The Consulting Pyramid Won’t Die. It’ll Change Shape.

Let’s start with the impolite part. The classic consulting pyramid - wide base of juniors, thick band of managers, tiny spike of partners - wasn’t built to find better answers. It was built to print better margins.

For decades it worked because clients bought effort dressed up as expertise. Armies in spreadsheets. Weeks of benchmarking. The theater of “we left no stone unturned.” Then AI walked in, kicked the props, and turned the stones in minutes. Meanwhile platforms slid upstream and started handing out “lite advisory” like diner napkins. Procurement is already asking who used AI, where, and how you checked the work. They’re buying outcomes and auditable process instead of bodies.

If you’re a boutique, you’re feeling a double squeeze: AI deflates task costs and platforms own the rails. The old pyramid creaks. The instinctive response is to ask, “So what shape comes next - flat, diamond, or monolith?” Better question: what must you own to stay valuable, and what shape makes that ownership easy?

This is a practical field guide for lean boutiques (5-20 people) that want to survive whatever comes next. We’ll scan near term (2025–2027), medium (2027–2030), and long (2030+). We’ll stress‑test the popular predictions, poke the fragile assumptions, and make a clean, opinionated call:

Thesis: Most boutiques will converge on one of two leverage shapes:

• The Obelisk: solo or near‑solo advisory - senior‑heavy, AI‑amplified, almost no base.
• The Diamond: lean firms with a strong middle (architects + tech enablers), a thin base of AI‑augmented apprentices, and a tiny apex of partners who orchestrate.

Forget the old pyramid and the fantasy flatland. Let me make my case.

‍

The Pyramid, Explained

Before we talk about collapse, remember why the pyramid existed at all.

In the consulting golden age (’60s–’70s), McKinsey, BCG, and Bain popularized “professional leverage.” The logic was brutally elegant: stack cheap labor under expensive brains. A few partners sell, more managers supervise, and a broad base of analysts grind.

Each layer fed the one above. Analysts dug up data, associates turned it into slides, managers shaped the story, partners sold it in the boardroom. Clients got the feeling of scale and rigor. Firms got margins. Juniors got an MBA’s worth of learning on someone else’s dime.

For decades, this model was unbeatable. It aligned perfectly with how big companies bought advice: in bulk, by the hour, through procurement teams who equated headcount with horsepower (not only in the private sector, but also in the public one - see the evidence here for the UK and here among the US federal government). The pyramid made the invisible tangible - “fifty consultants on-site” looked like value. It also solved the industry’s biggest problem: unpredictable project demand. When utilization dropped, you just rotated the pyramid - redeploy analysts, pause hiring, protect partner income. Elegant in its brutality.

The system wasn’t just an org chart. It created a career ladder, a predictable profit formula, and a rhythm for growing new partners. The only catch was that it depended on two fragile assumptions: that grunt work had to be done by humans, and that clients were willing to pay for it. Both are dissolving before our eyes.

And that’s where the cracks start to matter.

‍

Why the Pyramid Cracks (But Doesn’t Crumble)

The pyramid’s economics depended on two facts:

1. Clients couldn’t cheaply get analysis;
2. Training juniors on client time was socially acceptable.

Both are now fiction. Generative AI can digest 300 pages and spit a 60‑slide draft before your espresso cools. First‑pass diagnostics are near‑free. Clients know. Procurement knows. Your juniors know (they’re the ones prompting at midnight).

At the same time, execution platforms - Salesforce, AWS, HubSpot, Microsoft - keep moving upstream. They’re offering not only features but implementation support: onboarding playbooks, “success plans,” half‑respectable roadmaps. Is that orchestrated transformation? No. Is it enough to cannibalize low‑complexity work? Absolutely.

So what breaks? The base. The work that justified platoons of juniors is getting automated or absorbed by “success” teams. What doesn’t? Judgment, trust, orchestration - the messy human arts that make change stick. That’s the space boutiques must own.

But owning orchestration isn’t a slogan. It has operating consequences. It changes who you hire, how you price, how you guarantee quality at speed. It changes your leverage shape.

