February 7, 2025
Artificial Intelligence

What AI will – and won’t

  • January 22, 2025
  • 0

Fast Company ran a great article titled “Consultants beware, AI is coming for your job”. It got my attention for many reasons least of which is that consulting

What AI will – and won’t


Fast Company ran a great article titled “Consultants beware, AI is coming for your job”. It got my attention for many reasons least of which is that consulting has been my career for several decades. But it also got my full attention as I have seen this change coming for some time now.

The well-written Fast Company piece documented the experiences, concerns, observations, etc. of faculty at NYU Stern.  Collectively, they are teaching the next generation of consultants and are, correctly, addressing the challenges these students will face especially as AI changes the business landscape and their long-term careers.

Quotable article highlights

That Fast Company article has some great sound bites including: Consultants will no longer be the “smartest people in the room”. The authors note that “For the next generation (of consulting work), answers are the new commodities and immediate access to insights are table stakes.” I’d agree but like in most things, the devil is in the detail. AI is changing consulting but there are limits as to what AI will do. Smart consultancies will need to carefully restructure their business models, their approach to developing talent, how they bill and more. It’s a complex change that’s underway.

What I took from that piece is that the changes impacting the consulting industry and its participants will impact other industries and careers as well. AI’s ability to deal with tabulation, summarization and other time-consuming tactical activities in very short timeframes frees up consultants to work on more creative ‘craftsmanship’ roles. Every job/profession impacted by AI must re-discover what exactly makes them ‘valuable’ to clients including consulting.

Consulting historically

Consulting was always been a multi-faceted profession. The best consultants possess numerous technical, analytical, interpersonal, business and other skills. They also have great listening and detective skills. They can read a tough group of executives like it’s nothing special at all. They can turn over all kinds of rocks and find the key piece of information that everyone else has missed. They can also guide clients through difficult changes even when the client fights them the entire way.

Great consultants act in the client’s best interests. They take on a fiduciary relationship, have impeccable ethics and act with great judgement.

I love being a consultant. Over the years, I’ve learned that many clients genuinely want access to a consultant that:

  • Brings a significant measure of their own political capital, intellectual property and expertise to the project
  • Speaks the straight, honest truth. These executives have no use for a yes-person/toady. While the truth can sometimes hurt, great consultants find sucking-up to be unprofessional or unethical.
  • Acts fairly and with a measure of diplomacy
  • Inspires client personnel
  • Drives material change
  • Supports their recommendations with data
  • Etc.

Consulting problems

AI tools can be great for many tasks that used to generate untold billable hours for top-drawer (and down market) consulting firms. Newer AI tools (and related advanced technologies) are capable of eliminating a lot of low-value-added work. Those AI tools can:

  • Compare financial, operational and other metrics across a client’s competition
  • Search for solutions/competitive responses that other firms have used
  • Generate work plans for proposed initiatives
  • Create lists of relevant actions the client might want to consider
  • Develop executive presentations
  • Forecast and model future states for the client

And more

This is where consultants and MBA students are struggling today. In the past, a typical management consulting gig might see 95% of billed hours spent on collecting data and analyzing it. Very little time was actually spent advising the client. If AI materially shortens the data collection and analysis activity, then fewer consultants are needed and total project costs can be reduced.

AI tools can also tackle a number of tasks that have been the purview of others that call themselves consultants. This can include outsourcers, software implementers, managed service providers, etc. These tools are capable of taking over large aspects of work activities such as:
 

  • Package software configurations, setup and upgrades
  • Software selections
  • Application software testing
  • Performance monitoring
  • Data cleanup
  • Report development
  • Identifying anomalous transactions
  • Improving systems security
  • Robotic process automation

And more

The tasks AI is taking over are often transactional, repetitive and/or rarely provide competitive advantage. The lack of competitive advantage is a byproduct of the data within the AI training databases. These tools can only play back ideas, concepts, data/results, etc. that other firms have already tried and implemented. AI databases are full of old insights not radical, novel, revolutionary concepts. In other words, if a company wants tailormade, innovative ideas, it may benefit from tapping into the imagination and idea cross-fertilization possible within the minds of clever human beings.

If fewer staffers are needed in these consultancies, what will that do to billing rates, chargeable hours and the apprentice training that the traditional consulting business model is based on? The short answer is that consultancies won’t need large staff-to-partner pyramids. That, in turn, can have an adverse effect on partner earnings unless new ways to monetize value delivered to clients is implemented. Without a robust apprenticeship process, where will the next generations of consultants come from and what will they know/learn if they don’t do a fair bit of the research, analysis, etc. that AI tools will do? The apprenticeship model of consulting is being challenged by AI.

What AI can’t replace

Over the decades, I’ve learned what consulting really is and what it is that clients really want and value. And, those things are not strengths of AI. In short, these are the defining characteristics of great consultants:

Pushing – In a nutshell, clients pay me to push them. Yep, they are using an external person/firm to push, shove, cajole and even badger them into moving off of the status quo. Frequently, clients already know what they need to change, but, they just don’t have either the leader, the plan and/or the means to motivate people off of the incumbent processes, systems and business methods. Consultants get rich every time they hear a potential client waxing eloquently about their old practices. These firms can talk wistfully about old business methods like your grandmother can reminisce about ‘the good old days’.  These clients need someone to not just shove the firm into the present but to keep it moving until the project is fully done. Inertia and nostalgia may be common client shortcomings but they’re gold to consultants.

Last I checked, AI can’t do this.

