Skip to content
upshift
The Three Shifts

Your offerings live in your senior people’s heads. AI rewards the firms that move them out.

Custom service, named offering, repeatable methodology, licensed IP. Four stages of making firm expertise sellable without one person at the center of every engagement. Most firms are stuck on the bottom rungs.

Most professional services depend on a specific person being in the room. The work is judgment, the judgment lives in their head, and the offering walks out when they do. That’s how it has worked for decades. AI is making it expensive.

Codified IP scales at near-zero marginal cost. Methodology that exists outside one person’s head can be trained and replicated. AI does the codification work that used to require a dedicated team. It captures the patterns senior people follow without thinking about it. It surfaces the rules they apply on autopilot. Firms that move out of that dependence keep what they build. Firms that don’t will watch their best ideas walk out with senior staff.

What you can train a junior to deliver is what you can sell. Everything else depends on one person.

The four stages compound. You can’t codify what you haven’t named, and you can’t productize what only your senior people can deliver.

What’s already in the field.

Future Ready Accountant Report: 93% of accounting firms claim to offer advisory. Almost none have systematized it well enough that the work survives the senior person being unavailable. That’s the climb most firms are still at the bottom of.

Deloitte moved some advisory clients to subscription billing on codified methodology. The same expertise that used to ship as project hours became a platform with monthly recurring revenue. The codified offering is the asset; the subscription is the revenue model that matches. Codification is what turns AI from a productivity tool into a margin lever.

The four stages, and what AI does at each.

Which rung does your highest-revenue offering sit on today?

Scales without you

Productized

The licensed IP

What AI does

AI compounds at this stage. The methodology gets smarter with every engagement.

What’s working

Stop selling time. Start selling the system.

What AI does

AI works as a margin lever once the methodology is documented. Capacity expands.

What’s working

Build the method. Train the team. Keep the judgment.

What AI does

Defined offerings hit a ceiling fast: the senior person is still the bottleneck.

What’s working

Naming is not delegating.

What AI does

AI can’t capture what isn’t documented. The offering walks out when the partner does.

What’s working

Name the work before you can sell it.

Lives in heads

Ad-hoc

The custom service

Name the work before you can sell it.

The work runs through one specific person. Their judgment, their relationships, their pattern recognition. What they do is real and valuable, but it walks out when they do.

What AI does

AI doesn’t fix ad-hoc work. It can’t capture what isn’t documented, can’t train what isn’t taught, can’t replicate what only one person knows. Firms at this stage are running on one person’s calendar. As that person approaches retirement or simply stops accepting new clients, the firm’s most valuable work has no successor and no buyer.

What’s working

Pick the most valuable thing your senior people do and write down the decisions behind it, not the deliverable it produces. What do they ask in the first five minutes? What signals trigger which recommendation? Name the deliverable and you get a category like “AI-augmented advisory,” which is marketing copy no client can buy. Name the judgment and you have something a junior person can be trained against. That’s the prerequisite for every rung above this one.

Defined

The named offering

Naming is not delegating.

The offering has a name, scope, deliverable, and price. Clients can buy it; marketing can write about it. But it still requires a senior person to deliver. Defined doesn’t mean someone else can deliver it.

What AI does

Defined offerings under AI hit a ceiling fast. The senior person is still the bottleneck. AI speeds up parts of the work (research, drafting, first-pass analysis), but the judgment that makes the offering valuable still lives in one head. The firm is selling a productized version of one person’s time, which is better than ad-hoc but still not scalable. Margins improve modestly; capacity does not.

What’s working

Document what the offering actually requires: inputs, decisions, quality bar. The discipline surfaces the parts that aren’t senior-judgment-dependent, usually most of the work. Expect the partner who owns it to say the work is too bespoke to write down. That objection rarely survives one real engagement studied closely. The repeatable pattern shows up, the genuine judgment is the smaller part, and naming it is what frees the senior person from the rest.

Codified

The repeatable methodology

Build the method. Train the team. Keep the judgment.

The methodology is documented well enough that the work no longer depends on the senior person who developed it. Decision trees, frameworks, checkbox criteria, AI-augmented workflows. Senior people review the high-judgment calls; junior staff and AI handle the structured parts. The offering survives the senior person being unavailable.

What AI does

AI works as it should at this stage. Workflows automate the structured parts; humans apply judgment to the structured outputs. AI-assisted research, drafting, and analysis compress the cost of delivery while keeping the value to the client intact. Margins improve materially. Capacity expands without proportional headcount.

What’s working

Extract the methodology from the senior people, then build it into the workflow your team runs every day. The documented way this fails is the binder nobody opens: a methodology written as a separate project, parked in a shared drive, disconnected from the actual work. It has to live where the work happens. Junior staff and AI run the structured parts; senior people review the judgment calls and check the output against what they know is right. Codification is what turns AI from a productivity tool into a margin lever.

