The New Economics of Professional Services

The conventional narrative about AI and professional services billing is as follows: AI compresses delivery time, rendering the billable hour increasingly difficult to justify, and prompting firms to consider a shift to outcome-based pricing. However, this view is overly simplistic. It assumes professional services sell only effort, when in reality they provide a bundle: output, judgment, accountability, defensibility, responsiveness, and institutional trust. AI does not reveal a single, hidden truth about these offerings. Rather, it separates the bundle into two distinct components. One component is commodity production research, including drafting, analysis, review, and related tasks, where AI delivers significant efficiency gains. The other component is accountable judgment, synthesis under uncertainty, organizational alignment, decision support, external credibility, and the assumption of responsibility when outcomes do not meet expectations. These components follow different economic dynamics and will be priced accordingly.

This report contends that effort-based pricing is harder to justify in production-heavy workflows, but it only shifts when four conditions are met: outcomes are measurable, the provider affects results, buyers accept metrics, and providers can absorb liability. If any condition is missing, the billable hour remains because the economics do not yet support alternatives. The main constraint is not a single factor but the mix of provider risk appetite, outcome measurability, buyer habits, and regulations. Firms that succeed will not adopt one pricing model but will distinguish between commodity work and judgment-based services, pricing each appropriately.

Firms are currently capturing the AI surplus, though the factors sustaining this advantage are deteriorating unevenly across the sector.

A 2025 Axiom study of 600+ senior legal leaders in eight countries found 79% of law firms use AI for client delivery. Only 6% charge less for AI-assisted work, while 34% charge more. Law firm revenue rose 13% in 2024. Billing rates increased 6.5% in 2024 and 7.3% in 2025, the fastest since the financial crisis. The AI productivity gain is currently flowing to provider margins.

There is precedent. When Westlaw and LexisNexis automated legal research in the 1970s-1980s, firms took the efficiency gains, and rates kept climbing. E-discovery platforms in the 2000s triggered similar results. In consulting, analytics, benchmarking, and knowledge management, improved delivery but did not change fee structures.

Three developments are putting more pressure on the current equilibrium than before. First, delivery-cost compression is significant. A Harvard Business School/BCG experiment found that AI users completed 12.2% more tasks, 25.1% faster, and with over 40% higher quality. In legal services, a complaint-response system cut associate time from 16 hours to under four minutes. A Stanford study found AI-using accountants finalized statements 7.5 days faster with 12% higher quality. These are not minor improvements; they represent major shifts visible to buyers.

Second, competitive alternatives now price in AI efficiency. The alternative legal services provider market hit $28.5 billion in 2023, growing 18% annually. In-house counsel numbers rose 87% since 2008. In IT, global capability centers employ 1.8 million people and take on work once done by outsourcers. These are real alternatives for clients.

Third, buyer expectations are changing quickly. Source Global Research found 58% of consulting clients now expect fees to rise, up from 27% in late 2024 and 5% pre-pandemic. An ACC/Everlaw survey of 657 in-house counsel across 30 countries found 61% plan to seek changes to legal delivery and pricing.

It is necessary to clarify what these signals represent. Rising client expectations are genuine, but such shifts have previously outpaced reality in professional services procurement. The ALSP market is expanding rapidly, yet it primarily addresses commodity work, while high-value advisory continues to be directed to traditional firms. Stock declines and headcount reductions in IT services and agencies reflect sector stress and delivery model adjustments, but do not directly indicate changes to billing models. The 2026 Georgetown Law/Thomson Reuters report on the US legal market describes a “standoff,” with clients seeking innovative billing models and firms reporting that clients ultimately compare hourly rates. Both parties are hesitant to move first. While the conditions for disruption, including productivity gains, competitive alternatives, and evolving buyer expectations, are more visible, simultaneous, and commercially salient than in any prior technology cycle, the current equilibrium remains intact. The strain, however, is tangible.

The pricing mechanics are mostly recycled; the genuinely new question is what justifies the fee.

By 2016, 97% of law firms invoiced at least some work on a non-hourly basis, and alternative fee arrangements accounted for $21.1 billion of outside counsel spending. Yet hourly billing remained dominant for complex work. The alternatives, fixed fees, retainers, hybrid structures, and blended rates were available but did not displace time-based billing at scale.

The reason is instructive. Most alternative pricing models change the unit of billing but not the price the client pays. A fixed-fee engagement moves scope risk to the provider but not outcome risk. A retainer shifts capacity risk but not quality risk. These are meaningful shifts in commercial mechanics, but they do not change the basic economic relationship between effort and payment.

