"McKinsey in a box": The end of strategic consulting?

At the same time, the late Professor John D. C. Little of MIT and others in the field of management science speculated that the massive amounts of data being culled from optical scanners, credit cards or pharma supply chains (what we call "big data" today) would go unanalysed due to a lack of human data scientists.

In the 1990s we all speculated that the future would bring about a three-layer stack solution. Sitting on top of bare metal, an operating system and some databases, a layer of "AI PhDs" would analyse the data using a multitude of methodologies. A second "authoring" layer would write about the findings of the analysis in a variety of formats (email, research report, academic study, PowerPoint presentation, video, audio, images, etc.). Finally, a third distribution or marketing layer would automatically distribute the generated content (the output from NLG or natural language generation, a branch of what we call generative AI today).

The original data could be of any format (e.g. text, numbers, images, etc.), and the output format would likewise be of any format. That future is now here.

DMG in a nutshell

Consulting is formulaic. Consultants often promise a "method", as opposed to an outcome, especially in the field of strategy. Rather than promise a doubling of share prices, they will "have their best people look at it from multiple perspectives, using the latest data and theories, to come up with a customised recommendation to create sustainable value." 

Translation: they will use whatever makes sense, and bring in experts as needed.

DMG is a fancy way of restating what consultants often promise. As the name implies, it allows content to be generated interactively, with the flexibility to change methods (i.e. algorithms, business frameworks), data sources, and outputs in real time while leveraging a "best practice" approach.

DMG envelopes:

• Multi-methods: rule-based, cognitive, symbolic, computational, control, reinforcement
• Multi-modal inputs/outputs: text, images, videos, audio, PowerPoint presentations
• Mix of experts: algorithms mimicking weather forecasters, agronomists, marketers, etc.

A prompt such as "I need a growth strategy for barbless fly-fishing hooks sold by a company that has only 50 employees located in France" yields a 10-page PowerPoint deck that replicates the consulting formula in about two minutes. It uses data and information with provenance and draws strategic conclusions. 

This is not confined to growth strategy but any strategy: HR, import/export expansion, IT strategy, AI strategy, mergers and acquisitions… you name it. Any value chain or geographic product market, any size/revenue of the enterprise and even down to an individual's personal career strategy.

Early tests

An INSEAD executive education client, a multinational bank with approximately 50,000 employees, wanted to reduce salespeople's time spent on researching their customers, especially SMEs, and increase revenues. Their sales team aspired to become "trusted advisers" as opposed to people pushing financial products. Soft-selling often translates to more revenue compared to hard-selling. 

During a co-creation workshop, one of the salespeople defined trusted adviser as "you know, like McKinsey". The "McKinsey in a box" phrase was born. The team wanted the output to be dynamic, with rotating topics, while fully tailored to the specific SME and their value chain. After initial pilots, the DMG methodology was scaled within the bank’s home market, with plans to expand to other markets.

This bank has now become Xavier AI’s early commercial adopters, using the platform to refine their strategic planning and sales approach at enterprise scale.

AI career strategist

The DMG methodology was shared with companies, executives, INSEAD students and others who used it to write strategic plans, pitch decks and, of course, class assignments. Students and alumni also used the system to help them plan personal career strategies.

For example, Xavier AI can produce a “concierge” deck that points MBA students to the classes to take, professors to engage with and logical paths to success based on the student’s CV, LinkedIn profile page and career aspirations. 

Once a student sets an industry or position to target, Xavier AI can offer more granular data, including employment options and companies that are recruiting. Finally, when a student has secured an interview, Xavier AI dives into the company, its strategies, how the student's background can help the company fulfil its mission, details about the company’s typical interview process, and how to stand out vis-à-vis other candidates.

Commercial launch

Since its first applications as a market-driven AI pilot and classroom tool, Xavier AI has evolved into a commercial platform that addresses a fundamental market inefficiency. Its AI solution delivers affordable, high-quality consulting through AI-generated decks for strategy, management and marketing functions.

Unlike traditional language AI models which are prone to hallucinations, Xavier AI leverages a proprietary reasoning engine and specialised agentic workflows that tap into relevant industry databases, news, public financials and internal documents.

This represents a fundamental advancement over existing AI approaches that often struggle with factual accuracy in complex business contexts.

So what's next?

Xavier AI represents something more profound than mere automation of business advice. It democratises strategic decision-making at a scale never before possible. Small businesses and start-ups can now compete on more equal footing with larger corporations. 

Traditional consulting firms, already under pressure from a post-Covid slump, will need to pivot towards higher-value implementation and change management services. Because the question is no longer whether AI will disrupt strategic consulting, but how quickly businesses will adapt to a new reality where quality insights are available to anyone with an internet connection.

Philip M. Parker is a Professor of Marketing at INSEAD and the INSEAD Chaired Professor of Management Science.

Disclaimer: Any references to companies or organisations reflect the author's individual views and do not constitute endorsement by INSEAD.