There is no longer any question that artificial intelligence can make a monumental difference within an organization – from solving big problems to completely evolving a business model. That’s why Microsoft created AI Business School two years ago to provide guidance on strategy, culture, responsibility and other critical topics for key decisionmakers across industries.

Implementing AI holistically and successfully is no easy feat – and we certainly don’t have all the answers. But as we have worked with business leaders through AI Business School and in customer relationships, we’ve seen a few companies who set the bar for what an AI-powered organization can be.

And that’s why we created Best of Business AI 2021: to showcase customers who have shown us what it looks like to have a comprehensive strategy and execute across the organization to create business value and momentum with digital transformation. Best of Business AI 2021 builds on the frameworks from AI Business School and features 10 customers who are accelerating their AI journey by connecting a strong leadership approach to their business goals and technical capabilities.

What follows are a few selected tips from our featured customers. Visit Best of Business AI 2021 to fully immerse yourself in their success through video, stories and interactive elements that will give you insight into how they did what they did and how you can follow in their footsteps. We’re proud to share their journeys with you.

Involve everyone in your organization

Companies that successfully implement AI understand that it is not just a topic for technical teams or leadership. They have meaningful conversations with employees to understand how each part of the business can benefit and how it may be impacted. Outokumpu involved employees across the organization from the beginning, and it gives them access to data and AI models so they can make the best decisions in their day-to-day work.

Stefan Erdmann, Chief Technology Officer at Outokumpu, says, “You have to get trust from the leadership team, talk to people about rollout and get everybody on board.”

Use AI to increase business value

AI can help you solve difficult problems, but it may also bring new opportunities for growth and success. At HOCHTIEF, the company is monetizing their new AI expertise by offering AI applications to customers as a service.

“We will offer all these solutions to market and we will have a new business model for the company,” says David Koch, Chief Risk, Organization and Innovation Officer for HOCHTIEF.

Choose a strategic starting point

Start by finding the right use case to apply AI, then be sure you are sharing your results regularly. At CES 2019, Bell used an augmented reality experience to demonstrate its vision for AI and how it could transform the business and the industry. This showcase inspired excitement within the company and the tech community, and the company has built on that momentum with regular progress toward its big goals.

Matt Holvey, Senior Manager of Intelligent Systems at Bell, says, “Highly iterative, incremental proofs of concept—demonstrating something every three to six months—are the best way to gain and continue the championship and endorsements from upper leadership.”

Bring business and technology together

Everyone in your organization is there because of their expertise. Empower them to collaborate, innovate and help come up with creative AI solutions. At Mondelēz International, AI experts and product experts teamed up for the most impact.

Rob Hargrove, Executive VP of Research, Development and Quality at Mondelēz International, says, “On one side we have our data science, modeling, and simulation experts. One the other we have our biscuit, chocolate, gum, or candy development experts in our product teams. Neither side knows the other’s job perfectly. So they know they can’t work in siloes because then neither side will be successful.”

Put values into action

As with all significant technological innovations, the use of AI can also have unintended consequences, including privacy issues and security breaches. Implement principles, practices, tools, and governance to help your organization anticipate and mitigate any risks – and to be sure the end result aligns with your values. At AXA, creating governance tools and practices was not only the right thing to do, it also helped the company learn more about their business and how they can best help their customers.

“Our responsibility is to properly manage the data that the customer opts to share with us,” says Jerome Lafon, Head of Connected Car Business Domain, Data and Tech Innovation at AXA.

Work with other entities

Implementing AI responsibly often requires collaboration with other organizations, interaction with end users and coordination with public entities. Look for guidance from others– and use what you learn to help where you can. Nonprofit OceanMind works with partners to understand the impact of human activity on oceans.

Kanit Naksung, Director of Fish Quarantine and Fishing Vessels Inspection Division at the Thailand Department of Fisheries, which is partnering with OceanMind, says, “We can use AI to help enforce fisheries law and to help authorities make better decisions.”

Prove it, then scale it

Once you’ve seen that something works, invest in expanding it across your organization. Dedicate resources toward evaluating AI solutions and developing strategies to expand their reach. At DHL, this meant building the right infrastructure, then providing support and communication to go from proof of concept to productization.

Markus Voss, CIO & COO for DHL Supply Chain, says, “In almost every one of our facilities, we are deploying those mature digital solutions. It is my vision to have every customer, every site, and every one of our employees sensing that the world is changing in supply chain.”

Prioritize for the most impact

You can’t solve every problem at once, so prioritize your efforts, exploring the costs and benefits then determining exactly what success will look like. You may find, like CSIRO did, that AI is wanted everywhere, but it’s important to choose carefully so your work can be of real value.

Jon Whittle, Director of CSIRO’s Data61, says, “Everybody wants to work with our AI experts, both within the organization and also outside CSIRO…The danger of that is that you get pulled in too many different directions, and you are no longer able to really make a difference in any area, because you tried to do too many things.”

Empower everyone

Good ideas can come from anywhere, which is why the “AI everywhere” approach at Grab has been so successful. Non-technical employees work with specialists to turn ideas into solutions.

“When people embrace and have conviction of what you’re trying to do with it, you really unlock the value of AI,” says Wui Ngiap Foo, Head of Technology at Grab.

Put your data to work

Data can be used in unexpected ways, helping to enhance and scale human expertise. At WPP, data sparked imagination and new ideas for the company’s designers and marketers, empowering them to be even more creative.

Di Mayze, Global Head of Data & AI at WPP, says, “Data isn’t boring; it can really inspire and surprise. And that’s what we want, to make sure that data and creativity work together and party together!”