Stating the (Hard, Plain) Truth – When it came to break a particularly obstinate or ornery horse on the ranch, I sometimes saw a ranch hand head into the corral with a piece of 2X4. He wasn’t trying to be mean, vicious or cruel but some horses needed more than whispered thoughts to get their attention.

Clients can be the same way. Some will stubbornly stick with old business paradigms, people, business practices, outdated notions of their competitors, etc. While many of those items were relevant at some time in the company’s history, they aren’t anymore.

Consultants sometime succeed in getting top management’s attention to what’s really going on even though others in the firm have tried to (and failed) many, many times previously. The telling though is a bit of artwork and something AI can’t directly address. This is where top consulting partners really earn their pay. They know when, how, where and what to say for maximum impact. Sometimes they can even make it seem like it’s the client’s idea after all.I lacks empathy and can’t read a room.

Observation – Many, many times in my consulting experience, I’ve uncovered massive items that an AI tool might not have uncovered. These items might be one-off, well-hidden, undocumented, soft (like culture or interpersonal dynamics), criminal, etc. items. Whatever they are, they don’t necessarily show up in an AI training DB. Here’s just a short sample of some of the big things I found while working on consulting engagements:

  • One company had a global EVP of Manufacturing running their production plants at full capacity so as to meet a key inventory metric of zero stockouts of finished goods. This had three knock-on results: the company’s supply chain was choked with inventory; many products expired long before they got anywhere near a retailer or end consumer; and, the EVP got a bonus equal to twice his considerable salary. The CEO suspected something but it was my observation that the EVP was possibly putting personal financial gain ahead of the company’s interests that triggered a leadership change.
  • In another situation, I uncovered another executive behavior/compensation plan problem. This time, the wrinkle was that the CEO of this huge publicly traded firm was so scared of the leaders of his largest manufacturing plants that I had to suggest less impactful projects. That was definitely a one-and-only situation in my career.
  • Another client had grown significantly over the last 30 or so years. The growth was mostly via some 50 or so inorganic acquisitions. Prior regimes had quickly assimilated each acquired firm but these post-merger integrations were mostly just skin-deep. Of all of the unfinished work, management was almost exclusively focused on the dozens of redundant and different ERP systems in use. But some of the biggest things my team observed that needed fixing were that the company had not altered its supply chain, fulfillment, distribution logistics and other factors that would make everything from order handling, transportation and carbon consumption more valuable for the company, its shareholders and its customers. Maybe an AI tool could have noted some of this but I doubt it could have rationalized all of the different and incomplete data let alone have built the one-off plan to bring order to this colossal set of challenges. That skill requires an architect’s attention to detail, internal politics, compensation, customer relationships and more.
  • I sold and closed a $66 million shared services project simply because I observed who my competitors were and correctly guessed how they’d approach such a deal. I then crafted a proposal unlike the competitions’ and one that hammered on a number of concerns this Swiss client had – notably in the areas of security and privacy. My approach even surprised the client’s Group CFO who wanted to meet the shoot-from-the-hip American/Texan consultant who took these concerns far more seriously than his staid, conservative UK competitors.

AI is great at regurgitating old data and it can also react to the data you feed it. But, it doesn’t observe anything and humans (like my German Shepherd) can spot things that don’t fit a well-worn pattern or may never have been previously documented. AI can’t quite replace that.

Connections – The very best consulting leaders have an unbeatable rolodex. They can connect top executives to peers, to thought leaders, etc. They can find not just a suitable project team from within their firm’s personnel but they can get the very best people for this client’s project. They can also introduce a client’s leaders to the top people in key tech firms, potential acquisition candidates, potential alliance/partnership firms, etc.  AI’s rolodex is non-existent.

Get it Done – The best consultants aren’t just idea people. They eat what they kill. They have to have the data, facts and courage to bring their proposed initiatives to life. The best consultants put their professional reputation on the line with every project.  In my opinion, no amount of LLM training will replace the tacit and explicit knowledge that a consultant with many successful projects under their belt possesses.

Prioritization – Prioritization refers to the ability of a consultant look at the universe of potential client initiatives, the timing and resource requirements for each, the probabilities for success, the magnitude of the change, the potential benefits, the sub rosa political dynamics, and many other factors into a logical, economical and powerful game plan. Lists of projects are of little value to clients just as are plans that fail to take into consideration dozens of human factors. AI might make a start at this but a seasoned hand on the tiller might generate a far better and more successful program.

My take

The article’s authors are correct: AI will fundamentally change consulting but the real challenge is knowing exactly where and how AI will impact this industry. For example, the whole apprenticeship model must be rethought. Value billing will be another issue. Consulting partners will likely have spirited internal conversations on the effect of smaller pyramids on partner income. I can go on but you get the point. A little tech, like AI, can sometimes have a profound, disruptive impact on an entire industry.

Buyers of consulting services will have opinions about AI consulting ‘solutions’ too. If your firm wants to be like their other run-of-the-mill competitors, you will find an array of AI capabilities available to you from AI vendors or via slimmed-down consultancies. But for those who want more than this, a real, strategic consultant may be needed. The challenge is finding a real strategic consultant with the expertise, insights, etc. that have been tempered by real world client work. This is a human factor that AI cannot replicate. If you need ‘innovative’ ideas, I wouldn’t rush for an AI solution.

The Fast Company authors got this point dead-on and it’s one we should end on: 

This is consulting’s true essence, and as long as AI knows everything but understands nothing, there will always be room for humans in consulting.  

True. Now, go read the article!



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *