The Codification Trap: How Consulting Firms Engineered Their Own Commoditization

The strategy that made firms scalable made them interchangeable. The firms know it. The buyers are starting to figure it out.

In March 1999, the Harvard Business Review published a paper that shaped the next two decades of consulting economics without anyone noticing until the damage was done. The authors were Morten Hansen and Nitin Nohria, both at Harvard Business School, and Thomas Tierney, then the worldwide managing director of Bain & Company. The paper was called “What’s Your Strategy for Managing Knowledge?” and it described two approaches to building a consulting business.

The first was personalization: knowledge lives in people, gets shared through conversation and mentorship, and the firm charges a premium for access to specific experts with irreplaceable judgment. McKinsey and Bain operated this way. You hired them because you were hiring them, the specific partners and engagement managers whose pattern recognition could not be extracted from their heads and stored in a database.

Codification was the alternative: knowledge gets extracted from individuals, written into documents and templates and frameworks, stored in repositories, and reused by anyone in the firm. Ernst & Young was the paradigm case. The paper described how E&Y had built a Center for Business Knowledge with 250 people whose job was to create “knowledge objects” by pulling interview guides, work schedules, benchmark data, and market-segmentation analyses out of completed engagements and filing them for reuse. The economic model was explicit: invest once in a knowledge asset, then reuse it across dozens or hundreds of engagements. The authors attributed E&Y’s consulting growth, then running at 20% or more annually, to the economics of reuse.

Hansen and his co-authors were careful. They argued that firms should make one strategy dominant and use the other in a supporting role, rather than trying to balance both equally. They even understood that individual services could mature into commodities; the paper discusses reengineering consulting as one that had, and quotes an E&Y partner on the logic:

The objective was to commoditize expertise quickly enough to exploit scale and reuse.

What Hansen did not examine was the industry-level consequence: what happens when most large firms industrialize similar bodies of knowledge at the same time, producing convergence between competitors rather than just commoditization within a service line.

We lived through that consequence. Every contributor to this publication watched it arrive. The codification strategy achieved exactly what it promised: any consultant could apply the framework. The part no one planned for: any consultant at any firm could apply the same framework. The firms engineered their own commoditization, step by step, for entirely rational reasons, and most of the partners who oversaw the process did not recognize what was happening until the buyers started to.

What Codification Was Built to Fix

The codification model worked for a long time because the firms that pursued it were solving a real problem. Consulting in the 1980s and early 1990s depended too heavily on individual partners. If a partner with deep sector expertise left, the firm lost that capability until someone else rebuilt it from scratch. Client relationships were personal; institutional knowledge was personal; the ability to deliver was personal. That dependency constrained growth. A firm could only sell as much work as its best people could personally supervise, and it could only grow as fast as it could develop new people to that level.

Codification broke the constraint. If you could capture what your best partners knew and store it in a system that a competent senior consultant could access and apply, you no longer needed the specific partner on every engagement. You could staff projects from a wider pool. You could grow faster, operate across more geographies without waiting years for each office to develop its own senior expertise, and build a bench of competent generalists who could be deployed almost anywhere.

Ernst & Young built the model. Andersen Consulting (later Accenture) built a version of it. The Big 4 firms followed, each constructing their own internal knowledge infrastructure through the late 1990s and 2000s. E&Y had its Centre for Business Knowledge; Deloitte built practice-specific methodology libraries; PwC and KPMG invested in knowledge management platforms with dedicated staff to curate, tag, and maintain them.

The scale arrived faster than anyone expected. By the time the Hansen paper was published, E&Y’s system already indexed over a million knowledge objects across more than a thousand repositories. The early systems were heavily curated, with dedicated teams that knew which materials were current, which applied to which industries, which had been tested in live engagements. But the volume was growing faster than the curation could keep pace with. The overload problem was not a later corruption of a system that began modestly. It was latent in the success of the codification model itself.

Four Hundred Results for “Market Entry Framework”

Anyone who worked inside a Big 4 firm between 2005 and 2020 knows what happened next. The knowledge repositories grew to contain thousands of documents: frameworks, methodologies, accelerators, templates, past deliverables, interview guides, engagement plans, benchmark databases, sector overviews, regulatory summaries, model analyses. Every completed engagement was supposed to contribute sanitized outputs back to the repository. Every practice group maintained its own sub-library. Every office added local variants. Every annual planning cycle produced new methodologies to reflect the firm’s current strategic priorities.

The maintenance teams could not keep up. Tagging was inconsistent. Version control was unreliable. The same framework existed in four versions across three practice libraries, and none of them was clearly marked as current. The search function returned 400 results for “market entry framework,” and the consultant who needed one on a Tuesday afternoon had no way to evaluate which of those 400 documents was the one that would actually help with a consumer goods company entering Southeast Asia.

So consultants did the rational thing, the thing every senior manager living through this adopted without any formal decision: they ignored the system and rebuilt from memory. You knew the framework. You had used it fifteen times. You could reconstruct it in PowerPoint faster than you could find the approved version in the knowledge base, verify it was current, and adapt it. So you did. And because you were reconstructing from memory rather than copying the canonical version, your reconstruction drifted. It picked up habits from your last three engagements. It dropped elements you never used. It simplified where you thought the original was over-engineered.

Multiply that by several thousand senior managers across every Big 4 firm, and you get a decade of quiet divergence from the official methodologies, followed by periodic “methodology refresh” initiatives where the knowledge teams tried to re-establish a canonical version, followed by another cycle of drift. The codification strategy was designed to make the firm’s knowledge independent of individuals. In practice, the knowledge had become so unwieldy that individuals reverted to carrying it in their heads anyway, with the added complication that they no longer had a reliable external reference to calibrate against.

Fewer Frameworks, Same Frameworks

The firms tried to fix this, and their fix created the next problem.

Generalization. If the knowledge base was too large and too fragmented to search, reduce its size. If consultants in different industries were maintaining separate versions of conceptually identical frameworks, consolidate them. Strip each framework down to its most portable, most broadly applicable form. A market-entry methodology that could work across consumer goods, financial services, and industrials was more useful than three industry-specific versions that no one could find. A maturity assessment that applied to any organizational capability was more maintainable than twelve variants tailored to specific functions.

This made sense internally. Fewer frameworks meant less maintenance, easier training, more consistent quality across practice groups. A new consultant could learn five general-purpose frameworks and be staffed across a wider range of engagements than one who needed to master twelve industry-specific ones. The training burden dropped. The staffing flexibility increased. The utilization rate, which is the number the partnership economics actually care about, improved.

The problem is that “more general” and “more similar to what every other firm uses” are the same thing. A market-entry framework stripped of its industry-specific elements looks like every other market-entry framework stripped of its industry-specific elements. A maturity assessment built to apply across all organizational capabilities looks like every other maturity assessment built the same way. The generalization that solved the internal search problem simultaneously erased the analytical differences between firms.

There was a time, and several of us were working during it, when firms approached the same problem differently. Some strategy practices emphasized market structure and competitive positioning; others, with stronger operational roots, began from organizational capabilities, processes, and execution constraints. Deloitte’s technology practice had a genuinely different delivery culture than PwC’s. These were not marketing distinctions; they were analytical traditions that produced different diagnoses of the same client situation, which is precisely what made the competition between firms useful to the buyer. (And the supposed “personalization” firms were not immune. McKinsey and Bain both invested in codification and knowledge management as they grew, blurring the clean distinction the Hansen paper had drawn. By the 2010s, the question was not which strategy a firm had chosen. It was whether anyone still remembered the difference.)

The generalization trend ground those differences down. As frameworks were simplified and made portable, the firm-specific intellectual traditions that informed them washed out. Senior partners who carried the old analytical DNA retired or moved on. The new generation, trained on the generalized versions, had no memory of the distinctive approaches that preceded them. By the mid-2010s, the convergence was visible to anyone paying attention on either side of the table.

The Packaging Game

Rather than fix the convergence, the firms redirected their energy. The intellectual substance of the frameworks had converged, but the packaging had not. So the competitive game shifted from building better frameworks to building better marketing for interchangeable ones.

Proprietary names for identical methodologies. Trademarked terminology that meant the same thing across firms. Branded colour schemes, visualization formats, new quadrant labels on the same 2x2 matrix. BCG would present it with one set of labels and Deloitte would present it with another, but the axes measured the same dimensions and the strategic implications of each quadrant were identical.

Every fiscal year brought a refresh. The framework did not change; the label did. A “digital maturity assessment” became a “digital readiness index” became an “AI transformation diagnostic.” The internal analytical structure was the same grid with the same scoring criteria producing the same distribution of client scores. What changed was the slide template, the proprietary terminology, the colour palette.

We know this because we did it. Several contributors to this publication led methodology refresh initiatives at their firms. The honest version of what those initiatives accomplished: we spent six to eight months working with practice leadership and the marketing team to rename and re-visualize frameworks that had not been analytically updated in years. We produced new training materials so that consultants could present the renamed version fluently. We created competitive differentiation briefs explaining why our “Accelerated Value Framework” (or whatever the name was that year) was distinct from the competitor’s equivalent. Those briefs were difficult to write, because the honest answer was that they were not distinct in any way that mattered to the client.

The marketing worked, to a point. Clients who saw three pitches in a row and heard three different proprietary names for the same approach sometimes concluded that each firm was offering something unique. The partner who could explain their firm’s version most fluently won the mandate. What was being evaluated was presentation skill and relationship quality, not analytical differentiation. When the partner who won the pitch had been at a different firm two years earlier, pitching the same client with the same framework under a different name, the fiction was complete.

(Readers of this publication will recognize this from Post #7, where we described exactly that scenario from the buyer’s side of the table.)

