Playing To Win
Will Artificial Intelligence Eradicate Practitioners of Strategy?
Yes & No
The question I get asked the most these days is whether Artificial Intelligence (AI) will wipe out the human practice of strategy. I have done a couple of pieces on AI previously — Strategy & AI and Investment Strategy & AI — but not on the impact of AI on the business of strategy. So, this Playing to Win/Practitioner Insights (PTW/PI) piece is on Will Artificial Intelligence Eradicate Practitioners of Strategy: Yes & No. And as always, you can find all the previous PTW/PI here.
The Practice of Strategy
There are many people working either full-time or part-time on the practice of strategy. While the most obvious might be the consultants working in the strategy practices of management consulting firms, like McKinsey, BCG and Bain, there are probably more working at in-house strategy functions at companies. In total, there are likely tens of thousands of both around the world. I will collectively call them practitioners of strategy — or PoS, for short.
The answer to the question is a bit of both ‘yes’ and ‘no.’
The ‘Yes’ Part
Yes, AI is going to eradicate a lot of PoS tasks and jobs. Many PoS are engaged in performing tasks that AI can do faster, more thoroughly, more accurately.
I have watched this before. During the summer of 1980 between my two years of Harvard Business School, I worked as an investment banker, at the time the second most popular destination for our graduating class. I spent the vast majority of the summer producing documents with what would now be considered barbaric technology. Prospectuses were printed on hot lead presses, and we spent countless hours at the printers proofing prospectuses, finding errors, having the typesetters replace letters, and then proofing the next iteration without the benefit of redlining. Client pitch decks were produced on IBM Selectric typewriters.
Investment banks needed scores of us willing MBAs to perform these tasks. But soon electronic publishing replaced hot lead, and prospectuses were proofed on computer screens, aided by spellcheck and redlining of changes. Sophisticated graphics software turned pitch decks into a snap thanks in part to databases of templates from prior presentations which made cut-and-paste the norm.
Lot and lots of those MBAs were no longer needed and now those investment banks hire more for their sales and trading desks and quant departments than for investment banking.
AI is the latest technological disruption to high-end professional jobs — maybe the most disruptive yet. There are innumerable PoS doing customer research, creating market maps, putting together analytical decks, and explaining strategy principles. Most of that is completely going away.
I will illustrate that with an extremely rudimentary use of Claude — my favored AI tool these days. I asked Claude: “Create a typology of all winning business strategies.” It gave me six categories (e.g. market position strategies, competitive behavior strategies) and under the six, twenty sub-categories, one of which was differentiation.
I then asked: “Please provide the various ways to win with differentiation.” I got back five core differentiation approaches, including product/service quality differentiation and customer experience differentiation plus four implementation strategies, including pricing approaches and communications strategies.
But I wanted to figure out how to make winning long-lived, so I asked: “How would I make sure that my advantage is sustainable?” I got back eight elements including secure your supply chain and develop proprietary technologies.
That was three questions and about five minutes. The implication: no company is going to pay a high-priced PoS to explain the various means of differentiation and the assorted avenues for making it sustainable. Similarly, they aren’t going to pay high-priced strategy PoS to put together a ‘market map’ or a lot of other things.
These are tasks that have become algorithmic, not dependent on expert heuristics — per my earlier PTW/PI AI piece. The big management consulting firms have long since figured that out and taken the interim step of offshoring algorithmic consulting tasks to cheap labor jurisdictions.
Those offshored PoS jobs will soon be history — as with many, many PoS jobs in high-income countries. But how far will it go?
To get more insights, I went back and did a more sophisticated Claude prompt. A friend recently asked for advice on this strategy challenge. I gave Claude the exact situation — but I disguised which medical profession and what product:
“I need help with strategy. We are a provider of a medical product used by medical professionals routinely in their practice. The individual medical practitioners purchase our product through one of two very large and broad distributors. It is a relatively small cost item in the range of items the medical practitioners purchase from the distributors. Our product is branded and is has a long history of use. We are the market share leader with 70% global market share and enjoy attractive profitability. However, our product is not physically differentiated. What strategy should we pursue to protect our dominant market share and our profitability?”
Claude’s answer:
Based on your situation, here are strategic recommendations to protect your dominant market share and profitability:
1. Strengthen relationships with distributors
Develop exclusive or preferred partnerships with your two key distributors; Create incentive programs that reward distributors for prioritizing your product; and Provide superior service and support to distributors compared to competitors
2. Brand reinforcement
Invest in marketing that emphasizes your product’s heritage and reliability; Create educational content for medical professionals highlighting your product’s consistency and trusted history; and Develop a robust testimonial program featuring respected practitioners
3. Bundling and value-added services
Offer complementary products or services that enhance the value of your core product; Create educational programs or certifications for medical professionals; and Develop digital tools or apps that integrate with your product
4. Loyalty programs
Implement practitioner loyalty programs with meaningful benefits; and Consider volume-based discounts or rebates for consistent purchasers
5. Incremental innovation
Pursue small, meaningful improvements to your product; Focus on packaging, ease of use, or complementary features; and Patent these improvements when possible
6. Active competitive intelligence
Monitor competitor activities closely; Maintain pricing vigilance and be prepared to respond strategically; and Track market trends and evolving practitioner needs
That is a good answer — especially for free and in another five minutes. Claude gives a good list of sensible things to take into consideration — and how to think about them. The creation of a thoughtful list is a valuable act. The Five Forces model is one of the greatest strategy tools ever invented and it is simply a thoughtful checklist of things about your industry (and their implications) to which you are wise to pay attention as you are formulating your strategy.