‍

The Three Shapes Everyone’s Arguing About

Now that the old pyramid is wobbling, everyone has a theory about what comes next. Some argue for a flatter, senior‑heavy future. Others see a swelling middle of expert generalists as AI eats the base. And a third camp says people won’t be the leverage at all - platforms will. Let’s unpack the three contenders: the Obelisk, the Diamond, and the Monolith.

‍

1. The Flattened “Obelisk”

The obelisk is consulting on intermittent fasting: cut calories, keep strength. Tall, lean, almost no junior base, just enough mid‑layer to coordinate. The idea surfaced in HBR think pieces and hushed comments from MBB partners. As one put it, “A senior with an AI co‑pilot is the new case team.”

The bet: AI replaces 70-80% of junior analysis, and clients accept compact, senior‑heavy teams if quality holds. A single partner juggles two or three pods of AI‑augmented experts. Pricing drifts toward retainers and milestones because utilization gets spiky when one person is “the product.”

Why it might work:

• AI does analyst work - research, synthesis, first‑pass decks.
• Clients prefer to pay for judgment, not body counts.
• Seniors can run more in parallel without obvious quality loss.

Why it might crack:

• Overhead doesn’t disappear. Coordination still takes time.
• Without juniors, there’s no apprenticeship pipeline.
• It relies on steady, high‑fee work. One slow quarter and your economics implode.

The obelisk is elegant in theory and brutal in practice. It may work for ultra‑specialists and C‑suite whisperers, but for most firms, it’s too brittle to scale.

‍

2. The “Diamond”

The diamond is the balanced response - and probably the most credible. PwC, EY, and mid‑sized firms say they’re “moving from pyramid to diamond.” Translation: the junior base shrinks as AI eats tasks, while the middle - specialists and architects - swells.

In the diamond model, AI handles the repetitive, humans handle the relational. A small layer of tool-literate facilitators runs the tech stack, a thick middle of domain experts designs and orchestrates change, and a compact apex of partners sells, steers, and owns outcomes.

Why this model is gaining traction:

• It follows the value: orchestration, integration, adoption.
• t preserves apprenticeship by proximity (facilitators learn from architects).
• It scales outcomes without re-inflating headcount.
• It matches how clients buy now: small expert teams, measurable impact, fast cadence.

Assumptions under the hood:

• AI saves 30–50% consistently and is routinized.
• Firms can upskill or hire a new “expert middle” fast enough.
• Clients will trust leaner teams if quality control is transparent and auditable.

Skeptics warn it can re‑bloat - partners hoard control, hire too many mid‑levels, and the base creeps back. Still, it’s the consensus favorite. The first viable post‑pyramid equilibrium.

‍

3. The “Monolith” (Platform Hybrid)

If the obelisk is consulting on a diet and the diamond is consulting rebalanced, the monolith is consulting reinvented. Here, leverage comes from products, data, and code, not people.

Tech‑adjacent players - BCG X, Palantir, and “consulting platforms” - push this. The future firm looks less like a partnership and more like a product company with advisory wrappers. The human pyramid becomes a slim control layer atop reusable IP, diagnostics, and agents.

Imagine a boutique that builds proprietary benchmarks, diagnostic bots, or workflow tools - then charges clients subscriptions for ongoing access while layering high-value advisory on top. Margins depend on adoption, not utilization.

The bet:

• You can codify knowledge faster than it walks out the door.
• Clients will pay for access (data, frameworks, or tools), not just for people.
• You can fund product development without starving the firm.

Risks and reality:

• Building platforms is expensive and culturally alien to services firms.
• Software economics and service culture mix like oil and water.
• Productizing expertise can expose how generic it was all along.

Still, for the few who pull it off: high switching costs, recurring revenue, and SaaS‑like multiples. For the rest: unfinished code and unpaid invoices. For boutiques, the pragmatic path is “build a spine, not a product” - a small internal OS that compounds craft without full product risk.

‍

Each of these shapes tells a different story about what consulting will become. The obelisk bets on efficiency, the diamond bets on orchestration, and the monolith bets on assets.