What distinguishes the current moment is that a small number of firms are beginning to fundamentally restructure their commercial relationships. According to industry reporting, approximately 25% of McKinsey’s global fees now come from outcomes-based arrangements, measured against scorecards that include investor targets, revenue goals, operational metrics, and customer satisfaction. In a structurally different context, Intercom’s AI resolution agent charges $0.99 per successful customer resolution. If the AI fails, the client pays nothing, and this approach has quadrupled revenue year over year. PwC is building toward managed services, combining its Agent OS platform with advisory overlays, aiming to capture 20–25% of advisory revenue from such arrangements. KPMG targets 15–20% of US consulting fees from similar models.

The risk absorption spectrum ranges from time-and-materials at one end to full gainshare at the other. This is a useful lens for comparing arrangements. But risk absorption does not fully explain pricing shifts. A firm may want to take on downside risk but be unable to do so if the outcome cannot be measured, if the buyer's process cannot handle contingent fees, or if regulations block the arrangement. Any serious market view must consider all four conditions: measurability, provider control, buyer acceptance, and liability capacity. Where all four are present, new pricing models emerge. If any is absent, the billable hour continues, and not just from inertia.

Five sectors under different kinds of strain

The rate of change in billing models is not uniform across professional services. It varies based on regulatory protection, client leverage, and the degree to which AI automates the production component of the work. It is important to distinguish between evidence that signals sector stress versus actual change in billing models. For example, a stock decline signals investor anxiety, and a headcount reduction signals leverage-model adjustment. Only changes in actual contract terms and fee structures constitute evidence of a billing model change. The sectors below experience all three kinds of pressure, but with differing intensity and proportion.

IT services face the most direct pricing pressure. India’s IT services industry confronts what analysts describe as its most consequential inflection since outsourcing began in the 1990s. The Nifty IT index fell 19% in February 2026. Jefferies estimates that application-managed services, which represent 22–45% of revenues, face sharp revenue deflation. But the clearest evidence of actual repricing comes from executive statements: LTIMindtree’s CEO has confirmed that clients are seeking to convert time-and-materials contracts to outcome- or managed-services-based structures. HFS Research reports clients demanding up to 20% price reductions. The entire business model, labor arbitrage multiplied by headcount, is being challenged by AI substitution, and the sector has neither regulatory barriers nor brand premiums to cushion the adjustment.

Digital agencies face acute client-driven pressure. A 2025 Typeface survey found 60% of senior marketing leaders are spending less on agencies as a direct result of AI. The largest holding companies’ share of US advertising expenditure dropped from 44.6% in 2019 to 29.6% by early 2024. Brands are demanding 25% fee reductions. The Omnicom-IPG merger eliminated approximately 4,000 positions and targets $750 million in savings. The evidence here is strongest for sector stress and delivery-cost compression. Evidence of fundamental contract-model change is thinner — many agencies have operated on retainers and project fees for years, and the current shift may represent fee compression within existing models rather than a migration to new ones.

Management consulting is experiencing delivery transformation faster than pricing transformation. McKinsey’s internal AI tool Lilli processes over 500,000 queries monthly, with 72% of employees as active users. BCG’s AI consulting grew from zero to an estimated $2.7 billion in two years. McKinsey’s workforce shrank from approximately 45,000 to 40,000 by mid-2025. But well-resourced clients do not mainly pay top consulting firms for access to information or analytical labor. They pay for synthesis under uncertainty, organizational alignment, implementation horsepower, external credibility, and someone to own part of the risk of being wrong. AI compresses the first of those value layers, the analytical production work, while potentially raising the relative importance of the others. The bigger risk to consulting is not that AI democratizes knowledge. It is that it forces the profession to articulate which parts of a $3 million engagement are commodity production and which are accountable judgment, a distinction that the current pricing model conveniently obscures.

Accounting occupies a regulated middle ground. EY’s Helix platform now analyzes 100% of a client’s journal entries versus traditional sampling. KPMG’s continuous auditing capabilities reduced time to detect financial irregularities by 75%. Graduate job postings dropped 44% year-over-year across the Big Four. Delivery transformation is substantial. But PCAOB and FRC regulations mandate licensed human sign-off on audit opinions, creating a regulatory floor beneath which efficiency gains cannot directly translate to fee reductions. The KPMG/Grant Thornton precedent in which KPMG demanded and received a 14% audit fee reduction from its own auditor, explicitly citing AI efficiencies, demonstrates that even regulated work is not immune to repricing pressure. But it remains a single, conspicuous case. The broader market has not yet followed.

