Why traditional consulting is doomed

In its current form traditional management consulting, the business of providing advice for money, is doomed.

This may seem quite a bold statement. But I believe that anyone who understands the industry and takes a long unemotional look at the developing technological landscape will conclude the same thing. A couple of years ago Christensen et. al.’s HBR article Consulting on the Cusp of Disruption caused quite a stir. I think that in ten years we will look back on the article and it will read like a rather tame advertisement for McKinsey Solutions.

The thesis of “Cusp” — that sometimes clients just want data and an answer instead of a heavy and complicated engagement with a consulting firm — is absolutely true. But it rather understates the pressures consulting is going to face in light of developments in artificial intelligence and big data. Some people, somewhere have undoubtedly had this realization already and are hard at work building the consulting firm of the future. In the meantime, the rest of us in the industry owe it to ourselves to wake up.

Of course all broad prediction about a “traditional” industry being “doomed in its current form” can degenerate into a clever sort of shell game. Industries are always changing and it’s easy to be right if you don’t define all your terms clearly up front. Consulting has always been changing. In the 60s and 70s, for example, strategy firms like BCG had production departments that produced transparencies for presentations delivered via overhead projectors (hence, incidentally, the term “deck”). Nowadays, of course, we have PowerPoint, and even senior partners can crank out their own presentations when they have to. These sorts of shifts in the division of labor are going on constantly in consulting, as are growth and shrinkage in the size of the industry and of individual firms. Even new rent-a-consultant companies like SpareHire and the cringeworthily-named HourlyNerd are not truly disruptive. These brands are fundamentally parasitic; advertising that you can offer “ex-McKinsey consultants starting at $60 an hour” assumes that McKinsey exists and commands buyer respect. Such companies may nibble at the fringes of big-firm revenue, but they will not replace those firms.

What we are discussing here is not “change as usual” but much more fundamental set of changes that will be wrought by technology. The issue of consulting’s disruption becomes clearer when you focus clearly on what consultants do. A careful examination reveals that the four principal activities of consultants are all inherently disruptable. Taking them in order:

Diagnosis and problem identification, the art of determining “what is the problem,” is a machine-duplicable process that has been managed by expert systems in other fields for decades. The medical expert system Mycin beat doctors in diagnosing bacterial infections way back in the 1970s. Expert systems gained traction in numerous fields throughout the 1980s, limited only by the expense of computing resources and the high cost of engaging experts to build knowledge bases. Nowadays, computing resources are cheap and the developing world is full of experts on every topic, who can be engaged easily and remotely. It is only a matter of time before expert systems are applied to simpler, less life-critical problems like “What is the best way to structure and run my project office?”

Data collection is traditionally done through interviews, surveys, and research. All of these activities can be replaced or at a minimum improved through the use of big data. The ability to crawl the web and create a broader, richer data set will inevitably change how consulting research is done. Having a big-data answer to a business problem will become an expectation rather than a novelty.

Analysis, the art of crunching data and using it to develop insights, will become computationally-driven analytics. Of course in many cases consulting firms are using analytics already; the difference is again that this will become an expectation rather than a novelty, applied to a broader range of problems as the world’s store of easily accessible data increases.

Reporting, the task of writing and developing reports and presentations, will become increasingly automated due to the use of natural language generation software. Companies like NarrativeScience are already developing sophisticated products that greatly simplify such work and create human-readable sentences and paragraphs.

Of course the common objection to all such arguments is “Machines could never do those tasks as well as people.” This objection, even if true (that remains to be seen), is irrelevant. Even if machines “can’t replace human beings” in consulting work, they can certainly augment human capabilities and undermine the current division of labor in an unprecedented manner. Under new technological conditions the traditional pyramid structure of most firms and the economics of the billable hour might no longer make defensible business sense. The management consulting firm of the future might consist of a handful of skilled experts and data practitioners, providing statistically informed recommendations in highly specialized areas of expertise.

As routine work moved offshore during the 1990s and 2000s, many firms worked to become “more strategic” and instead solve higher-level problems for client sponsors. In the consulting world of the near future, this might turn out to be exactly the wrong approach. Implementation work may turn out to be far more durable, since it involves complicated day-to-day interaction with client personnel on real projects. Conversely strategy work may end up coming the most disruptable consulting of all, insofar as it involves providing speculative recommendations based on research and experience, which is ideal work for expert systems. Indeed, business school case-cracking is the exact sort of thing that expert systems excel at, insofar as it involves abstract reasoning and the application of knowledge from similar cases. That is one reason that prestigious firms have always been able to deliver this work with bright young MBAs who are highly educated, but not necessarily experienced.

If consulting were disrupted in the above manner, who would benefit? I believe the winners would be a select few disruptors, plus consulting clients themselves, who would have access to better insight and advice at much lower cost. The losers would be anyone in consulting with a muddled or unclear value proposition. Disruption might also open up completely new markets, such as a “retail” model of consulting at a lower price point, that could be used by more buyers with smaller budgets.

Regardless of how consulting develops in the next 5-10 years, consulting firms themselves should beware anyone who says they can help “guard against disruption” by helping them do any or all of the above preemptively. This claim misses the entire point about disruption which Christensen made in The Innovator’s Dilemma, which is that any step you take to preempt your own disruption may in fact bring it about. That is why disruption is a “dilemma” and not simply a problem to manage.