Winning AI Strategies – Part 2: Established Companies

This article is the second of two on winning AI strategies, focused mostly on established companies.

Traditional companies may have the most to gain from adopting AI, and in some cases pivoting the company it

You may find it surprising, but much of the value from AI will be captured not by big tech companies or startups, but instead by established companies. These companies have the data, from their own operations, their customers and suppliers. And they are rich in resources. In many cases, they are just starting to collect data about their products and services, and how their customers use them. AI talent and tools are becoming more of a commodity for which they can partner with Big Tech platforms and startups.

According to McKinsey & Company, applying AI just to marketing, sales and supply chains could create economic value of $2.7 trillion over the next 20 years. Beyond making operations more efficient and effective, these companies can use AI to create new products based on information locked in their systems. According to a recent Accenture study, about 80% of large companies have adopted machine learning and other forms of artificial intelligence in their core business. Five years ago, that figure was less than 10%.

Incubating AI ventures within established companies
is not a bad way to start

Much of the same guidelines for startups discussed in the prior post applies to AI projects within established companies. In fact, many companies tend to think of AI projects like internal startup ventures. This can be a useful approach in the beginning but it leaves a lot of potential value on the table as we’ll talk about shortly.

The majority of big companies using AI, develop these tools as isolated applications, not integrated into the wider business or IT architecture. In essence, the companies create standalone AI startups within their company. While this may be OK to test the waters, most companies should re-architect around AI solutions if they want to fully reap the benefits. Leading companies see AI as a vital part of its IT infrastructure and potentially the future of its business, and even its product offering.

Like with startups, AI projects within large, established companies should have a clear, measurable goal that adds real value, leveraging the domain expertise of the company.

Established companies often have the data they need to derive economic value and competitive advantage – If not they need to mine it directly from their operations

On data strategy, large companies have a substantial advantage in that they likely own the data, or at least have access to it from their customers, suppliers and partners. Nonetheless, the stakes of mishandling this data is dramatically higher. For this reason, it makes sense for companies that are serious about AI to have a data strategy function or more colorfully titled, *Data Tsar.* Overcoming organizational resistance to sharing data, and setting priorities across AI and other data-hungry projects will require strategic oversight.

Established companies should leverage the platforms of big tech and talent and energy of startups to jump start their AI strategies

Like with startups, there is no reason not to leverage cloud-based tools for AI development. At the very least, companies can use these tools for rapid prototyping and experimentation with their data. Plus it allows them to move quickly without building a large technical machine learning team or IT infrastructure.

Walking before you run is important to AI projects in big companies, but for different reasons. While cashflow may not be important to most of these companies, overcoming organizational scepticism and competing priorities always favors projects that can yield measurable and meaningful results in a short time horizon.

Attracting AI talent is a different kind of problem for large companies, since they usually can’t offer the promise of big equity payouts that startups can. One way to overcome this challenge is to pool machine learning and data science talent together into a central organization. This would allow them to sharpen their cutting-edge expertise better than if they were scattered across teams. Also, centralization tends to yield greater job satisfaction and retention for these types of roles.

Long term, establish companies will need to build centers of excellence around machine learning and data science

On the other hand, this is a good area for big companies to consider partnering with startups to co-develop AI solutions. Alternatively, large companies can contract for this expertise from vendors and consultants, but at substantial expense.

Keep in mind, much of the work developing AI may not require formally trained engineers or data scientists, but may be carried out by sharp analysts who have become conversant with the technology and who are proficient at using the automated machine learning tools.

Ethical concerns and brand liability related to AI (and data) need to be addressed head on by establishing formal oversight early

Finally, large companies should establish ethics-related positions or review boards as a part of their AI initiatives. For example, Microsoft, has created an “AI Ethicist” role to provide guidance on topics like algorithmic bias.

Note: this article is based on a lecture given by me at UC Berkeley, Haas School of Business, as part of the “Business of AI” online course released December 2019