---
title: "AI Killed the Billable Hour"
date: "2026-03-30"
description: "Keynote presentation for professional services firm leaders. What AI means for your business model and what to do about it."
marp: true
theme: upshift
paginate: true
html: true
lang: en-US
---

<!-- _class: cover -->
<!-- _paginate: false -->

<img class="logo" src="logo.svg" alt="">

# AI Killed the Billable Hour

What it means for your firm — and what to do about it

Shawn Yeager · Upshift · March 2026

<!--
SPEAKER: Open with a question — pause and let it land.

"How many of you have bought AI tools in the last 12 months?" Wait for hands. "How many of you can point to a dollar of new revenue those tools created?"

That gap — between buying and earning — is what this is about. 90 minutes. Let's get into it.
-->

---

<!-- _class: agenda -->

## The agenda

1. The shift that already happened
2. Why efficiency isn't enough
3. What the firms that survive look like
4. How to start on Monday

<!--
SPEAKER: "Four acts. Each one builds on the last. I'm going to try to leave you with something you can actually do — not just think about."
-->

---

<!-- _class: light -->

## Shawn Yeager · Founder, Upshift

- 30+ years and $300M+ in revenue across five technology cycles — browsers, mobile, SaaS, Bitcoin, now AI
- CEO, operator, revenue leader. I've made the decisions you're facing now.
- I don't implement AI. I help firms figure out what to sell because of it.

<!--
SPEAKER: "Browsers. Mobile. SaaS. Bitcoin. Now AI. I've been through four of these before, and each followed the same arc."

"The new capability gets bolted onto the old model. Companies used the internet for digital brochures. They used mobile for smaller screens. They used cloud for cheaper servers. This phase feels productive. It is also a trap."

"Then a handful of companies ask a different question — not 'how do we use this?' but 'what does this make possible that wasn't possible before?' They build around the answer and move into open space while everyone else is still refining what they've already got."

"I've spent 30 years and built $300 million in revenue on both sides of that. The technology changes every cycle. The commercial pattern doesn't. That's why I'm here."

~90 seconds.
-->

---

<!-- _class: statement -->

## The shift

The cost of expertise delivery has **collapsed to near zero.**

<!--
SPEAKER: Three-beat pause after the slide appears.

"That sentence is worth sitting with for a moment."

"I don't mean AI made expertise cheaper. I mean the cost of delivering expert-level output — a contract, a financial analysis, a sourcing shortlist — has dropped so far so fast that it has fundamentally changed what clients will pay for it."

"That's not a future threat. It happened. And most firms in this room are still pricing like it didn't."
-->

---

<!-- _class: statement -->

Every professional services firm is now in the **business model business.**

<!--
SPEAKER: "What I mean by that: for 30 years, the business model of professional services was simple. Hire smart people. Bill their time. Grow by adding people or raising rates."

"That model worked because expertise was scarce and hard to access. AI broke that. Not just for your competitors — for your clients. They have access to AI tools too."

"So now the question isn't 'how do we do our work better?' It's 'what do we actually sell in a world where AI can do most of what we used to charge for?'"

"That's a business model question. Most firms aren't asking it yet."
-->

---

<!-- _class: data -->

## Your clients already moved

- 60%

  of senior marketing leaders already spend less on agencies because of AI

- 87%

  of companies use AI in at least one part of recruiting

- 50%

  of Harvey AI's business is now corporate in-house legal — up from 10% in a year

<!--
SPEAKER: "These numbers are clients pulling work back in-house. Not threatening to. Doing it."

"60% of marketing leaders are already spending less on agencies. Not considering it — already spending less. And that's a sector that was ahead of where law and accounting are today."

"Harvey AI is particularly telling. They started as a law firm tool. A year later, half their business is corporate legal departments replacing outside counsel. That's your clients replacing you."

"Meanwhile, the Big Four have invested over $10 billion in AI since 2023. PwC cut 5,600 employees globally while deploying AI agents. They're not adding AI to their headcount. They're replacing headcount with AI. And their cost structure now lets them profitably serve clients in your market — mid-market firms that previously didn't justify Big Four fees."

"But PwC went further. They launched a product called PwC One — a subscription platform offering automated tax guidance, M&A due diligence, and other services. No PwC person in the loop. Annual subscription. Their CEO told partners that anyone who thinks they can opt out of AI is 'not going to be here that long.' That's not an efficiency initiative. That's a productized offering competing directly with the work firms in this room do every day."

"So you have AI-native startups from one side and the Big Four selling subscriptions from the other. That's the squeeze."

"How many of your clients have mentioned AI in the last 90 days? What were they saying?"

