The habits organizations adopt in how they approach artificial intelligence have the power to speed them to its use – or to slow them down.
The most successful companies and governments (who say they are using AI now, not just piloting or experimenting) are substantially more likely to also report they follow four key habits in addressing it.
In a recent survey of 600 organisations about their AI plans, organizations who are now employing AI demonstrated, in many cases, one or more of these habits, which set them apart and propel them forward.
No single habit is responsible for succeeding or stalling – but engaging in multiple good habits can elevate a company and make it more likely to weather the inevitable challenges that AI projects can fall prey to.
Adopt as many of these habits as you can to go from succeeding in your AI experiments to making AI something you use commonly in your organization.
Centralise oversight but bring together people from across the organisation
Successful organisations are more likely to start with a task force to oversee all AI initiatives. Such teams are best when they draw participants from most departments within the business, and when the teams have people who think in various ways. The most successful employ the most roles on the teams, including AI researchers, AI engineers, data scientists, machine learning specialists, data engineers, application designers, software developers, strategists and project managers.
Formalise AI-performance accountability, decision making and budgeting processes
Highly effective organisations in AI are more likely to assign specific and influential stakeholders to AI initiatives, often including C-level executives within the company. Our recent survey revealed that more than three out of four organisations who are employing AI indicated that they’ve assigned C-suite level stakeholders the responsibility of the AI initiatives’ performance, giving these stakeholders the responsibility to meet performance indicators and be accountable for how the AI projects perform. In contrast, businesses not employing AI are more likely to say they have made business units accountable for AI success.
Part of this habit is assigning budgetary responsibility to a C-level or equivalent executive. More than 40% of organisations utilising AI today have assigned the budget to a corporate function, while only 34% of organisations not employing AI follow this practice.
Limit your proofs of concept to focus on the most promising opportunities
Although it’s tempting for companies to explore as many AI projects as they can, the more successful organisations focus their resources and planning on the most promising opportunities. Narrower focus makes organizations more effective in the long term.
Indeed, organisations who invest in fewer efforts to develop pilots and proof of concepts (POC) are today more likely to be utilising AI effectively within their business. In contrast, businesses that are not currently employing AI have consistently conducted more POCs and the results show that organisations employing AI today have done fewer POCs and pilots in every category of AI application about which the survey asked.
Commit to financial and risk analyses of all efforts both before and after implementation
Performing financial or risk analyses and practicing careful project selection are key habits that go with success in employing AI. About 50% of organisations that are running financial or risk analyses of their AI efforts report that they are employing it today. We believe that such discipline and rigor go together with successful efforts.
Organisations that are best able to defend and promote those activities, in general, are taking the step of calculating value by:
1. Establishing a baseline to compare performance
2. Determining key measurements to analyse
3. Tracking them in detail through the project’s lifetime
These businesses can show that they are better able to achieve the goals that executives and sponsors have identified — or they can show why they did not achieve these goals and how they can improve performance through alternative approaches.
Focus on the long-term commitment that will lead to long-term payoffs
AI will persist as a major new aspect of software and services for organisations. It will have profound impacts in IT departments and significant effects throughout the business. These habits demand and benefit from long-term commitment, and they will foster long-term payoffs as well. While top executives in organisations are often impatient for AI boons, they will also benefit from adopting these general approaches to improve overall project and product performance in the future.
Whit Andrews is Vice President and Distinguished Analyst in Gartner Research