Artificial Intelligence (AI) - Challenges and Opportunities

John Hagerman

President/The BrandGate - AdvisoryCloudTM

Overview

Artificial Intelligence (AI) has been implemented and is delivering on its promise at least at large companies including Facebook, Google, and Netflix.  Retailers are using AI-powered robots in their warehouses, Utilities use AI to forecast electricity demand, Automakers are using AI for autonomous cars, and Financial Services companies are using AI to better understand their customers, look for potential fraud, and to identify new products/services customers will want.  But as we look beyond the technology segment and these specific examples, AI adoption is still at a very early level. 

The typical company who has successfully adopted AI is larger than their industry peers, have already embraced the digital transition and well on the path to implementation, employ AI in the core of their value chain, adopt AI to increase revenue, and have the full support of executive leadership.  

The good news (according to the McKinsey Global Institute) is companies who are leading AI adopters have profit margins from 3 to 15 percentage point higher than industry averages in their market segments. So why isn’t everyone jumping into AI?

Challenges

A number of companies have implemented AI programs and understand that a full implementation touches all aspects of the business.  The insights gained have led to transformational changes that have fundamentally changed their business.

From implementing Predictive and Prescriptive Analytics solutions we have found the challenges fall into 4 major areas:

  • Cost – supporting not only the initial outlays for software and costs for cloud support but the on-going costs for training employees and continued training of the AI system when business processes change.
  • Culture clash – any changes will be viewed with suspicion by the organization. For example, I have had employees whose primary job responsibility is to generate a weekly report which with the implementation will be done instantly by the system. Also, because no one has really analyzed the data you may find that markets and certain promotions really don’t work as commonly believed. 
  • There are so many technology options – so it is hard to narrow down the potential choices and select the best implementation path.
  • Inability to precisely predict ROI especially at the start of the project will be a source of friction.

While these challenges must be addressed, it is important to keep the team focused on the long-term goal and the benefits AI will deliver.  As with all transformative changes, clear and open communications will keep the team engaged and maintain management support.  In addition, companies should understand that a re-training program will be valuable for those personnel affected by the business process changes.

Opportunities

As stated earlier, the profit margins for companies who adopt AI is much higher than their industry peers.  For example, McKinsey predicts that a retail company could generate profit margins 9 percentage points higher than industry average.  In Health care and Financial services, the margins will be even higher. 

Also, the additional benefits of higher customer engagement/satisfaction, smarter more efficient R&D, optimized production and maintenance, and more efficient and effective sales/marketing campaigns will begin to create a sustainable competitive advantage.  And competitors who delay adoption may find themselves in a very difficult position.

Where to Start

Because of the attention and far reaching impact of AI implementation programs it is advisable to start with: a small project, that can be implemented quickly, and provides value to the company.

The criteria for an initial project are:

  • Start with a few small, quick to implement projects. Leverage this to build support within the organization on the AI initiative by demonstrating early successes.
  • Engage key stakeholders early and often. Keep everyone informed on objectives, timelines, progress to goals.  Rumors tend to fill a void when the project does not publish regular status updates.
  • Bring on an Implementation Partner. Few IT teams have the time, technical knowledge, experience, or general capabilities to design and implement an AI project.  By bringing in a partner for the initial implementation you will increase the probability of success.  With a positive proof of concept, approvals for additional AI projects should be easier.

Summary

AI offers the ability to transform your organization, improve operational efficiencies, generate higher profit margins, and provide the insights to create positive customer relationships. The journey to implementation can be complicated and frustrating but, in the end it can be highly profitable.  Those that successfully complete the journey can create a sustainable competitive advantage founded on a highly profitable business model with a very loyal customer base.