Room discussion — two or three responses.
-->

---

<!-- _class: image chapter -->

![bg](images/office-windows-night.jpg)

## Part 2

# Why efficiency isn't enough

<!--
SPEAKER: "The obvious response to everything I just showed you — the instinct most firms are already acting on — only gets you partway there."
-->

---

<!-- _class: statement -->

## The efficiency trap

AI makes your work faster.

Faster work means smaller invoices.

**Smaller invoices mean less revenue** — not the same revenue, faster.

<!--
SPEAKER: "Let me make this concrete. If a task that used to take 10 hours now takes 2 — and you bill by the hour — your invoice just dropped 80%. The efficiency is real. But under hourly billing, the math works against you."

"This is not a hypothetical. Among law firms that widely use AI, 20% already report challenges meeting billable targets. Not future challenges. Current ones."

"And the instinct is to say: 'We'll just do more work.' But you're not getting paid to do more work faster. You're getting paid for the output. And the output is getting cheaper."

"The trap isn't efficiency itself — it's efficiency without model change. In the last year, have any clients pushed back on a bill because the work happened faster than they expected?"
-->

---

<!-- _class: light two-col -->

## The old game vs. the new game

### Old game

- AI is a productivity tool
- Do the same work faster
- Protect the billing model
- Free up our people for higher-value work

### New game

- AI is a commercial catalyst
- Redesign what you sell
- Price on outcomes
- Build revenue that compounds

<!--
SPEAKER: "The old game isn't stupid. It's the obvious response. And it works — up to a point."

"Most firms playing the old game describe it as 'freeing up our people for higher-value work.' That sounds right — until you ask: what is the higher-value work? What are you charging for it? Most firms can't answer that. The efficiency gains stay trapped in the P&L. Productivity goes up, pricing stays flat, and the margin improvement is real but temporary."

"A firm playing the new game uses AI to offer a service their clients couldn't get before — continuous monitoring, real-time reporting, proactive analysis — and charges a monthly fee for it."

"Same AI. Different question asked."

"I've seen this before. I was at NYDIG working with major banks on Bitcoin. Most of them treated it as another asset to put on the balance sheet — the efficiency play. Meanwhile, companies like River and Strike built new products around digital payments and captured the customers. The banks that bolted Bitcoin onto their existing model watched the revenue go somewhere else. Same pattern."
-->

---

<!-- _class: contrast -->

## The ceiling

- **25–40%**

  Efficiency gain from bolting AI onto existing models — then it stops

- **3×**

  More likely to see real EBIT impact when you redesign workflows instead

<!--
SPEAKER: "Nick Kervin is the managing partner of BDO — a top 10 accounting firm. He said publicly that efficiency within the existing business model has a ceiling of 25 to 40 percent. After that, you've extracted everything there is to extract."

"McKinsey's data shows that only 6% of organizations see meaningful profit impact from AI — and those firms are three times more likely to have redesigned their workflows rather than just adopting tools."

"The firms that are winning aren't the ones who bought the best tools. They're the ones who asked a harder question."
-->

---

<!-- _class: statement -->

The enemy isn't AI.

It's mistaking **efficiency** for **strategy**.

<!--
SPEAKER: "I want to be precise about this, because it's easy to hear all of this and conclude that AI is the threat. It's not."

"The threat is a mindset — specifically, the mindset that says: if I just do what I'm doing faster and cheaper with AI tools, I've done what I need to do. Efficiency gets you 25 to 40 percent. The question is what you do after the ceiling."

"The firms that are going to struggle aren't the ones that ignored AI. They're the ones that adopted it enthusiastically — and stopped at efficiency. They optimized a business model that's incompatible with what AI makes possible."

"Faster is not the same as better positioned. Those are different things."
-->

---

<!-- _class: statement -->

Not adoption. **Reinvention.**

<!--
SPEAKER: Three-beat pause. Let the word land.

"That's the fork in the road. Every firm in this room has adopted AI. Very few have asked what AI makes possible that wasn't possible before — and rebuilt around the answer."

"The last time this happened, that difference created Amazon and killed Borders. Created Netflix and killed Blockbuster. The technology was the same. The question was different."

TRANSITION: "So let me show you what the firms that survive this actually look like."
-->

---

<!-- _class: image chapter -->

![bg](images/skyscraper-dark.jpg)

## Part 3

# What the firms that survive look like

<!--
SPEAKER: "I've been doing business model transformation work for 15 years — across every major technology cycle since cloud. What I'm about to show you isn't theory. These are the four patterns that show up every time an industry goes through this shift."

