4 key questions to ask before starting an artificial intelligence project

4 key questions to ask before starting an artificial intelligence project
Photo by Shutterstock

Nowadays, more and more companies are starting to start artificial intelligence projects, but some projects have not been successful. Before starting to implement the first artificial intelligence project, companies need to understand some key issues.

According to a survey conducted by the research organization Gartner, companies’ implementation of artificial intelligence projects is expected to double this year. By the end of 2020, 40% of companies will deploy artificial intelligence projects. Such statistics will put pressure on the CIOs of other companies, because their corporate executives want to know why they are not innovating in this field.

There is a cruel fact hidden behind all the propaganda and hype surrounding artificial intelligence. A study conducted by the Massachusetts Institute of Technology Sloan School of Business and the Boston Consulting Group found that 65% of companies believe their artificial intelligence projects are worthless. When many projects fail to realize value, how can they bring success to the company’s business? Here are four key questions that companies need to ask before starting their first artificial intelligence project.

1.Where can artificial intelligence provide quick wins?

People are hearing every day how artificial intelligence will change business. Although it can be changed, this should not be the goal of the company’s first artificial intelligence project. Instead, implement a small-scale project that can bring rapid success, and success nurtures self-confidence, allowing the company to embark on a path of sustained success.

In the first artificial intelligence project, the company hopes to acquire knowledge and demonstrate the impact of artificial intelligence on its business, and can choose projects with visibility at the top of the enterprise. Find content that closely matches existing business processes so that the impact can be felt. After successfully delivering the project, you need to find a way to motivate each successful person. If artificial intelligence is expected to be infectious throughout the organization, department heads should be concerned about how artificial intelligence technology can bring about meaningful changes.

2. What is the data like?

The success of artificial intelligence and machine learning depends on large amounts of data. Companies need to analyze data storage to see which restrictions may hinder project implementation. Is there very little data collected? Does it need to be cleaned up more? If it takes years to fully compile enough data, the project is not feasible. If the collected data is very messy, you must determine what effort the data scientist needs to clean up.

In any case, there is no perfect data, but you can’t hold back because of it. Don’t choose a project with less impact just because another data set is more complete. The discovery phase is the perfect time to enter and explore possession. Companies need to spend some time modeling data to determine whether they can tell stories with fewer resources.

3. Are artificial intelligence projects creating value?

When deciding to implement a project, adding value should always be the focus of the company. This could be cutting costs, increasing revenue sources or simplifying business processes. So where are the inefficient processes? Where can better decisions be made? The value proposition should always be supported by data, not intuitive. Companies need to show top management why they want to implement this plan and what they expect from it.

When people see potential artificial intelligence projects, they want to determine the task, not make a large-scale modification. The ideal approach is to choose a process that is repetitive, rules clearly defined, prone to human error, and has data to support them. Companies need to build logic around these processes in order to reduce the gray area.

4. Know what is the definition of success?

The difficulty of delivering successful projects is not unique to artificial intelligence. This problem plagues many project teams for many reasons. It can usually be attributed to some unrealistic timetables, over-budget, scope expansion, and a combination of not having the right expertise to execute correctly. And project planning is the key.

Companies need to eliminate the is landing effect. Artificial intelligence engineers and data scientists need to work hand-in-hand with business analysts and end users to understand the problem and discover what the successful outcome is like. The team leader not only needs to be integrated into the interdisciplinary team, but also can talk about artificial intelligence solutions in simple language, so that key stakeholders will have a clear understanding of what kind of impact artificial intelligence will have and what kind of impact it will not have.

In addition, companies should not think that they can succeed in their own way. They also need to cooperate with trusted partners to obtain the necessary artificial intelligence expertise and solve the technical obstacles encountered in the initial project.

According to McKinsey & Company’s survey, by 2030, artificial intelligence will bring about 13 trillion US dollars in global GDP growth. According to a study by PricewaterhouseCoopers, 72% of executives believe that artificial intelligence will be a future business advantage. This is not a question of whether companies want to implement artificial intelligence, but a question of when. By thinking about these key issues, it can become one of those rare success stories. This success will help companies build a culture that enables artificial intelligence to flourish and improve business.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store