All three will exist. But read my last piece for the Future of Consulting, and you’ll understand why I’m betting big in the idea of boutique consultancies becoming orchestrators. If you’re a change‑oriented boutique (digital/AI adoption, cloud, cyber, ESG), the Diamond is the only shape that scales outcomes without re‑inflating bureaucracy.

The obelisk caps capacity. The full platform play is capital-intensive. The diamond is the pragmatic middle: enough expert middle to run multiple pods, thin base because AI eats the grunt work, and a light product spine so your craft compounds.

Think of it as: small base, strong middle, tiny apex + internal “OS.”

‍

Zooming In: What the Diamond Actually Looks Like

In a 12‑person boutique, the future doesn’t look like a mini‑McKinsey. It looks like a pod network riding a platform spine.

• 1 or 2 partner‑orchestrators frame the problem, own the room, and carry the scope and outcomes. They drop into key ceremonies - charter sign‑off, quarterly reviews, scope changes.
• Beneath them, engagement architects turn ambition into action: chunk outcomes, run governance, mentor. And tech architects build the leverage: retrieval, agent workflows, eval harnesses to keep AI reliable.
• Then 1 or 2 AI facilitators handle synthesis, summaries, and first‑pass decks. Instead of old‑school analysts, they are co‑pilots for the firm’s AI stack, turning chaos into structure. These people probably can be spotted by their strong writing, ability to sit steady in rooms, and systems brain. Paid like top‑tier analysts, with faster progression.
• Around them: a fractional specialist bench (cloud, cyber, data governance, regulation) that spins up as needed.

Pods run lean - three to five people - yet move at the pace a team of ten used to. Automation eats the grind, so one partner can steer two or three charters without quality slippage. Small teams, crisp accountability, and a shared platform that compounds the firm’s intelligence.

‍

Looking for Pushback

This all sounds interesting and clever, but is it grounded in reality? The only way to know was stress testing this thesis with real consultancy founders. So that’s what I did, through a series of informal conversations.

First, I stated the thesis:

Most boutiques converge on one of two shapes: An obelisk (solo or near‑solo, senior‑heavy, AI‑amplified, minimal base) or a diamond (strong middle (architects + tech), thin AI‑augmented base, tiny apex that orchestrates).

Then I listened. The three most pragmatic counter‑arguments:

1. “This structure is way too complex.”
2. “Small firms will live inside ecosystems.”
3. “Talent economics make the diamond unsustainable.”

I’ll briefly share what they’re all about, and the counter-counter-arguments I use to defend the thesis further.

‍

“The market won’t reward structure this advanced.”

Some founders argued that sub-10-person boutiques don’t face enough complexity to justify architects, AI roles, or platform spend.

“We live deal-to-deal and ship bespoke. A diamond feels like over‑engineering.” Fair concern. Why build a middle when you can subcontract or automate on demand?

First: short‑term reality ≠ long‑term viability. Ad-hoc benches work for a while. As clients adopt their own AI workflows and expect integrated delivery, “solo brilliance” loses to repeatable orchestration. The diamond matches how buyers shift - outcomes, not hours.

Second: the middle layer can emerge without hiring. AI systems and reusable workflows act as “virtual architects.” Many diamonds will be virtual: a partner, a part‑time engagement lead, a facilitator, plus a template library that does quiet work. I’ve seen 1 partner + 1 EA + 1 facilitator run five large engagements, with ~40% time shaved off prep/synthesis.

Last but not least, we can see proof from adjacent fields. Design studios, product shops, and fractional CXOs have already evolved in this direction. They deliver via “pods” that mix human orchestration and system leverage, not via bespoke craftsmanship. Consulting will follow as price pressure and expectations rise.

‍

“Small boutiques will just cluster around ecosystems”

Another counter-argument I’ve heard is that, instead of building a middle, boutiques will embed in platform ecosystems - AWS, Salesforce, HubSpot, Notion, alliances, marketplaces.

Ecosystems offer structure, tech, and pipeline, letting boutiques stay flat. Tempting… until dependency bites.

Dependency risk is the new existential risk. Ecosystem alignment may bring early leads, but it also caps margins and differentiation. Platform partners eventually commoditize their experts, as seen with Salesforce Gold Partners or HubSpot Agencies. A diamond - owning orchestration - lets you escape that gravity well.

