WFA Musings - Autumn 2023
AI’s Impact on Strategy
This second in our AI series tackles the issue of AI’s impact on business strategy. Given the significance of this topic, we have posted a more detailed paper on this subject on our website. The first in our series argued against any potential “pause” in the development of generative AI. I want to credit my partners, Peter Gates, Bob Goad, Jill Kravetz and John Trustman, for their significant contributions to these white papers.
The most important takeaway we would offer you is this: it is imperative that every business revisit and adapt its strategy in this AI era. The question is not whether but how and when AI will impact your business. It is likely AI is already impacting your business. We at Acropolis Advisors know from decades of experience that there is never a one-size-fits-all answer, and we welcome the opportunity to discuss your specific business needs.
While AI has been the darling in the news over the past year, it has been around for decades. There are already myriad applications where AI is present, from customer service interactions to robotics to managing various enterprise workflows and many more. Where humans are responsible for creating the algorithms for various functions, we are in the world of “Software 1.0.” When AI is responsible for creating new algorithms with no human involvement, we have entered the world of “Software 2.0.” This second world is what has caused the recent stir in the news because it can create unknown algorithms which may have good consequences or bad. The ability to create new algorithms without human intervention is only possible when there are multiple layers of algorithms developed by humans stacked upon one another with a massive amount of data to draw upon such that new algorithms can be based on learnings over millions of iterations. Good consequences arise when the algorithms in each stack leading up to the generative ones are sound and neither erroneous nor sinister. The other prerequisite is that the data that is being mined is “accurate.” This criterion is particularly hard to meet as it requires forensic inspections at times. Worse, most of our new AIs are trained off the data on the internet and, while the internet is a treasure trove of good data, it is also a storehouse of bad data and there are no easy ways to tell the difference. As you can imagine, meeting these two conditions is not a trivial exercise.
In addition to the software driving this revolution, it is also enabled by having the hardware to power it. The special purpose, limited instruction set computers or co-processors which do the specialized matrix algebra previously critical mostly to rapid rendering of graphics, have proven ideal for the huge number of simple multiplications and additions necessary for training and executing the AI math.
At its best, AI is able to distill reams of data and convert it into positive action without human intervention. Initially, this will occur for basic tasks followed by more advanced tasks. However, because there is rarely a “perfect” answer for mission critical decisions, human intervention will always be necessary. AI makes decisions based on stochastic distributions of prior results with a little randomness thrown in for creativity. As such, AI allows for better-informed decision-making based on probabilistic outcomes but will never be able to replace judgment amongst competing interests. AI will also never replace groundbreaking creativity. The DaVinci’s and Einstein’s of the world will continue to be the trailblazers.
When companies construct a strategy, they scour the landscape for intelligence on customer behavior, needs, and pricing sensitivities; competitor capabilities, costs, and value propositions; supplier capabilities and costs; and the societal requirements for the business. They then compare that intelligence to their own capabilities and costs. This baseline of information forms the basis of developing a sound strategy that can achieve sustainable competitive advantage. If the company is a follower in its industry, the strategy will be geared to gain share through cost leadership or differentiation. If the company is a leader, it will be geared to bolster its leadership position by combining both.
Since the AI era is now in full gear and ubiquitous, it’s important to realize that it will impact all of the above inputs to strategy development to some degree, necessitating a “real time” monitoring capability that addresses all aspects of the business. But companies comprised of people cannot be constantly turned on their heads. That would result in corporate ruin. However, when the cumulative evidence mounts that the conditions upon which the strategy was developed have been sufficiently altered, then the strategy needs to be revisited. Successful strategy development will create a “base case” along with reasonable alternative scenarios which can be pursued when the underlying assumptions change.
We submit that while the timing and impact of AI on strategy will be quite different by industry, there are two general takeaways. First, the fundamental principles of architecting competitive advantage as espoused by our co-founding Senior Partner Emeritus, HBS Professor Michael Porter, remain firmly intact. Second, the competitive landscape is subject to rapid change unlike ever before and, accordingly, developing a real time monitoring system for the key drivers applicable to all stakeholders is critical. Strategy has not changed, but cycle times have compressed.