"They're not about AI adoption. They're about what you offer and how you charge for it. AI is the enabler. The shift is commercial."
-->

---

<!-- _class: pillars -->

## Four shifts: how you deliver

- Episodic <span class="arrow">→</span><br>Continuous

  From discrete projects to always-on delivery

- Reactive <span class="arrow">→</span><br>Proactive

  From waiting for the call to surfacing problems first

<!--
SPEAKER: "Four patterns. Not every firm makes all four — but the ones seeing the biggest revenue impact are making at least two."

"These aren't about AI adoption. They're about what you offer and how you charge for it. AI is the enabler. The shift is commercial."

"First two."
-->

---

<!-- _class: pillars -->

## Four shifts: how you charge

- Hours <span class="arrow">→</span><br>Outcomes

  From billing time to billing results

- Bespoke <span class="arrow">→</span><br>Scalable

  From custom work per client to codified platforms

<!--
SPEAKER: "Last two. The fourth one — bespoke to scalable — is the hardest. And it's the one with the most durable competitive advantage once you've done it."

"Let me walk through each with real examples."
-->

---

<!-- _class: light process -->

## Episodic → Continuous

- Accounting

  Real-time delivery

  Client transactions feed an AI engine that updates the return continuously — not once a year.

- Staffing

  Always-on pipelines

  Continuous talent pipeline engagement — not scrambles after a requisition opens.

- Consulting

  Subscription relationships

  Monthly base plus usage billing. The client relationship doesn't end when the project does.

<!--
SPEAKER: "The episodic model — a matter, an audit, a search, a project — is how every firm in this room makes money today. It has a fundamental problem: every engagement starts at zero. You have to sell again, staff up again, ramp the client relationship again."

"The continuous model eliminates that. AICPA has a Dynamic Audit Solution that does real-time data ingestion. Deloitte has moved some clients to subscription-based advisory with monthly billing. Mastercard runs an always-on talent pipeline — not a requisition-based process."

"Notice what's different: it's not that they work harder. It's that they changed when and how value is delivered — and that changed when and how they get paid."
-->

---

<!-- _class: process -->

## Reactive → Proactive

- Accounting

  Find problems first

  AI monitors client data in real time — flagging anomalies before anyone notices.

- Staffing

  Anticipate needs

  Kuehne+Nagel: 22% higher candidate conversion by anticipating demand before the role opens.

- Consulting

  Generate insight, not just analysis

  McKinsey's Lilli: used by 72% of staff — research time down 30%, capacity redirected to proactive work.

<!--
SPEAKER: "The reactive model — client calls with a problem, you solve it — is also how every firm here operates. The client is in charge of the relationship."

"The proactive model flips that. You call them. You surface the issue before they knew it existed. That changes the power dynamic entirely."

"Think about what it means for client retention when you're the one who finds the problem. Think about what it means for billing — you're not waiting to be engaged, you're already engaged."

"Kuehne+Nagel ran AI across their workforce data to anticipate talent needs before managers submitted requests. 22% higher conversion. 20% faster time-to-fill. Same people, different question asked."
-->

---

<!-- _class: light two-col -->

## Hours → Outcomes

### The shift is already happening

- 59% of law firms now use flat fees alongside or instead of hourly
- 34% more flat-fee billables in 2025 vs. 2016
- 73% of consulting clients now prefer outcome-tied pricing

### The counterintuitive finding

- Clients expect AI to cut costs 30%
- But 100% would pay more for AI-enabled advisory that generates more value
- Outcome pricing captures the upside. Hourly pricing doesn't.

<!--
SPEAKER: "Here's the piece that surprises most managing partners."

"Clients expect AI to make your work cheaper. They're right. But they also say — 100% in Source Global Research's survey — that they would pay more for advisory that generates more value."

"The firms that are suffering are the ones racing to pass the cost savings to clients. The firms that are winning are repackaging the savings as new capability — and charging for the outcome."

"Flat fees aren't just a billing preference. They're a signal that you're confident enough in your delivery to price the result, not the time. That confidence is what clients are buying."

"I spent years backing music-tech companies that were building for a world where streaming was the norm — while the labels were still fighting it, trying to keep selling packaged product at $15 a unit. The companies that moved to subscription pricing captured the industry. The labels that licensed early instead of suing set the terms. That's the same shift from hourly to outcome-based — the firms that move first don't just survive, they set the terms for everyone else."
-->

---

<!-- _class: quote -->

> To transition from bespoke to scalable, firms must cultivate the courage to codify their knowledge.

**CB Insights** · Future of Professional Services

<!--
SPEAKER: "The fourth shift — bespoke to scalable — is the hardest one, and this quote names why."