Productized

The licensed IP

Stop selling time. Start selling the system.

The offering is a product, not a service. It scales without proportional headcount. Recurring revenue, replicable across clients without re-staffing. The firm has built something that exists independently of any one person.

What AI does

AI compounds at this stage. Each client engagement makes the underlying methodology smarter. Patterns surface and decision rules refine. The offering improves on its own. The firm’s competitive moat is the IP, not specific people. New competitors have to build the methodology from scratch; the firm’s installed base of refined IP is the asset. Margins keep improving as the marginal cost of delivery approaches zero.

What’s working

Before you build the platform, get a paid yes on the brief. The way this rung kills firms is the reverse: build the system first, then go looking for buyers. Write the offering down tightly enough that a named client can say yes or no to it, and take it to three of them. Encouraging conversations aren’t demand; a signed pilot is. Build the platform, subscription, or licensed methodology only for the offering that earns a real commitment. The firms that validate first set the price and the terms.

What ties them together.

The four stages aren’t equally exposed to AI. The firms whose offerings live in specific people’s heads are the most fragile. The firms who climb the ladder turn their expertise into something that exists outside any one person and stays at the firm even when senior staff don’t.

Climb the ladder. Each rung makes the next possible. Skipping doesn’t work: you can’t codify what you haven’t named, can’t productize what you can’t train someone else to deliver.

What stalls firms here is comp, not capability. The time a partner spends writing down how they work is time they don’t bill, and in a firm that rewards individual billings, that reads as pure cost. The method gets built where the firm decides that time is the investment it is. That’s the courage the name points at.

This ladder is the third axis of business model change AI is forcing. The Pricing Ladder covers what you charge, The Delivery Ladder covers how you show up, and the Productization Ladder covers what you actually sell. Move on one without the others and the gap shows up six months later.

The Productization Ladder runs from bespoke to scalable, one of the Three Shifts. The Workshop specifies 2–3 of those offerings, priced and defined tightly enough to put in front of named clients for a real yes or no. Your team or an implementation partner builds the ones that earn it. For how it lands in your sector, see the industries page.

Questions

A productization framework maps the stages a professional services firm moves through to make expertise sellable without depending on one specific person. The Productization Ladder names four: ad-hoc, defined, codified, and productized. Each stage changes what AI does to delivery costs and what it takes to scale.

Productizing means turning expertise that lives in specific people’s heads into something that exists outside them. A named offering, documented methodology, structured workflow, and a delivery pattern that doesn’t require any specific person to deliver. The destination is IP that scales: replicable across clients, sellable without re-staffing, and durable through senior turnover. Most firms don’t get there in one move. Productization is the top of a four-stage climb (ad-hoc, defined, codified, productized) where each stage builds on the previous.

AI does most of the codification work that used to require a dedicated team. Decision trees, structured outputs, captured patterns from past engagements: all of it gets cheaper and faster with AI in the workflow. The firms that climb the ladder use AI to compress the cost of delivery without compressing the value to the client. The firms that don’t climb are running AI as a productivity tool on top of an ad-hoc service, which improves margins modestly and changes nothing structurally.

A defined offering has a name, scope, and price; a client can buy it. A codified offering has a documented methodology anyone trained on it can deliver. Defined is the marketing layer; codified is the operations layer. Most firms stop at defined and wonder why scaling doesn’t work. The senior person is still the bottleneck because the methodology lives in their head, even if the offering has a name on the website.

Pick the most valuable thing your senior people do and start writing down the decision pattern behind it. Not the work itself: the questions they ask, the signals they look for, the recommendations that follow. Most firms skip this and try to define an offering before they’ve named the underlying expertise. That produces marketing copy, not a sellable service. Spend the time on naming first; everything downstream gets easier.

Document what the offering actually requires. Inputs, decisions, deliverable structure, quality bar. The discipline of writing it down surfaces the parts that aren’t actually senior-judgment-dependent, usually most of the work. Once that’s on paper, junior staff and AI can handle the structured parts; senior people review the high-judgment calls. That’s the climb from defined to codified.

Productization is when the codified methodology becomes the asset and the firm’s revenue model changes to match. Subscription billing, licensed methodology, platform delivery. The marginal cost of serving the next client approaches zero because the methodology already exists. Deloitte moved some advisory clients to subscription billing on codified methodology. Big Four firms are productizing due diligence and audit prep. The pattern is the same across professional services: codified offerings shipped as platforms instead of project hours.

The IP. The accumulated methodology, the refined decision rules, the body of past engagements that makes the next one better. A competitor trying to displace you has to build the methodology from scratch and earn the trust that comes from a track record they don’t have. Keep refining: more clients, more patterns, faster iteration. Productized is the floor of a different kind of competition, not the finish line.

Ready to figure out what you sell next?

Thirty minutes with Shawn Yeager. No pitch. Talk through what the Workshop looks like for your team.

Book a conversation

Or see where you stand