Orchestration is the ecosystem’s missing layer. As platforms bundle more “lite advisory,” clients will need neutral orchestrators who coordinate across them. Boutique diamonds fill that gap. You become a trusted integrator across multiple vendor environments.

‍

“Talent economics make the diamond unsustainable.”

This last counter-argument is the one I’ve heard the most.

Finding and retaining “architects” (people who blend strategy, tech, and orchestration) is expensive. AI fluency plus consulting judgment is rare. Most small firms can’t afford mid-level talent; they either hire juniors and train them, or rely entirely on the founder. Without scale, the diamond collapses.

There are two key questions for boutique consultancies here: First, “can we afford to pay for this kind of talent?” And second, “what happens with the apprenticeship model? How do we train managers with less juniors?”

From an economics perspective, it’s worth remembering that the diamond is more profitable, not more expensive. A tight 10‑person diamond can out‑margin a 20‑person pyramid. Mid‑level talent costs more, but multiplies throughput. The unit is leverage per person, not headcount.

(Back-of-napkin:

• Diamond pod (12‑week charter): 4 people fractionally (partner 20%, EA 80%, TA 40%, facilitator 80%) ≈ 1,200 billable hours. At $275/hr, revenue ≈ $330k. Loaded delivery ≈ $145/hr → ≈ $174k. Gross margin ≈ $156k (47%). Partner at 20% can oversee two pods.
• Pyramid (same scope, 8–10 people): blended rate compresses to ~$230/hr; coordination waste eats hours; loaded cost per billed hour climbs to ~$170/hr. Revenue ≈ $276k; cost ≈ $204k; gross margin ≈ $72k (26%).

Same scope, similar billed hours - but the diamond’s higher rate and lower coordination loss yield ~21 pts more margin and better partner yield.)

Even if I’ve completely botched the math, several of the founders I spoke to mentioned that AI flattens the talent premium. The cost gap between junior and mid-level talent is narrowing because AI handles much of the middle-skill work.You need fast learners more than unicorn polymaths. Architects can emerge via apprenticeship + platform memory. Which brings us to the second question.

Shrinking the junior layer doesn’t mean you stop developing people, it means you compress the learning curve. In the old pyramid, analysts spent years grinding through Excel and PowerPoint before they were trusted to think. In the diamond, the base is smaller but faster.

Facilitators arrive AI‑fluent. First six months: master the firm’s spine - precedents, prompts for synthesis, how to question and evaluate machine output. By month twelve, they’re running workstreams under an architect, not formatting slides at midnight. Target: prompt‑executor to problem‑owner in 18 months, not five years.

Every boutique that survives this transition will build a platform spine that encodes its craft - templates, exemplar decks, playbooks, prompt libraries, decision memos, “before/after” cases that show where judgment moved the needle. Tacit know‑how leaves heads and lives in systems. The platform is a living textbook: open a new engagement and you’re surrounded by precedents and workflows.

This all sounds promising - but it mostly accelerates hard skills. The room is won by soft skills.

The biggest challenge for the diamond-shaped firm of the future, after all of the research and conversations I’ve had, come down to one question: How do you accelerate the acquisition of soft skills?

‍

Soft Skills Are The Bottleneck

The World Economic Forum’s Future of Jobs work points to a simple shift: by 2030, tech skills fade faster; soft skills - creative thinking, leadership, empathy - matter more. AI reshapes +40% of core skills. Translation: the half‑life of hard skills is shrinking; the compounding is in soft ones.

Soft skills are hard to learn, and they’re what’s left for humans as software eats tasks. Your moat lives there. We ignored them for years. We can’t now.

As AI accelerates, a lot of yesterday’s techniques won’t work tomorrow. Meanwhile, you need managers with fewer juniors. That’s a soft‑skills problem.

Traditionally, juniors built soft skills via structured training, on‑the‑job reps, and personal development. Shadow a senior in a pitch, get feedback on clarity and presence, attend a workshop on stakeholder alignment, rehearse in mock projects, reflect with peers. It takes years. Accelerating it feels impossible.