"Courage. Not technology. Not capital. Courage."

"What it means in practice: the knowledge that lives in your senior people's heads — the judgment calls, the pattern recognition, the 'here's what I'd do' — has to get written down, templated, systematized."

"That's uncomfortable. It feels like giving something away. But the firms that do it create a moat — because their expertise is now embedded in a platform, not a person. That's very hard to copy."
-->

---

<!-- _class: light process -->

## Bespoke → Scalable

- Law

  Codified expertise

  Garfield.Law delivers legal services entirely through AI — per-document pricing, no hourly billing. Expertise is in the platform, not the person.

- Accounting

  Productized advisory

  BDO built AI-powered advisory tools that package senior partner judgment into repeatable, scalable client offerings.

- Consulting

  Knowledge platforms

  McKinsey's Lilli codified institutional knowledge — 72% of staff use it daily. Research time down 30%, expertise accessible firm-wide.

<!--
SPEAKER: "This is the hardest shift — and the most valuable once you've done it."

"Garfield.Law is the clearest example. They took everything a solicitor knows and embedded it into a platform. A regulator approved it. They charge per document, not per hour. The expertise isn't in a person — it's in the system."

"BDO did something similar on the advisory side — they took the judgment calls that only senior partners used to make and built tools that deliver that judgment at scale."

"And PwC did it at the largest scale of all. PwC One packages six services into a subscription product — no person involved. Tax guidance, M&A due diligence, automated and delivered directly to clients for an annual fee. That's the biggest firm in professional services saying: we codified our expertise, built it into a platform, and now we sell it without billing a single hour."

"McKinsey's Lilli is the consulting version. Seventy-two percent of their staff use it every day. That's not an experiment — that's a new operating model."

"The common thread: each firm had the courage to write down what they know and build it into something that doesn't require a specific person to deliver."

"Microsoft hired me to work with about a dozen of their software partners making this exact shift. The ones that saw cloud as just cheaper distribution — a way to cut hosting costs — lost. The ones that saw it as a business model opportunity and redesigned what they sold won. That's the fork in the road you're looking at right now."
-->

---

<!-- _class: contrast -->

## The revenue divergence is already measurable

- **2×**

  Revenue growth among law firms that adopted modern delivery models

- **−50%**

  Revenue decline among firms that didn't — same four years (Clio Legal Trends)

<!--
SPEAKER: "This is four years of data. Same market. Same economic conditions. Firms that shifted delivery models — not just tools, but models — nearly doubled revenue with only 50% more clients."

"Firms that didn't lost half their revenue."

"I want to be honest: this is early data from legal. The equivalent studies in accounting and consulting aren't published yet. But the direction is clear, and the mechanism is the same."

"The divergence is already underway. The question is which side of it you're on."
-->

---

<!-- _class: image chapter -->

![bg](images/city-dawn.jpg)

## Part 4

# How to start on Monday

<!--
SPEAKER: "Last act. Everything I've covered is real — but the goal isn't to leave you impressed by the data. The goal is something specific you can do this week."
-->

---

<!-- _class: statement -->

## The decision

You can keep using AI to **work faster.**

Or you can use AI to **sell differently.**

_Most firms are only making the first decision._

<!--
SPEAKER: "I want to name the decision precisely, because it's easy to conflate these."

"The efficiency decision gets made by default. Your people are already finding AI tools, using them, doing work faster. That's happening whether you plan it or not."

"The commercial decision is different. It asks: given what AI makes possible, what should we offer that we don't offer today? What can we sell that we couldn't sell before? What clients should we be going after?"

"That decision doesn't get made by default. It requires the managing partner to stop, look at the business model, and ask a harder question."

"How many of you have had that conversation in the last 90 days?"
-->

---

<!-- _class: pillars -->

## The other side

- Revenue that compounds

  Not projects that reset to zero. Clients stay because you're embedded in how they run.

- Pricing that goes up

  You sell outcomes, not time. Firms offering advisory services earn 30%+ higher monthly recurring revenue.

- A firm worth more than its hours

  One that creates value clients can't replicate — and can't replace you for.

<!--
SPEAKER: "This is what you're building toward. Revenue that doesn't reset every quarter. Clients you don't have to re-sell every engagement. Pricing that goes up as your delivery improves — not down as AI commoditizes the work."

"Keep this destination in view. The rest of this act is about how to start getting there."
-->

---

<!-- _class: light pillars -->

## Three things in the way

- Where to start

  53% of mid-market firms feel only "somewhat prepared." The top barrier isn't AI — it's no one in the room who knows what's commercially possible.

