Is AI lowering consulting firm valuations? What buyers ask in 2026

Is AI lowering consulting firm valuations?#

AI is lowering valuations for some consulting firms and not others, and the difference is how much of a firm's revenue rests on hours that AI can compress. Buyers have responded by examining exposure one firm at a time. Adopting AI does not settle the question either: Bain's September 2026 analysis of private equity portfolios found little correlation between AI spend and value at most portfolio companies, so buyers look for AI effects in pricing, margins and retention rather than in tool budgets.

The effect shows up in two places: the multiple a buyer is willing to pay, and the shape of the offer. A firm with exposed revenue may see more of its price moved into an earnout that pays only if revenue holds after the sale. A firm with protected revenue may see the opposite, because a buyer wants to secure capabilities that AI makes more valuable.

Which consulting work faces the most AI pressure?#

AI pressure concentrates in work that was sold as analyst capacity. It is weakest where the client pays for access, accountability or evidence nobody else has, and a buyer will map your revenue against that divide early in diligence.

Most firms have a mix. The useful exercise is to tag each offering in the PSA by type, so revenue by exposure level comes straight from your own reports rather than from a buyer's estimate.

The five AI questions buyers ask in diligence#

Buyers and their advisors have added a set of AI questions to standard commercial diligence. Each one asks, in a different way, whether earnings will survive as clients and competitors keep adopting AI.

The questions are less hostile than they sound. A buyer asking them is usually trying to decide how much of the price to pay at closing and how much to defer, so every question answered with a document rather than an assurance is a direct argument for more cash up front.

How to answer the pricing and retention questions#

Pricing and retention questions are answered with reports from your own systems, so the work is mostly assembly and reconciliation. Pull revenue by pricing model from the accounting system and PSA, and client-level revenue history from the CRM, then reconcile them so a buyer's analyst gets the same answer from every source.

If most revenue is hourly, show the plan and the early evidence for moving offerings to fixed or outcome fees, rather than leaving the buyer to assume the worst. A price book with even a few non-hourly lines, and signed engagements under them, is more persuasive than a slide about intentions.

For retention, explain churn rather than just reporting it. A client that left because a program ended is different from one that replaced the firm with an internal team using AI, and the buyer will want to know which is which.

How to answer the IP, AI policy and data questions#

IP, AI policy and data questions are answered with documents, and most firms can assemble them without outside help. The ownership memo deserves the most care: a buyer that finds client-owned material inside a supposedly proprietary benchmark will start doubting the rest of the IP story.

The data question carries more weight than it did a few years ago. TechCrunch reported in October 2025 that Mercor's CEO described AI labs recruiting former senior staff of consulting, banking and law firms because the firms themselves would not hand over data that could automate their work. A firm that can show what it holds, and who owns it, is showing a buyer an asset others are trying to reach indirectly.

For the AI policy, evidence of practice matters more than the policy text. A buyer will ask how the policy is enforced, so keep the usage log, the training attendance record and the notes from any client that asked about AI use, and show how each request was handled.

• 1A methods binder: each named method, who wrote it, where it has been used and which templates support it.
• 2An ownership memo separating firm-owned frameworks and benchmarks from client-owned deliverables, citing the MSA clauses that decide each.
• 3The AI use policy, the approved tools with their data terms, and a record of staff training.
• 4A record inventory: systems, years of accessible history, record families and known restrictions.
• 5Any data licenses or benchmark subscriptions already in place, with their exclusivity and term.

Illustrative: an HR consultancy reframes its story for a buyer#

Illustrative: a fictional HR and organizational design consulting firm receives an approach from a larger professional services group. The buyer's first data request asks for revenue by pricing model and a description of how the firm uses AI.

Most of the firm's revenue is hourly, and its compensation benchmarking survey, its strongest asset, has never been documented as firm property. Before the next meeting, the firm documents the survey method and the client terms permitting aggregate use, writes down its AI policy, and shows two offerings already moved to fixed fees. The buyer still proposes an earnout, but the discussion shifts from whether AI erodes the firm to how quickly the new pricing can scale.

Where SourceX fits in an AI-era sale#

SourceX helps a firm build the record inventory buyers ask for, using the SourceX Enterprise Data Value Framework to describe which records show expert work linked to outcomes and what rights apply. The first fit check uses metadata only, so nothing confidential leaves the firm during a sale process.

Firms that choose to license records go through the SourceX five-step transaction, and each package carries a SourceX Evidence Packet, so any existing license is documented in a form a buyer's counsel can review quickly.