Formerly limited to theoretical and somewhat obscure applications, Artificial Intelligence (AI) and Large Language Models (LLMs) have, for slightly over a year now, been making significant inroads into day-to-day business operations. The launch of OpenAI’s ChatGPT4 in November last year is considered by some as a pivotal event in this development. As I’ve emphasized in my blog, the resulting shift in the labor market could be unprecedented. This shift is notable not just for the sheer number of activities and workers impacted, but also for the type of workers affected: white-collar professionals. This demographic has rarely been so directly targeted by technological innovation. Strategy consultants are certainly not exempt from this trend. However, I firmly believe that these technological advancements do not spell the end for the world’s leading consulting firms (McKinsey, BCG, Bain or ‘MBBs‘). Instead, they suggest a transformation in the role of the strategy consultant, especially at junior levels, which is likely to be beneficial.
First, let’s examine the types of consulting-related activities that have shown potential for AI-related productivity gains:
- Since its inception, ChatGPT4 has proven itself exceptionally capable in various tasks, even passing rigorous tests with ease. More commonly, LLMs have demonstrated their prowess as top-tier copywriters. I confess to using it occasionally to ensure my non-native English appears fluent.
- Notably, tech giants like Microsoft and Google, albeit to a lesser extent, have been integrating AI and LLMs into their products. Microsoft Viva Topics, for instance, continually scans an organization’s vast array of data and documents, simplifying internal knowledge management.
- More recently, Microsoft revealed Copilot, which is now being incorporated into the latest Microsoft 365 updates. This tool is expected to reach the laptops of consultants soon, aiming to facilitate the creation of emails, documents, and slides from existing materials and user prompts.
- In the consulting industry, comprehensive studies have been relatively rare to date. To my knowledge, BCG is the only firm that has recently published an in-house study on the impact of ChatGPT. Among various findings, the study indicates that ChatGPT excels in creative and brainstorming tasks. In fact, human interventions often seem to diminish the quality of the output in these areas. Conversely, ChatGPT appears to struggle with complex business problem-solving tasks, where its use has been found to reduce human output quality.
Assuming that these technologies will remain outside the consulting realm, due to either a lack of adoption or concerns over data security and privacy, is illusory. The risk of falling behind competitors who embrace these technologies is far too great. Consulting firms, like all other businesses, must adapt to this change rather than trying to avoid it.
For junior consultants, I anticipate that these technologies, and their subsequent evolutions, hold the potential to be game-changers. Practically speaking, I expect the significant portion of time currently dedicated to gathering and organizing client-specific and public data (including, but not limited to, quantitative data), as well as creating slides, to decrease markedly. This might seem like great news for young graduates at first glance.
However, the reality is more nuanced. I believe the time saved will be partially reallocated towards improving work-life balance to some degree, depending on labour market conditions and the intensity of the competition for talent within consulting firms. Additionally, more time will likely be spent on complementary activities where AI currently shows limited effectiveness, thereby positioning AI as a ‘complement’ rather than a substitute. For instance:
- For AI-powered tools to perform optimally, consultants will need to provide them with context-rich, easily understandable ‘prompts‘. This means that consultants will have to invest more time in thoroughly understanding their clients’ requests and breaking down larger problems into smaller, machine-digestible components. BCG’s study suggests that larger, more complex problems are currently too challenging for these tools.
- Furthermore, it’s important to remember that AI outputs cannot be taken at face value. Studies indicate that AI algorithms can generate inaccurate or even fictitious information. Consultants must reinvest some of the time saved to scrutinize the machine’s conclusions, pose relevant follow-up questions, and probe the data themselves when necessary.
- Finally, as I have previously discussed, it’s essential to involve clients in the AI journey. This includes not just the business aspects (what AI can contribute to their business) but also ethical, security, and privacy considerations (what can be performed and under what conditions). Consultants must act as catalysts for change throughout client organizations, not just within their IT departments.
Looking backwards, what does this imply for the typical consultant recruitment journey? The situation is even more open-ended here:
- The interview process, which currently includes a combination of numerical/quantitative testing, informal soft skill assessment, and business acumen evaluation (through case studies), seems well-suited to the current needs and should remain largely unchanged. However, some consulting firms might introduce a ChatGPT prompt test to assess candidates’ ability to interact effectively with machines.
- The outlined day-to-day tasks will likely place greater emphasis on business experience over pure intellectual capability. We might witness consulting firms shifting their hiring focus more towards post-MBA hires at the expense of recent graduates. Additionally, more frequent ‘secondments’ to client sites could become common, allowing consultants to enhance their business knowledge through real-world exposure.
In any case, the consulting industry is likely to undergo a complete transformation. BCG’s study indicates that performance rankings are significantly disrupted by the introduction of ChatGPT in areas where it excels. Today’s top performers might not retain their status tomorrow. The implications for performance management and training, especially in topics currently absent from today’s consultants’ curricula, are clear and significant.
In conclusion, the rise of AI and LLMs, especially in strategy consulting, demands a proactive stance from firms and consultants. Far from a passing trend, these technologies signal a major shift, reshaping roles and industry standards. Firms must integrate these changes into their operations and training programs. For consultants, particularly new joiners, this era is a call to enhance skills, understand AI’s nuances, and adeptly address client needs. Success in this evolving landscape indeed hinges on blending, and not replacing, human expertise with AI.