Even for an expert, a thoughtful list of things to consider is important. My lawyer wife is also a jet pilot and she knows that regardless of experience level, all pilots are expected to follow their jet’s preflight checklist before taking off. The same holds for doctors in operating rooms as Atul Gawande describes in Checklist Manifesto.
Right now, PoS mainly carry such checklists in their heads and use them to power their work. LLM/AI makes it almost entirely free in time and cost to acquire super-valuable checklists, which use everything written in the world of strategy to develop them for you — often with illustrative examples.
Measured in hours spent, what percentage of PoS tasks are going to be done faster and cheaper by AI — like these? While that is hard to predict, my estimate is that it will be at least 50% of all current PoS hours worked.
The ‘No’ Part
But for some tasks, PoS work is not going to be irradicated by AI. And because I believe the surviving PoS work will be much more highly compensated than the eradicated work — probably twice as high — I suspect the surviving PoS compensation will be on the order of 70% of the current pool.
While the following may change some day, but I don’t see AI on a trajectory of being capable of handling these three PoS tasks:
1. Intelligent Prompting
Even in the first simple example, based on the first two questions that almost anyone could prompt, I didn’t receive a real answer to winning. Those answers told me nothing about what would be necessary for a useful (in this case) differentiation strategy because to be useful at all, a strategy has to sustain its advantage. That third prompt could only come from a user who is reasonably knowledgeable about the subject of strategy. Claude couldn’t figure that out on its own. To get a useful answer, I had to know that sustainability is critical.
In the second example, to enable Claude to be intelligent in generating an answer, I had to frame the question with enough sophistication about which variables in the strategy equation are critical. Without the right features of the problem, Claude would have probably hallucinated.
As the great pragmatist philosopher, John Dewey argued, a problem well-framed is half solved. And that is a limiter on AI eradicating PoS. AI needs to have a strategically intelligent interlocutor.
2. Application to Specific
As is obvious from the small examples above, AI provides a general answer to a question related to a specific situation. In the first, my questions set up a very general context and returned very generic answers from Claude. I was far more specific in the second case but still received a general answer.
That should come as no surprise. As I have argued before, LLM/AI is a mode-seeking device. It surveys its giant database and comes back with relevant elements found most frequently — as with the six potential solutions to my second problem.
Claude left it to me to figure out which of the six is best matched to my specific context. And on this front, it is turtles all the way down. I can ask the next more specific question, and the next and the next and will get diminishing returns on applying the general strategy advice to my specific problem context. It essence, this loops back to the first limitation: the need for a strategically intelligent interlocutor.
PoS who have this ability to apply the general knowledge generated by AI to their specific situation will continue to be valued. But they need to have that specific expertise.
3. Generation of Unique Solutions
The other limitation of LLM/AI as a mode-seeking device is that it is utterly incapable of generating unique new solutions. Such solutions simply don’t exist in the LLM/AI’s world. Their world is the catalogue of what currently exists, and their search algorithm focuses on what is found most frequently within what currently exists.
PoS who have the ability to generate unique solutions will continue to be valued — and very highly.
Practitioner Insight
If you are a PoS, there are three ways in which you can continue to be valuable in a post-AI world. You can guide AI with intelligent prompting that is based on sophisticated understanding of strategy. You can take general answers produced by AI and determine how to apply these insights to the specific situation you are tasked with transforming. And you can go beyond the range of answers provided by AI to generate unique solutions to problems with which you are presented.
If you develop a personal heuristic for performing one of those three tasks with distinction, you will have a fulfilling job as a PoS. If you do for two of the three, that job will be fulfilling and really lucrative. If you do all three, you will be in the catbird seat.
However, if you are only capable on the algorithmic aspects of strategy, you should start planning for a career outside PoS. New technologies always take time to ripple through the labor market, so it won’t happen tomorrow. But obsolescence is certainly lurking out there.
Note that the three happy paths for PoS are not in the form of human versus AI. Each is a case of human PoS augmented by AI — humans using AI to get the most useful set of general answers from which they can build context-specific solutions and against which they can compare unique solutions. Not being an adept user of AI in PoS work will soon be a non-starter.
Thus, the answer to the question of whether artificial intelligence will eradicate practitioners of strategy is that it will eradicate many but not all. And the ones who survive and prosper will be those with a valuable personal strategy heuristic and who augment themselves with AI. They will do just fine!
******
As a reminder, I previewed two months ago that I am doing a PTW/PI podcast series with friend Tiffani Bova. The second in the series is on LinkedIn here on Wednesday, March 26th at 12 noon EST and 9am PST. Look forward to seeing you there!
