This article was co-authored with Andy Thurai , vice president and principal analyst with Constellation Research, and is a follow-up to our previous article on AI’s secret sauce.)
AI, show us results
getty
The transition from artificial intelligence merely being a collection of disparate projects to adopting the secret sauce of business use cases requires some organizational moxie and strategic planning, but does not have to be an arduous or daunting process.
While AI can make organizations very successful, success doesn’t happen randomly. There needs to be effort put into identifying the right use case, getting the right support, getting the needed funding, and finally successfully deploying it quickly.
The following are ways to elevate your organization’s AI efforts into successful business use cases:
- Don’t pitch “artificial intelligence.” Pitch business growth. Make the case for business advancement using AI that can deliver much better results than the current way of doing things. While the term “artificial intelligence” will certainly get everyone’s attention, remember that you’re not selling a technology implementation – you are selling an improvement in process, cost efficiency, a new revenue stream, or a way to gain new insights for decisions. If you can’t justify the project in any of those manner, then the project doesn’t have a chance to be successful right from the beginning.
- Learn the pain points from the field and from executives. Identification of a potential use case can be very tricky. Especially in the organizations where AI is not widely adopted, the business and field users may not be aware of what AI can potentially do for them. Instead of explaining AI to them, an easier alternative could be to ask them about their struggles and what can make their life easier. If there are commonalities across units, geos, or even partners that will be an easier business case to explore. If there is potential, it can be easily validated back with the identifiers with the suggestion of what AI can do for them in this specific case. Not only will this make it easy on them, but they will be very invested as this can potentially solve their major pain point. They can obviously help sell this to their executives who can even consider funding this exercise as their problem is getting solved.
- Strive to democratize AI. Where do promising AI development activities go awry? For starters, if it’s tied too closely to the technology itself, and not the business – to the point where it’s black art to business users. “AI needs to be in the hands of everyone, not just the experts,” according to Mona Chadha , director of category management at Amazon Web Services. “AI tools need to become easier for line-of-business users to apply and to get value. There is a shortage of AI experts and data scientists capable of leveraging sophisticated AI frameworks and infrastructure.”
- Identify advocates. Identify proponents within the organization who can sell AI to executives and managers seeking better ways to address their business problems or opportunities. These individuals need to understand the scope of their company’s AI needs more than the developers or data teams building or incorporating the AI solutions. They need to speak the language of business, and help business leaders understand how AI will address their worst pain points.
- Build trust with potential users. Executives and managers may be enamored with the technology itself, especially with ones that are complicated such as AI, and may be hesitant to bet their businesses on it. That could be because of a perceived trust gap in terms of the insights or recommendations AI may deliver versus what they see in the field. This is where a solid demonstration of successful use cases either from within the organization or from outside where similar implementations were very successful.
- Follow emerging examples of success. There are now working examples of successful AI initiatives, with proven value to businesses. Examples include the use of AI to diagnose diseases and provide personalized treatment, untangle traffic jams, optimize supply chain flows, provide proactive inventory tracking, help secure sensitive data, personalized customer engagement through conversational AI, training, coaching, and measuring marketing returns. Look for areas where competitors have used AI to solve business issues successfully, or look for successful use cases from adjacent industries and use that as a starting point.
- Establish success metrics. You can’t manage or improve what you can’t measure, let alone build a robust and expansive business use case. Look at variations between cost savings, efficiency improvements, revenue achievements, or any other success metrics that were established from before AI solutions were put in place to those following deployment to prove how this AI project is widely successful.
A good AI strategy is only a starting point. Without proper execution, it is just a hallucination.