Interactive: The truth about AI from companies who are using it

While there is a lot of discussion around cognitive, it is mostly early adopters that are actually deploying it. This can come down to a number of reasons, but as the barriers to entry become lower, many more organizations are looking at the opportunities cognitive technologies can provide and learning from the mistakes of the early adopters.

Starting a project the right way can make or break it. A systematic approach to engaging with and implementing cognitive technologies may seem to require more effort than the “just do something” approach in the short term, but it is more likely to enable success and require fewer resources in the long term.

Whether it’s deep learning or robotic process automation (RPA), projects that don’t take a systematic approach can fail for a number of reasons. Often teams struggle to define a good use case, or perhaps don’t use the right technologies for the problem. Like any transformative project, there tends to be high risks for high rewards, so it’s not surprising that they often fail at the pilot level. Additionally, some of the projects impact the organization’s existing technology architecture, but as IT groups may not be involved in the initial stages, it can be difficult for the technologies to be integrated into an existing architecture.

Knowing the common pitfalls is a good way to enable projects not only to get off the ground, but also to have the desired impact. The steps below are some of the most effective things you can do to navigate the cognitive technology landscape:

• Educate senior management on the technologies and their likely impact. While they may hear about cognitive technologies from the media and peers, confirming they have a firm grasp on the different types of technologies and the impacts they can make on their businesses is vital.
• Select the right technology for your business problem. From RPA, to natural language processing (NLP), the alphabet soup of solutions can be hard to digest. It’s important that organizations understand the proper uses of each technology and the specific way to employ it to get the desired results.
• Recognize that “low hanging fruit” projects tend to have a much greater chance of succeeding. Even though they may have less potential value, starting with a simpler project is often much more likely to achieve a meaningful result.
• Build in expectations for learning and adaptation. By definition, many cognitive systems need to be trained and improved over time. Rarely does the initial “go live” mean that the pilot works at an optimal level. A good measure of effectiveness can be based on the ability of both the team and the system to adapt and improve over time.
• Get a portfolio of projects going. With such a range of technologies available, it’s important to gain experience quickly. All projects should be developed with agile, “minimum viable product” approaches.
• Discontinue some projects, scale up others. With the portfolio approach, you can be better positioned to end the projects that fail and devote resource to scaling up those that show the greatest potential.

Discover $DFS more about cognitive technologies and their potential impact on businesses here.

This article was created for Deloitte by Quartz Creative and not the Quartz editorial staff.