Legal services show the strongest structural resistance to billing model change. Despite AI productivity gains comparable to other sectors, the 2026 Georgetown/Thomson Reuters report found that 90% of all legal dollars still flow through hourly billing. The profession’s self-regulation through bar associations, partnership economics that incentivize the status quo, and a compensation system built entirely around billable hours make the legal sector a place where delivery transformation and pricing transformation are most disconnected. The primary disruption vector is not incumbent firm pricing innovation but competitive substitution through ALSPs, in-house teams, and AI-native legal technology platforms that price services differently by default.

The contract layer is being rewritten and may matter more than the fee structure.

Beyond pricing, the terms of professional services agreements are undergoing rapid revision that most industry commentary underestimates. These changes span disclosure, data rights, intellectual property, liability, and governance, and they are creating a compliance infrastructure that is becoming a source of competitive differentiation.

The American Bar Association’s Formal Opinion 512, issued in July 2024, established that existing professional conduct rules on competency, confidentiality, and reasonable fees apply to AI-assisted legal work. Disclosure is mandated when clients inquire about AI use, when AI-related costs will be billed, when confidential data enters AI systems, or when court rules require it. State-level rules are proliferating: Florida requires informed consent when client data enters AI tools; Pennsylvania requires disclosure of AI-related expenses; California’s AI Transparency Act takes effect in January 2026.

At the contract level, provisions increasingly common in sophisticated enterprise agreements include clauses preventing providers from using client data to train AI models, requirements for complete sub-processor disclosure, data isolation provisions, and deletion rights with certification. Intellectual property ownership of AI-generated work product remains legally unsettled. Liability frameworks in higher-risk enterprise contexts are evolving toward two-way indemnification covering IP infringement, data breaches, and algorithmic bias.

Governments are building procurement scaffolding. The EU published Model Contractual Clauses for AI Procurement in March 2025. Australia released Model Clauses v2.0 the same month. The EU AI Act, with full enforcement of high-risk provisions by August 2026, will impose tiered obligations that reshape how professional services are contracted.

Firms that build AI governance compliance, auditability, provenance tracking, and quality assurance into their delivery models rather than treating them as bolt-on requirements are positioning for the premium end of the market. Whether this positioning translates to durable pricing power depends on whether governance expertise remains scarce or becomes rapidly commoditized through template adoption. The early evidence is ambiguous. What is not ambiguous is that the contract layer now contains more commercially consequential decisions than the fee structure in many AI-augmented engagements.

Where pricing power will concentrate and where it will not

The direction of travel is clear in broad terms: production-heavy work will be repriced; judgment-heavy work will command premiums; and the boundary between the two will be contested in every sector. But precision matters more than directionality. Three implications follow from the analysis above.

The first is that the billing model change will be concentrated, not universal. Pricing shifts where all four conditions are present: the outcome is measurable, the provider can influence the result, the buyer accepts the metric, and the provider can bear the liability. In IT application management, all four conditions are increasingly met: outcomes are measurable (uptime, defect rates, resolution times), providers can control them, buyers accept the metrics, and providers can absorb the financial risk. In a contested M&A strategy, the first condition alone is often absent, which constitutes a “good outcome” that may not be knowable for years. Firms should map their portfolios against these four conditions rather than adopting a single pricing philosophy.

The second is that contract modernization is a more urgent priority than pricing innovation. Most professional services agreements were written for a world in which humans did the work, with the main commercial variables being scope, timeline, and rate. AI introduces new variables, data rights, model governance, output provenance, liability for AI errors, and human review requirements that existing contracts do not address. The regulatory environment is hardening faster than market pricing is shifting. Firms that proactively build AI governance into their contracting templates gain both a compliance advantage and a trust advantage with sophisticated buyers.

The third is that the firms best positioned for the next era are those that can clearly articulate to themselves and to clients which parts of their work are commodity production and which are accountable judgment. The billable hour will persist for years, embedded in compensation structures, procurement processes, and professional identity. But it will increasingly be confined to work where the buyer genuinely needs to purchase judgment, accountability, and the provider’s willingness to be wrong, not merely the provider’s time. The firms that recognize this distinction, and price each component honestly will define the terms on which the transition occurs.