If I had to sketch a soft‑skills acceleration playbook for a next‑gen boutique, I’d probably explore stuff like:

1. Rehearse the room: monthly recorded role‑plays of tense stakeholder meetings with rubric‑based feedback (clarity, empathy, negotiation).
2. Write‑to‑own: decision memos and POVs that must land with execs; track first‑pass accept rate.
3. Feedback density: two short loops per week beat one long one per month.

(Please consult with experts before implementing these, I know close to nothing about upskilling methods.)

If the thesis I’m presenting here get proved wrong in a decade, it will be because we couldn’t accelerate soft skills fast enough.

‍

Can You Re-Imagine Your Boutique?  

Now let’s imagine the thesis does materialize. What does a firm look like?

I reached out to clients and acquaintances who run boutique consulting firms to ask for their best guess on how their businesses would look like if my thesis is correct. Soloists sketched obelisks. Boutiques sketched diamonds. Here are a few snapshots.

‍

Soloists and Their Obelisks

Let’s look at where most founders actually start: as trusted guides running obelisks. These are lean, senior-heavy practices where the founder is both product and brand, assisted by AI tools and perhaps one or two operational supports. High-trust, high-impact, but hard to scale.

1. The Executive Coach. Elena runs North & West Advisory, a solo coaching and leadership development practice for tech executives. Her obelisk is as narrow as they come - just her at the top, supported by an AI co-pilot that summarizes transcripts, drafts follow-up reflections, and tracks client progress against leadership metrics. She personally handles all client sessions, onboarding, and business development. Her leverage is intimacy and depth. Her only “base” is a part-time operations assistant and a digital toolchain that automates prep and note synthesis.

2. The Martech Advisor. Arun leads SignalBridge, a micro-consultancy helping B2B SaaS firms integrate marketing automation platforms. His obelisk includes himself, a part-time implementation specialist, and a library of custom GPT workflows. He personally scopes and orchestrates every engagement, from vendor selection to data migration. Clients hire him for judgment and technical clarity, not a team. AI handles research, documentation, and audit summaries, letting him operate three clients concurrently without staff.

3. The Workforce Facilitator. Maya founded WorkKind Studio, a practice focused on culture transformation and employee experience. She runs strategy sessions herself and uses a combination of AI-assisted survey analysis and visualization tools to generate diagnostics. A single assistant helps manage logistics and participant communications. Her brand relies on emotional intelligence and trust - AI stays in the background, accelerating reflection summaries and workshop design. Like most obelisks, her firm’s value rests entirely on her reputation and ability to orchestrate complex human conversations.

All three operate sustainably at small scale, but their limitations are clear: everything depends on the founder. The next step - if they want to build a firm that outlasts them - is to grow a diamond.

‍

What a 12-Person Boutique Will Actually Looks Like?

Three diamond‑shaped boutiques, in practice:

1. The Digital Transformation Boutique: A 12-person consultancy helping mid-sized manufacturers modernize their operations.

Two partners-orchestrators manage client relationships and define 90-day modernization charters. The middle group consists of six architects: three engagement architects who run transformation pods (governance design, process mapping, and adoption metrics), and three tech architects who build and maintain the firm’s agentic platform (data retrieval pipelines, diagnostic bots, and evaluation systems). Two AI facilitators handle meeting notes, synthesize documentation, and assemble draft playbooks for client review. A rotating bench of cybersecurity and ERP experts is called in when needed.

Each engagement runs as a self-contained pod - partner, one engagement architect, one tech architect, and one facilitator - supported by the shared platform spine. The model allows the firm to handle multiple transformations simultaneously without losing coherence or quality.

2. The Marketing Advisory Firm: A 10-person brand and growth consultancy for B2B technology companies.

A single managing partner sets direction and curates client relationships. The architect layer combines creative and technical expertise: three engagement architects (brand strategy, content orchestration, and analytics) partner with two tech architects who maintain the agency’s AI-powered content generation and insight engine. One AI facilitator handles data cleaning, campaign synthesis, and post-campaign reporting.

