Last week, McKinsey and Company launched their generative AI tool, Lilli.
Here’s how they described it:
If you could ask the totality of McKinsey’s knowledge a question, and [an AI] could answer back, what would that do for the company? That’s exactly what Lilli is,
According to them, “Employees can prompt Lilli with a question, and the tool will scan the knowledge base, locate five to seven relevant pieces of information, summarize key points, include links, and identify experts in the field.”
This is the first of many use cases that will help us reshape our firm.
While every company and their mothers are claiming that AI is going to change everything (and hoping their investors agree), I think this is really interesting for McKinsey.
A little context on my perspective first:
In 2008, I joined McKinsey’s research center in Boston. I was focused specifically on manufacturing and supply chain knowledge and worked directly with client teams around the world working on these kinds of projects. I had a background in manufacturing engineering, had worked at GE, and also had experience working on factory floors, which made me a good fit for the role.
McKinsey was one of the first firms to invest heavily in “knowledge management” in the early 2000s.
Before going under the fancier sounding “knowledge,” most consulting firms had libraries with physical books and resources that were managed by professionals with a background in library science.
They were experts in helping people find information.
Over time as computers reshaped the work of consulting firms, the more talented information specialists also helped consultants synthesize information. This transformed into the role I was eventually hired into, the “research analyst.”
While I still spent some time searching for information, computers had made that much easier, and I spent most of my time doing research, analysis, and synthesis directly for client teams.
McKinsey had also invested heavily in making this a competitive advantage. Instead of buying off-the-shelf knowledge management platforms (like SharePoint and others), they built their own internal system. It was really good. You could search this database to find sanitized client project reports, industry and functional deep dives, listings of internal experts, best practices, and more. You could also upload your own material to this network.
Part of my job was to make sure that everything within the manufacturing practice uploaded to the system was top-notch. I myself uploaded lots of stuff (probably foreshadowing my future creator path) and I am guessing some of it is still being used by client teams.
I give you this background because I believe that AI systems will only be as powerful as the underlying information and knowledge that a company can leverage.
I think McKinsey will be at a massive advantage because it was one of the earliest firms to build a robust knowledge management function and the high-quality software to match. For years, it had more people in these roles than other firms (though BCG started investing heavily in the mid-2010s and Bain after that) and because of this built up a large internal database of valuable knowledge.
McKinsey claims that the current system, “already cut down the time spent on research and planning work from weeks to hours, and in other cases, hours to minutes.”
What this means is that McKinsey will likely be able to give its front-line client teams more valuable information, more quickly, and at a cheaper cost.
This means more profit and more investment to help them continue their impressive growth, despite being an already large company.
Big firms are set up to dominate even more…
I think AI tools will help firms like McKinsey, BCG, and Bain (though Bain was much later to invest in knowledge management) develop a bigger lead in management consulting.
These firms will likely be able to:
- Serve existing clients across a wider range of areas by mining their internal capabilities for opportunities.
- Invest heavily in software solutions to sell ongoing services to existing clients at higher margins. AI will give them leverage for coding support and identifying opportunities.
- Scale back on junior-level researcher hiring and bring in less but more skilled specialist / expert-level talent that can work with AI to deliver a high-touch experience with clients (this has already been happening, but I imagine will be accelerated, especially if AI tools can help automate some of the gritty work of building slides and reports)
For firms without these resources (my typical client), here are the questions I’d be thinking about:
- If I’m not investing in codifying our existing process and knowledge, who is going to be in charge of it? How will I build this capability?
- How am I thinking about injecting AI into our process to save time and money? What are the best use cases? How can I build this capability internally and not depend on external software solutions?
- What are the unique capabilities I can build that may be valuable if we were to get acquired in the future?
- Who are the people who may be undervalued in the talent market that we can hire who are comfortable working with large amounts of data? (For a long time, this was data scientists, but this is probably more accurately priced now)
I think there is going to be a lot of excitement around AI over the coming years and you’ll likely see a lot of M&A as the bigger firms acquire smaller firms that develop unique capabilities.
I also think you’ll see a lot of wasted effort on solutions that clients may not care about. The biggest advantages will ultimately go to the firms that have developed unique internal systems and capabilities that enable them to do great work in different situations AND the ability to change course when things are not working.
A lot of people look at consulting from the outside and think something like AI will hurt demand for consulting.
At the margins I think it will definitely reduce demand, but mostly for the kind of high-effort give-me-a-bunch-of-young-bodies kind of work that big companies sometimes use consulting for. AI will also increase demand for many opportunities that are ultimately unknown to both clients and consulting firms right now.
As the computer scientist Stuart Russell has noted, AI can only do what you tell it to, not what you want it to.
Which is to say, knowledge and experience will continue to be more valuable than ever.
Related Essays
And if you’d like to explore this thinking further, here are some resources you can spend time with:
Do you have a toolkit for business problem solving? I created Think Like a Strategy Consultant as an online course to make the tools of strategy consultants accessible to driven professionals, executives, and consultants. This course teaches you how to synthesize information into compelling insights, structure your information in ways that help you solve problems, and develop presentations that resonate at the C-Level. Click here to learn more or if you are interested in getting started now, enroll in the self-paced version ($497) or hands-on coaching version ($997). Both versions include lifetime access and all future updates.