Five ways for technology leaders to get the most out of AI
Lessons learned from Majid Al Futtaim CTO Richard Wingfield
During Microsoft Envision, we hosted a breakout session with Richard Wingfield, Chief Technology Officer (CTO) at Majid Al Futtaim – Ventures, part of Majid Al Futtaim, a leading shopping mall, communities, retail, and leisure pioneer across the Middle East, Africa, and Asia. The company’s portfolio includes 23 shopping malls, 13 hotels, four mixed-use communities, 355 film screens, and iconic leisure and entertainment facilities such as Ski Dubai, Orbi Dubai, and Ski Egypt. It owns franchises representing international brands such as Abercrombie & Fitch, AllSaints, lululemon athletica, Crate & Barrel, and Maisons du Monde.
As CTO, Wingfield and his team are spearheading a massive digital transformation that involves the analysis and deployment of advanced analytics technologies across the organization.
Onstage, Wingfield shared his digital strategy—digital for customers, for efficiency, for culture—and his approach to artificial intelligence (AI) through several hard-won lessons and use cases. Here are Wingfield’s five key takeaways for business and technology leaders looking to get the most value out of AI.
Begin with clean data
Going digital can do wonders for efficiency. Wingfield manages digital strategy in sectors as diverse as finance, fashion, and even cineplexes. Of course, a complex organization creates a complex web of data. Wingfield knew that his company’s data held untapped potential, but it took work to unlock it. His team began by pulling the “easy” data, like transaction data, into Microsoft Azure. Then, the team continued to stock its digital data warehouse with new streams, like customer behavior and product data. The focused, intentional build-out gave them the ability to test and learn to ensure the models delivered valuable insights, reliable results, and return on investment.
Avoid data swamps
As you think about building your own data warehouse, it’s imperative to understand what data you are collecting, why you are collecting it, and how you plan to use it. A data lake, or raw depository, is a fine approach. But avoid the temptation to create a data swamp—a massive, unorganized collection with very little meaning or value. To unlock the power of your data, start with your strategy. Be clear about the data you need to deliver against that strategy. Start with easy-to-access streams, and then build in layers until you have a broad view and can pull KPIs across all areas of your business.
Don’t expect a silver bullet
Smart as it is, AI is not an instant fix. Unlike software, the work doesn’t end once the code is written and the software is deployed. In fact, that’s where the iterative AI journey begins. Majid Al Futtaim constantly tests, tweaks, optimizes, and evaluates the models—and then goes back and does it all over again. Wingfield notes that AI is evolving quickly, and exciting new products are coming to market with data science and machine learning built in—thus lowering the threshold to leverage AI capabilities and accelerating speed to market.
Have patience. Fail Fast.
In Wingfield’s experience, it’s important to prepare for failure. He recalls frequently awaiting a model’s deployment with excitement and anticipation, only to scrap it when it did not prove valuable to the company. Success requires iteration, flexibility, and the acceptance that failure is part of any AI journey. Looking back, he is excited about what the organization has learned and values its ability to adapt.
Any major business transition needs to resonate from the inside out, and that is especially true of AI adoption. It’s critical for employees to feel that AI puts the right tools in their hands and makes their jobs easier. Wingfield applauds the steps taken by Majid Al Futtaim, which has built a companywide school of analytics to empower and educate its employees in all facets of the business. Ensuring business leaders and counterparts are engaged from the beginning on use-case definition, KPIs, and results is a must. He firmly believes this isn’t a technology-owned effort; rather, it requires both business and technology collaborating. In the end, it all comes down to trust—after all, data is inert, and models don’t run themselves.
See more on Majid Al Futtaim digital strategy and brand experience.
Want to kickstart your own AI project inside your organization? Explore this interactive guide as you build your plan.