The engagement architects translate business strategy into creative narratives and message architectures. the tech architects automate testing, content optimization, and audience analytics. Together, they turn branding into a measurable, living process rather than a deliverable.

3. The ESG and Sustainability Advisory: A 15-person boutique guiding industrial firms through ESG strategy and reporting.

Two partners-orchestrators handle executive alignment and stakeholder management. Their architect layer - seven specialists in ESG frameworks, carbon accounting, supply chain transparency, and regulatory compliance - drives project design and execution. Among them, four function as engagement architects focused on client governance and implementation, while three act as tech architects maintaining the firm’s ESG intelligence platform. Two AI facilitators manage synthesis: drafting sections of sustainability reports, summarizing interviews, and producing compliance checklists.

The firm’s diamond runs on integration. Engagement architects ensure that ESG strategy connects with operational change. Tech architects translate complex datasets into dashboards and scenario models. Output isn’t “a report,” but orchestrated change measured by audit readiness and stakeholder trust.

Across all three cases, the pattern is unmistakable: partners orchestrate outcomes, the architect layer turns judgment into design and execution, and a small base of AI-augmented facilitators amplifies everyone else. The pyramid may have shrunk, but its center of gravity has moved upward - toward the middle, where technology and expertise fuse to deliver speed, quality, and repeatability.

‍

A Practical Path To Build A Diamond Firm

“If I already have a team of 3, 5, or 8 people, how do I evolve toward the diamond model - without breaking what’s already working?”

The answer isn’t to restructure overnight. It’s to layer new capabilities onto your existing structure, one step at a time, so that your operating model naturally tilts toward the diamond.

Also, let’s not forget you can’t do that without a mindset change. The founder’s goal shifts from doing great work to building a system that does great work repeatedly.

So, if you already have a 3-, 5-, or 8-person team, you’re halfway there. You don’t need to tear anything down. You just need to:

1. Name your future model (diamond).
2. Start building its middle layer (architects).
3. Encode your collective craft into systems.

Do that for 12–24 months, and your firm quietly transitions from a founder-led boutique to a compounding organism - small, adaptive, and diamond-shaped by design.

‍

Step 1: Map Your Current Shape

Before changing anything, founders need to see their current leverage model clearly. Most boutiques under 10 people are an hourglass: a heavy top (founders doing strategy + sales), a small operational middle (project managers or specialists), and a thin base (analysts, freelancers, or ops).

Practical Actions:

• Run a “pyramid audit”: List everyone’s 3 core activities and approximate time share (sales, delivery, admin, client success).
• Identify which tasks are codifiable (can be templated, delegated, or automated).
• Clarify who owns orchestration now. Who ensures alignment between clients, teams, and outputs?
• Tag one person as a potential “architect seed”: someone who blends delivery, systems thinking, and leadership potential.

The goal is to see the actual shape before deciding what to reshape. Most consultancies can do that in 4 to 6 weeks, even if this is not a huge priority.

‍

Step 2: Strengthen the Middle

The hallmark of the diamond is the middle layer, your engagement + tech architects. Boutiques don’t need to hire this layer; they can develop it internally.

Practical Actions:

• Identify 1 or 2 team members who already act like informal orchestrators (project managers, lead consultants, or tech-savvy delivery leads).
• Give them clearer authority: “You own the quality, governance, and rhythm of this engagement.”
• Start documenting delivery rhythms: weekly rituals, decision gates, client handoff templates.
• Pair each of these “proto-architects” with the founder for shadowing and feedback loops.
• Introduce light-touch performance metrics: project NPS, on-time deliverables, and margin by engagement. Add one throughput metric (time from brief to decision) and one quality metric (rework rate) to tie orchestration to economics.

The founder’s role evolves from chief doer to chief designer of delivery systems. This builds the spine of your future diamond without any layoffs or new hires. It takes time for people to adjust, but if everyone’s on board you can pull it off in 6 months.

‍

Step 3: Build the Platform Spine

Once you have proto-architects running smoother projects, you start encoding what works. This is how your firm’s intelligence compounds.

