When the Taxi Meter Stops Working

AI helps consulting practices deliver faster. Much faster. Work that took ten hours a year ago can take two. The client gets the same result, often a better one, weeks sooner. That is the benefit of better technology.

The commercial model for consulting has not caught up, though. At an hourly rate, firms would now collect a fifth of their fee. The client benefits immediately, and firms recover the other eight hours' revenue only if they find more billable work to fill them. That is an imbalance we need to fix quickly. An hourly model leaves a consulting firm worse off for using technology to do better work sooner.

David Holender of Vaimo described the problem in a company press release: “Too often, the agency model works like a taximeter.”¹ The meter measures time. AI is making time a less useful measure of what the client gets.

For a defined engagement, firms can use a fixed fee. The client and consultants agree on the work, the acceptance criteria, the responsibilities, and the price. If deliverables are deployed faster, the fee stays the same. A fixed-price agreement lets firms set the price from the value of the work to the client.²

If firms build the fixed price from estimated hours and add a contingency, though, they are still pricing effort. AI reduces the estimate, and the fixed fee falls with it.

A performance fee can make the price better match the value created. It also creates a harder question about attribution. Revenue can rise in the same quarter that the client raises prices. Handling time can fall after a new policy takes effect. Neither result tells us how much came directly or even tangentially from the consulting work. If we leave that question until the invoice is due, clients and consultants will be arguing about money without an agreed way to settle it.

The Value Radius

Before firms price a performance component, they must deeply understand how an order, request, case, or other entity moves through the client's business and where people, systems, and policies shape it. Every project plan they write should include a map of the client workflows their deliverables will affect. The plan must cover deliverables, team, budget, and dates, as usual. But this new workflow map must also define the project’s impact: where the work lands, which handoffs it changes, and where effects may appear downstream.

Suppose a firm put an agent in front of customer intake, so requests reach qualification faster. The qualification team may not have capacity for the extra volume. The bottleneck moves down the hall and the customer still waits. Consultants can meet every project target and still leave the customer experience inadequate.

That is why firms must now trace the change through the workflow before they price it. If the delivered intake agent is meant to speed up qualification, they need to know what happens next: does the exceptions queue have room for more work? Can the client show the result in a system of record? Who decides whether team or agency staffing changes? Those questions determine whether a claim can support a performance fee.

I call the resulting set of provable claims the Value Radius. It contains the claims that the client and consultant can trace, measure, and settle fairly enough to perhaps include in a payment formula. The Value Radius builds on benefits mapping, which traces the dependencies between a change and a business result.³

The work of proving value

A Value Radius takes work to establish. Consultants must work with their clients to map the workflow, test the available data, and identify the baseline, dependencies, and risks to attribution. The client makes the relevant people, systems, records, and decision-makers available. Together, they decide which claims might be able to support a payment formula.

This Value Radius Modeling is a fixed-fee assessment in its own right. It produces a workflow map, a measurement plan, a list of claims that can be priced fairly, and a recommendation for the commercial model. The result may support a fixed-fee implementation, a recurring operating fee, a performance component, or a decision to make no outcome claim.

The client gets useful operational analysis either way. Consultants get paid for the judgment, experience, and design work that makes a later implementation contract credible.

If the client proceeds, the consultants should attach a measurement plan amendment to the implementation SOW. The SOW sits under a master services agreement that covers the terms of the relationship.

Agree on what counts

Take a touches-per-exception engagement. Value Radius Modeling establishes whether reducing touches can support a performance fee and what it would take to measure it. When the answer is yes, the implementation statement of work incorporates the resulting measurement plan. When the evidence is incomplete, the next engagement can be a fixed-fee pilot that establishes it before a performance component is added.

The plan names the source data and baseline period. It sets the comparison method and any workload or case-mix adjustments. It also defines the quality and cycle-time guardrails, review cadence, payment formula, cap on the variable fee, and a deadband for ordinary variation. It says what happens if source data goes missing or the rollout departs from the agreed design. In short, the plan needs to identify and quantify the variables the workflow affects or is affected by. ⁴

If the client has two comparable exceptions teams, firms should look to pilot with one and use the other as a comparison during the measurement period. If the comparison team's work is affected by the implementation, the comparison no longer isolates the effect of the consulting deliverables. Spillovers like this can bias the estimate, so the design needs to account for them.⁵

The master services agreement and SOW also separate the client's commitments from the firm's. If the client keeps team or agency staffing static, does not adopt the workflow, changes a policy, or blocks access to source data, those choices affect the result and the measurement period. The agreements need to say how they affect the payment.

From capacity to a fee

Consider an illustrative distributor handling 2,000 order exceptions a month. Each exception takes six manual touches. The proposed work reduces that number to 2.5.

That removes 7,000 touches a month. If measured handling time shows that each touch takes ten minutes, the work releases about 1,167 hours a month. At a loaded labor rate of $60 an hour, that capacity is worth $840,000 a year.

That figure describes capacity. A cash saving occurs when spending falls. The people may still be on the payroll, and the business may use the time to handle more orders or improve service. Released capacity and a direct reduction in spending are different benefits.⁶

Now assume the distributor reduces agency use from 800 to 200 hours a month. The agency charges $50 an hour. Over twelve full operating months, agency spending falls by $360,000.

