Marc Esconde, Senior Advisor October 2019
Artificial Intelligence (AI) has demonstrated a remarkable ability to improve business performance in many different ways. However, the sheer volume and diversity of AI applications, combined with the enormous amount of hype currently affecting the AI field, have made it increasingly difficult to fully realize AI’s business-value potential. This article presents a simple yet effective 7-step process through which your company can broaden and accelerate AI’s business value while avoiding wasting time and money on AI “dead ends.”
Proliferation of AI Business Applications
In a few years, AI has taken the business world by storm. Fueled by massive investments from big tech firms and thousands of start-ups, this general-purpose technology has given birth to countless AI-enabled business applications, which analysts now expect will deliver trillions of dollars in value over the next decade. They are likely to be proven right.
Indeed, many freshly developed AI business applications have already built an impressive track record. A recent study sponsored by Analytics2Go (an AI-as-a-Service company) identified 86 use cases where AI-based applications have unequivocally improved the performance of 15 key business activities: automate, classify, create, detect, enhance, facilitate, fake, improvise, optimize, personalize, predict, prevent, segregate, simulate, and translate.
These highly diversified use cases—a summary of which is presented at the end of this paper—provide a sparkling illustration of AI’s ability to enhance virtually every business function in every industry.
Separating the Wheat from the Chaff (Hype)
Seizing all these proven, AI-powered performance improvement opportunities in every corner of a business is a major challenge in itself. To make matters worse, one must also cope today with an enormous amount of hype (aka “fake AI news”), which has made it almost impossible to separate the AI wheat from the AI chaff, and has caused some chaos in the AI application market.
To get a vivid illustration of today’s AI chaos, just type “AI” along with the name of any business function or process (e.g. ‘AI accounts payable’, ‘AI recruitment’, ‘AI supplier risk management’, ‘AI event planning’, etc.) in your web browser. If you feel overwhelmed by the number of choices, promises, and lingo, not to worry, everyone does. AI is a new Wild West: a dangerous place filled with outlaws and dubious ventures. More precisely:
- Many AI solutions actually contain no AI at all; they are just existing Business Intelligence (BI) solutions with a fresh haircut.
- Many other AI solutions are still at the conceptual stage or going through initial experimentations, without any definitive proofs that they can deliver true business
In this hyped AI environment, it is no surprise that many AI business projects end up delivering limited to no value (the so-called AI dead ends), and at the same time, the truly valuable AI applications and services may never show up on the radar screen simply because of this blurred picture.
Accelerate AI Benefits
The seven-step process described below is a simple yet effective way to focus your company’s efforts on the most valuable AI applications without being distracted by the hype, eventually leading to increased AI business benefits, achieved faster and at a lower cost.
Step 1. See the summary of the 86 proven AI use cases here.
Step 2. Remove the use cases that are either not applicable to your business or are already part of your AI experimentations.
Step 3. Create a cross-functional task force composed of professionals equipped with solid business knowledge, a creative mindset, and a basic understanding of AI fundamentals. (If unsure about the basic understanding of AI fundamentals, we suggest that your task force members sit through the excellent non-technical, easily digestible “AI for Everyone” online course from deep learning.ai, whose lectures are available free of charge on YouTube).
Step 4. Ask members to share their views on each use case’s applicability and value potential, ranking them H (high), M (medium), L (low), or N (none). Encourage out-of-the-box thinking (e.g., identify a use case from the list that is not an exact fit but could provide an innovative AI opportunity that will leverage your company’s own unique data), and ask the respondents for further suggestions on the items rated “High.”
Step 5. Consolidate all inputs, then invite the task force members and representatives from your technical AI team to jointly analyze the ratings and suggestions. Eventually, come up with a shortlist of no more than 10 projects with a perceived high business value and a likely technical viability.
Step 6. Perform a detailed technical and economic viability analysis on each shortlisted potential project. Develop a time-phased plan and a budget for those deemed viable.
Step 7. Execute the projects, measure the business benefits, collect the lessons learned, and apply them to subsequent AI projects.
Other Ingredients for Success
Needless to say, building a consistently competitive business using AI solutions requires many other critical ingredients, such as:
- Company-wide education on what AI can and cannot do (an absolute must, starting from the top!)
- Sanitized internal data feeds
- A sound data-acquisition strategy
- Permanent access to adequately skilled AI engineers and data scientists to build, implement, and maintain your AI applications
- Orchestration of AI activities through a robust, company-wide strategy and governance framework
- Partnering with the most capable AI solution and service providers
- Using clear go/no-go criteria to launch larger scale deployments
- Rigorously tracking KPIs to measure AI benefits
- Continuously improving AI projects through cross-fertilization and sharing of lessons learned
- Continuously improving AI models
- Maintaining an active AI technology watch
Completing the recommended 7-step process should help your company quickly sharpen its AI focus on applications and services that will increase business benefits, reduce risk of failed projects, and accelerate time to value.