Practical Actions:

• Consolidate templates, frameworks, and checklists from the last 3–5 successful engagements.
• Create a single shared “Delivery OS” (Notion, ClickUp, Airtable - doesn’t matter) that holds:
  • Proposal → onboarding → delivery → review workflows
  • Prompt libraries for internal AI use
  • Example decks, playbooks, decision memos
• Assign one internal “AI facilitator” (it can be part-time) to experiment with summarization, synthesis, and first-draft creation for recurring deliverables.
• Begin automating repeatable tasks: reporting, meeting synthesis, status updates.

This platform spine turns the firm from a collection of freelancers into a learning organism - every project leaves behind codified IP. This is ongoing work, but you can get the ball rolling in one or two quarters of focused work.

‍

Keep the Ball Rolling

If you complete those 3 steps, you’ll probably be ahead of 95% of the boutiques in your space to build a leverage model that’s profitable and sustainable in the next decade.

There’s no manual or best practices out there on how to keep moving forward, but if I had to guess (and that’s what I’ll be doing with the Boutique Consulting Club), I would then:

• Accelerate AI-augmented apprentices to build the base of the diamond. I’ll probably retrain or hire someone as “AI facilitator” for synthesis, document drafting, and IP management. And then create a structured development path for a Facilitator to become an Architect in 12-18 months.
• Codify and communicate the new model. Once you have stronger architects, a working platform spine, and fast apprentices, your firm is already operating as a diamond - it just doesn’t have the language yet. I’d make the model explicit to my team (“Here’s how we run - Partners, Architects, Facilitators.”) and communicate it to clients as a mark of maturity: a consistent delivery rhythm, clear accountability, and AI-enabled efficiency.
• Future-proof through modularity when (and if) the firm hits 10–-12 people, to resist the temptation to grow linearly. Instead, I’d grow through pods - self-contained units that replicate your delivery rhythm. You can cap each pod at 3-5 people (1 partner, 1-2 architects, 1 facilitator, fractional specialists). The platform spine is shared across pods, but I’d rotate architects between pods quarterly to cross-pollinate expertise. Set a quarterly “show-your-work” where pods present decisions, outcomes, and lessons into the spine.

Is this a perfect plan? I’m 100% sure it is not. But if you buy into my thesis, that’s the best practical path I can see to making the diamond-shaped firm work.

‍

The Timeline

After sharing the draft of this essay among friends, they pointed out to me I was not explicit with the exact timeline I expect my whole thesis to come true. Here’s my best shot:

• Short term (2025–2027): augmentation inside familiar seams. The pyramid wobbles but stands. Juniors use AI to do the same work faster. Early boutiques adopt pods + platform. Procurement starts asking how you made the sausage. You move to charters, collars, and retainers.
• Medium term (2027–2030): restructuring. Entry-level roles compress, middle expertise thickens, diamonds become visible in org charts. Authority layers matter (benchmarks, councils). Outcome kickers show up where levers are clear. The boutiques who codified their craft into a platform spine pull away.
• Long term (2030+): hybridization. The best boutiques look like expert networks wrapped around instruments: pods that orchestrate change, authority platforms that compound, and selective risk exposure where they actually steer the ship. The “pyramid” survives only as a memory and a warning: don’t tie your economics to tasks the machine loves.

‍

A Final Word Before the Wave Hits

We can argue shapes all day, but shape is a side effect. Ownership comes first.

• If you don’t own change, you’re a vendor on a Gantt chart.
• If you don’t build lens (information or relationships), you’re price-comparable and forgettable.
• If you don’t take risk where you control levers, you’ll be undercut by those who do.
• And unless you control the rails (rare for boutiques), stop fantasizing about being a platform company. Build a spine, not a monolith. And remember: owning the room is a soft skill. Install judgment, cadence, and courage with deliberate practice.

The future of leverage in consulting is simple to say and hard to do: Fewer bodies, more orchestration. Fewer hours, more outcomes. Fewer secrets, more systems.

Winners won’t have the prettiest slides. They’ll install judgment, cadence, and courage in the rooms where decisions get made - and run on an OS that learns.

Will there still be pyramids in 10 years time? In museums. Next to the fax machine.