The comparison team receives the same general policy and system changes, but not the consulting workflow deliverables. After adjustment for workload and case mix, the agreed analysis estimates that agency use would have fallen from 800 to 700 hours a month without the consulting work. The calculation then excludes those 100 hours, which are worth $60,000 over the year.

Performance-fee calculation for twelve full operating months	Amount
Gross reduction in agency spending	$360,000
Less: reduction expected without consulting work	$(60,000)
Agency savings attributable to consulting work	$300,000
Less: additional software and running costs	$(36,000)
Performance-fee calculation base	$264,000
Illustrative performance fee at 12%	$31,680

The $840,000 capacity figure stays outside this calculation because it estimates the wider value of the freed-up team time. The performance fee uses the $264,000 calculation base because it has a baseline, a comparison, and records both sides can review. Assessment, implementation and performance fees still need to be deducted to calculate the client's overall net benefit.

The client may use released capacity to absorb more orders or improve service. Those may be meaningful benefits, but they sit outside this performance fee equation until the client and consultants have evidence that separates the consulting contribution from demand, fulfilment costs, and the client's own decisions.

Make the engagement worth doing

Value Radius Modeling has a price of its own. If the client chooses to proceed, firms can then charge an implementation fee for accepted delivery. When firms operate the workflow, they can charge a recurring or managed services fee. Firms should also add a performance component only where the measurement plan can support it.

The Value Radius Modeling fee pays for the assessment. The implementation fee has to make delivery worthwhile even if the variable payment never arrives. A payment based on a result a year later cannot be the funding plan for the assessment, implementation, support, and measurement work that comes first.

Client decisions still shape the risk firms carry. Consultants may show that the work released capacity but the client may keep the agency hours in place. Consulting work can free up capacity without reducing ledger cost. The contract needs separate commitments on adoption, staffing, operations, data access, change control, and other operational actions. If consultants take responsibility for a wider business result, they likewise need the authority and access to influence it.⁷

Some engagements need no performance component. A client can pay a fixed fee for an assessment or architecture that improves a decision without tracing it to annual profit or a measurable financial benefit. Even when a result is measurable, the expected performance payment has to justify the cost of measuring and verifying it.

What replaces the meter

Value Radius Modeling gives each part of an engagement a price that reflects its work and risk. The master services agreement covers the relationship. Each SOW names the work and its base price. A measurement schedule attaches when the evidence supports a performance component.

The client knows what they are buying before work starts, and most importantly, the Value Radius of the work is understood and agreed to. If a variable payment applies, the agreement says what it measures and which decisions affect it. Faster, better work improves the economics of the engagement when using this commercial model. Hourly billing might still be a component of a firm's commercial model, but it should be a much less impactful one. Consultants must of course still track time to manage their own effort. For engagements using this model, time should not decide what the client pays.

AI can make consulting approaches rooted in deep business knowledge more profitable for both sides. The client gets better work sooner. The consultant gets to redesign work, manage the dependencies around it, and share in results that can be proved.

Firms that keep selling hours will feel the fee pressure. Firms that change their offer, contracts, and delivery model can turn the same tools into better client results and stronger margins.

Sources

Vaimo, “AI Is Forcing a Rethink of the Billable Hour in Digital Consulting: Vaimo,” company press release distributed by PR Newswire, 10 February 2026. https://www.prnewswire.com/news-releases/ai-is-forcing-a-rethink-of-the-billable-hour-in-digital-consulting-vaimo-302683788.html ↩
Ronald J. Baker, “Pricing on Purpose: How to Implement Value Pricing in Your Firm,” Journal of Accountancy, June 2009. https://www.journalofaccountancy.com/issues/2009/jun/20091530/ ↩
John Ward and Elizabeth Daniel, Benefits Management: Delivering Value from IS & IT Investments, Wiley, 2006, chapter 4, which introduces the Benefits Dependency Network. Bibliographic entry and contents: https://www.oreilly.com/library/view/benefits-management-delivering/9780470094631/ . For an accessible explanation, see Hugo Minney, “Benefits mapping: the foundation for benefits management,” Association for Project Management, 2 December 2024. https://www.apm.org.uk/blog/benefits-mapping-the-foundation-for-benefits-management/ ↩
Umbrex, “Outcome-Based Pricing,” undated, especially sections 3 and 5. https://umbrex.com/resources/frameworks/pricing-frameworks/outcome-based-pricing/ ↩
J-PAL, “Data analysis,” especially the discussion of spillovers and research design. https://www.povertyactionlab.org/resource/data-analysis?lang=en ↩
HM Treasury, “The Government Efficiency Framework,” updated 24 November 2025. https://www.gov.uk/government/publications/the-government-efficiency-framework/the-government-efficiency-framework--2 ↩
Miles Underwood and Steven Kalma, “Outcome-Based Contracting in IT: finding the Holy Grail?”, Compact, 2022, no. 2, pages 49 to 51. The discussion of supplier control and revenue risk appears on page 50. Article: https://www.compact.nl/articles/outcome-based-contracting-in-it-finding-the-holy-grail/ . Publisher PDF: https://www.compact.nl/pdf/C-2022-2-Underwood.pdf ↩