When Professor Michael Porter authored his seminal works on strategy in the early 1980s, he mapped out in his Five Forces and Value Chain constructs as the key components to understanding how a company could attain leadership within its properly defined business arena. In the early 1990s, I developed a strategic construct called the Full Potential Paradigm™ which fully embraced the tenets of Mike’s work and brought quantitative approaches and tools on top of it. This approach has helped the business leaders I have advised to define the level of value creation that can be attained over a specified time period, typically 5-7 years. Together, in 2020 Mike and I developed a new construct called the 21st Century Stakeholder Optimization Framework which posited that the primary needs of all stakeholders can only be met while also generating sustainable increases in enterprise value. This built off of Mike’s work on Shared Value and has extended our previous constructs.
At the heart of our thinking is that attaining leadership within an industry requires the CEO to adopt a mindset of “maniacal realism” and “tenacity.” “Maniacal realism” is defined as “leaving no stone unturned” and “abolishing delusions” which in this Information and Digital Age means constantly assessing the market conditions for every stakeholder to ensure that their primary needs are being addressed and never fooling yourself. AI can allow CEOs to receive a real time feed from a combination of proprietary in-house data and third-party data to understand the pulse of the business. “Tenacity” is defined as turning the implications from this information into action plans AND the willingness to change course when new information upends critical underlying assumptions.
To repeat, every business must adapt its strategy in this AI era. The question is not whether but how and when AI will impact your business? Ignore it at your peril.
Take the current situation that exists between the “Big Three” automakers in the US and the UAW. If you are one of the OEM CEOs, you must be able to process myriad factors as well as how those factors will unfold over time. It is a probabilistic linear programming problem that cannot be solved without an objective assessment of the probabilities of the inputs and intermediate steps. If the OEMs were to accept the UAW proposals how would it impact profits, market share, employee satisfaction, supplier economics, product pricing, customer receptivity, community relations, relative cost position, and on and on? The reverse is true for the UAW. If they accept the OEMs proposals, how will that impact member compensation, recruitment, retention, dues, benefits, pensions, etc.? And these analyses must be done in both the short and long run, with respect to their other unionized North American competitors, but also to non-unionized North American competitors and to global competitors. More importantly, they must also figure in various scenarios both for EV adoption and the introduction of AI based robotics. Significantly, Chinese EV manufacturer Nio has recently announced plans for a motor manufacturing plant that would produce 400,000 motors annually with a staff of 35. Contrast this to the US, with 10 million engines made by approximately 60,000 workers. That’s a 98.5% reduction in the number of workers, union or otherwise, per vehicle. An “optimal” solution for the UAW that forces the OEMs to follow Nio’s lead will not deliver what they are fighting for.
Is there a Pareto optimal solution for both parties? Today, the parties present their proposals and counter proposals after they go back to their corners and painstakingly try to calibrate the outcomes. While better than the old days before computers, it pales next to having the ability to tap into an exquisitely designed AI programs which can immediately update the probabilities and outcomes. Even better, a “neutral” and shared AI could present both parties with the logical steps over time in the negotiation and allow them to mutually select a shared best outcome.
Our approach to help CEOs and their leadership teams understand the full impact of AI is to execute the following:
1. Impact Assessment
Macro: Create a framework and model to understand the impact of AI on the industry’s business model as depicted through Porter’s Five Forces and Value Chain frameworks on a dynamic basis.
Micro: Understand the impact of AI on the company’s value proposition to each of the company’s stakeholders over time.
2. Corrective Actions
Identify the short- and long-term actions necessary to offset vulnerabilities and enhance strengths in the company’s value proposition.
Understand how these assessments alter the value creation opportunity as depicted through Achtmeyer’s Full Potential Paradigm™ framework, delivering revenue, margin, and market value projections.
Translate impact into KPIs for each stakeholder.
3. Monitor Doggedly
Develop and deploy ongoing measurement and monitoring tools based on the Acropolis Advisors 21st Century Stakeholder Optimization Framework and Dashboard.
AI is here to stay. It will become more powerful and pervasive every day, month, and year. It can be a tremendous force for business leaders to leverage. “Great” companies will embrace and utilize AI to their advantage. As I mentioned earlier you are welcome to reach out to us to talk about your specific issues.