- The revenue trap

  Default AI adoption compresses revenue under hourly billing. More tools won't fix a broken model.

- Team alignment

  Workflow redesign had the single strongest contribution to business impact across 31 variables tested — and it requires the senior team, not just the managing partner.

<!--
SPEAKER: "These are the three things I hear most often from firms that know they need to move but haven't."

"Notice what's not on the list: budget. Resources. Technology. Those aren't the bottleneck."

"The bottleneck is a clear-eyed commercial conversation with the right people in the room."
-->

---

<!-- _class: light agenda long -->

## Questions to ask your partners this week

1. If AI cuts our delivery time in half, what happens to our revenue under the current billing model?
2. What would we sell if we could serve clients continuously instead of per-engagement?
3. Which of our services could we price on outcomes instead of hours?
4. What do our clients pay us for that AI can now do — and what do they pay us for that it can't?

<!--
SPEAKER: "I want to leave you with something concrete. Four questions. Not AI questions — business model questions."

"Don't bring these to your IT team. Bring them to your next partner meeting. Put them on the agenda. See what happens."

"The first one is the math. The second one is the business model shift. The third one is the pricing shift. The fourth one is the strategic question — where's the moat?"

"If your partners can't answer these, that's not a failure — it's a signal that you need a structured way to work through them together. That's what the two days are designed to do. But start with the questions."
-->

---

<!-- _class: statement -->

## The first move

**[upshiftco.com/assessment](https://upshiftco.com/assessment)**

_8 minutes. No account. Free._

<!--
SPEAKER: "Before you redesign anything, you need to know where you actually stand."

"Pull out your phone. Go to upshiftco.com/assessment. 8 minutes. You'll get a diagnostic — where your firm has real commercial gaps, how you compare to peers in your vertical, and a report you can take to your next partner meeting."

"It's free. It's private. Your answers never leave your browser."

"I work with firms on this — a full day with your senior team, then offering briefs ready to test with real buyers. But that's a conversation for after you see where you stand. Start with the diagnostic."
-->

---

<!-- _class: statement -->

## My goal

Not to convince you AI matters.

**You already know.**

_Shift from "we should probably do something" to "we start Monday."_

<!--
SPEAKER: "I want to close with this, because it's the only thing that matters."

"You came in knowing AI was important. I haven't told you anything you couldn't have read somewhere. What I've tried to do is give you a frame — a way of thinking about it that leads somewhere actionable."

"The enemy is not AI. The enemy is a specific mindset: that efficiency alone is enough."

"The firms that win this aren't going to be the ones who hired the most AI consultants or bought the most tools. They're going to be the ones where the managing partner sat down and said: what do we actually sell in a world where AI can do most of what we used to charge for?"

"That's the question. You start Monday by sharing the assessment report with your partner group. That's the conversation that leads to everything else."

Open for Q&A.
-->

---

<!-- _class: closing -->
<!-- _paginate: false -->

<img class="logo" src="logo.svg" alt="">

# Let's talk.

shawn@upshiftco.com

[upshiftco.com](https://upshiftco.com)

<!--
SPEAKER: Leave up during Q&A.

Post-talk opener: "What landed most for you — what's most relevant to where your firm is right now?"
-->

---

<!-- _class: light sources -->

## Sources

- 88% AI adoption; 6% see EBIT impact; 3× with workflow redesign _McKinsey, State of AI 2025_
- 60% of marketing leaders already cut agency spend _CB Insights, Future of Professional Services_
- 87% use AI in recruiting; $600B+ staffing industry threatened _Bloomberg, Feb 2026_
- 20% of AI-adopting law firms miss billable targets _Thomson Reuters / Georgetown, 2026_
- Big Four: $10B+ invested in AI since 2023; PwC cut 5,600 jobs _Investor disclosures_
- Harvey AI: $8B valuation, 42% of Am Law 100, 50% in-house _Harvey AI_
- 59% of law firms use flat fees; 34% more flat-fee billables _Clio Legal Trends, 2025_
- 73% prefer outcome pricing; 100% would pay more for AI advisory _Source Global Research_
- Efficiency ceiling: 25–40% _BDO, Nick Kervin_
- Kuehne+Nagel: 22% higher conversion, 20% faster fill _Kuehne+Nagel_
- McKinsey Lilli: 72% of 45K staff, research time down 30% _McKinsey_
- 2× revenue with −50% fewer clients over 4 years _Clio Legal Trends, 2025_
- Advisory firms earn 30%+ higher MRR _AICPA/CPA.com CAS Benchmark Survey, 2024_
- PwC One: 6 automated services, subscription pricing, no person in loop _CPA Practice Advisor, March 2026_