important role in supporting patients. From the porters to the technicians, all individuals understand how they contribute to the health and safety of patients even if they don’t have an advanced medical degree. Likewise, mandating a basic level of AI literacy in every role will better set up the organization for long-term success in scaling AI.
Democratizing AI literacy could: provide marketing teams with the knowledge to communicate clearly about AI services to customers and to understand how to sell them; empower legal teams to have a stronger grasp on regulatory implications and ensure an organization isn’t exposed to costly risk; enable recruitment to refine their hiring processes and requirements when seeking potential AI practitioners.
It should go without saying that building AI literacy needs to be a cross-functional focus. Cultivating confidence in AI through democratization needs to span well beyond the CAO, technology leads and their teams. As we discovered in our research, when you cut the distance between the C-suite and the AI practitioners, you improve your odds of delivering value. When everyone has a deeper understanding of AI, they not only perform their jobs more productively, but will also be able to better support scaling AI across the enterprise.
So now you see, professionalization is part and parcel of scaling your AI and data practices. And if scaling your AI and data practices promises to make you a more connected and agile enterprise, the question then becomes: What are you waiting for?
Companies that are ahead of their peers are not waiting for their industry or regulators to take the lead on Professionalizing AI. We’re working with them to distinguish clear roles, demand education and training, define processes and democratize AI literacy within their “walls.”
Are you ready to go AI pro? Let’s talk.
1 Accenture “AI: Built To Scale” November 14, 2019. Pg 3, 6.
2 Accenture “AI: Built To Scale” November 14, 2019. Pg 3.
3 Accenture “AI: Built To Scale” November 14, 2019. Pg 15.
4 Wall Street Journal, “Accenture Looks to Boost AI Capabilities Through Acquisitions,” by Jared Council, June 22, 2020.
5 Accenture “AI: Built To Scale” November 14, 2019.
6 Pioneering the future together: Accenture Applied Intelligence and Berkeley Institute for Data Science join in new program.
7 Accenture. “Governing Innovation”. January 14, 2020. Page 22.
8 Accenture. “Ready. Set. Scale.” December 4, 2019. Page 21.