Same Inputs, Same Outputs

Then AI arrived, and the fiction became unsustainable.

Between 2023 and 2025, every major consulting firm deployed an internal generative AI assistant:

• McKinsey built Lilli (launched mid-2023)

• Deloitte launched PairD (UK in October 2023, then broader rollout)

• Bain deployed Sage (2023, powered by GPT-4)

• EY released EYQ (October 2023, part of the EY.ai platform)

• PwC rolled out ChatPwC (reaching roughly 200,000 users across network firms)

• KPMG launched Workbench (June 2025, multi-agent platform)

• BCG built its own suite of tools, including Gene

Their timing and geographic reach varied, but by an April 2026 Consulting Huber practitioner comparison, the assessment was blunt: the internal AI assistant was no longer a differentiator. It was the cost of entry.

The firms connected these assistants to their internal knowledge bases, building retrieval layers over the repositories that two decades of generalization had filled with converging content. The same frameworks, the same maturity assessments, the same market-entry methodologies, the same phased approaches to organizational change, stored in different systems but containing interchangeable substance. When consultants at different firms asked their respective AI assistants to draft a competitive analysis or a transformation roadmap, the outputs converged because the inputs had already converged. No one has run a controlled side-by-side comparison yet. No one who has sat through three AI-assisted pitch decks in a single week needs one.

AI did not cause the commoditization. The codification strategy caused it. The knowledge-base overload accelerated it. The generalization response cemented it. The marketing-over-substance pivot obscured it. AI stripped away the part that was still hiding the convergence: the individual consultant’s judgment, experience, and rhetorical ability in presenting familiar material as if it were fresh. When the AI drafts the first version of the deck, the partner’s personal touch on slide 14 is no longer the product. The product is the framework, and the framework is the same one the client could get from any of the other firms pitching for the mandate, or from their own internal team using the same AI tools drawing on the same publicly available business school content that informed the frameworks in the first place.

Rational at Every Step

The painful irony of this trajectory is that it was documented in real time by people who understood it. Hansen, Nohria, and Tierney described the two strategies in 1999 and saw that individual services could mature into commodities. Academic research through the 2000s and 2010s tracked the knowledge management overload problem and called it what it was. Industry commentators began noting framework convergence years before AI made it undeniable. Maurizio’s May 2025 Consulting Intel essay on the non-fungibility of great consultants put it plainly:

The modern consultant is a fairly expensive human Lego brick: interchangeable, replaceable, fungible.

That diagnosis was available at every stage. The firms did not act on it because each step in the sequence was individually rational, and the cumulative consequence only became visible from a distance that no one inside the system had the incentive to take.

Codification was rational. It solved the dependency problem and enabled growth. Scaling the knowledge base was rational. Each additional framework represented captured institutional knowledge. Generalization was rational. It reduced maintenance costs and improved staffing flexibility. Branding was rational. If the substance converged, packaging was the remaining competitive lever. Deploying AI was rational. Every competitor was doing it, and the productivity gains were real.

Each decision made sense in isolation. Taken together over 25 years, they amount to an industry building the most sophisticated possible version of a commodity. The Big 4 and MBB spent two decades extracting knowledge from their best people, storing it in systems, generalizing it for reuse, branding it for differentiation, and then connecting AI to the result. The final product of that process is a set of frameworks that any firm can deliver, any AI can reproduce, and any sophisticated buyer can evaluate as interchangeable.

The Hansen paper offered a choice: codification or personalization. The firms that chose codification achieved everything the strategy promised. What they lost was the thing the strategy explicitly traded away: the dependence on individuals. It turns out that dependence was not a bug in the old model. It was the product. The specific partner who saw the problem differently, who brought twenty years of pattern recognition to a novel situation, who pushed back on the client’s framing because she had seen a similar framing fail at three other companies, who could not be extracted from herself and stored in a database. She was the value proposition. The firms spent twenty-five years engineering her out of the model, and they succeeded.

Whether the firms can reverse-engineer her back in is the question for the next decade. Personalization, the strategy Hansen described as the alternative to codification, means small teams of senior people working on difficult problems where the value is in their judgment, not in a reusable template, and where the firm’s economics reflect that. Alvarez & Marsal and AlixPartners have built versions of this model. Some boutique strategy firms never abandoned it. The Big 4 and MBB would need to restructure their partnership economics, their staffing models, and their growth expectations to pursue it. Given the alternative (selling a commodity at premium prices until the buyers finish noticing), they may not have a choice.

---

This post originated in the comment thread beneath Post #7, where reader Vince raised the codification trajectory as a factor in framework convergence. His observation was well-supported by the academic literature and deserved a full treatment.

At your firm, when was the last time a framework was genuinely improved rather than rebranded? Not renamed, not re-visualized, not refreshed for the new fiscal year. Actually improved, with new analytical substance that changed the diagnosis a client would receive. What triggered it?