Asset-Based Consulting: Why Advisory Firms Are Building Software and Managed Services

The economics of advisory work are being rewritten from the buyer's side. Forrester states that AI-powered delivery models drive labour productivity in some services by 50% or more, and concludes that traditional time-and-materials pricing models no longer make sense. That is an analyst's view, not a market-wide measurement, but it explains the direction of travel in procurement. When a client can see that the same deliverable takes half the effort, it asks why it should pay for the old effort.

Market growth gives little cover. A research report cited by Bloomberg Law found that the consulting market grew by just 2.9% in 2024, and expected 6% growth in 2025, to nearly $110 billion. A market growing at a low single-digit rate cannot absorb a large drop in the revenue per task, so firms that rely on volume of hours face a squeeze from both sides: lower realisation per hour and slower overall demand. Revenue per head may fall while costs of talent remain high.

The implication for a chief executive is that price is no longer the only variable at risk; the unit of sale is at risk. If clients buy hours, productivity gains accrue to them. If clients buy an outcome, a managed service or a licence, the productivity gains accrue to the provider, at least until competitors catch up. That is the core of the asset-based thesis: it is a way of keeping the benefit of the efficiency that AI creates, instead of handing it over through lower fees.

A first practical step is to measure where your own revenue is exposed. Classify engagements by how much of the work is repeatable, how much is judgement, and how much depends on proprietary information. Repeatable work with a measurable output is the most exposed to price compression, and also the most suitable for productisation. Judgement-heavy work is safer, but it is also harder to scale. The portfolio view tells you where to build and where to defend.

Pricing power also depends on who the buyer is. Large corporates with procurement functions are quick to ask for AI-related reductions, while owner-managed businesses often value speed and certainty more than the last unit of price. A firm can use that difference: sell fixed-scope, fast, productised offers to the second group, and reserve bespoke work for clients who value judgement. Segmenting the client base by buying behaviour is a cheaper first move than any technology build.

The most visible evidence sits in Accenture's filings. For fiscal 2025 the company reported revenue of $69.67 billion, of which Consulting was $35.11 billion and Managed Services $34.57 billion. Three years earlier the gap was far wider: fiscal 2023 Consulting revenues were $33.61 billion against $30.50 billion for Managed Services, and fiscal 2024 figures were $33.20 billion and $31.70 billion. Consulting was roughly flat across the period while managed services grew steadily.

Bookings point the same way. Accenture's fiscal 2025 new bookings were $80.62 billion, of which Managed Services accounted for $42.98 billion and Consulting for $37.64 billion. Bookings lead revenue, so a larger managed-services share suggests that the revenue mix will tilt further. The company also reported generative AI new bookings of $5.9 billion for fiscal 2025, after $3 billion in fiscal 2024, showing that AI work is being sold through both channels.

Other large firms report similar emphasis. KPMG's fiscal 2025 revenue was $39.8 billion; its advisory business grew 2.9% with continued demand for managed services, while tax and legal services grew 7.5%, driven by client demand for an AI-enabled managed service. PwC reported that Advisory revenue rose 4.5% to US$24.3 billion and that it continued to invest nearly US$1.5 billion across its network in AI capabilities. EY Global reported consulting growth of 5.2 per cent. These figures are each firm's own reporting of differently defined service lines and currency bases, so we do not compare them like for like.

The lesson for a mid-sized firm is not that it should copy Accenture. Accenture has global delivery centres and balance-sheet capacity that most firms lack. The lesson is that the market leader has already moved a large share of its business to recurring, contracted services, and that clients expect this model. Smaller firms can respond with narrower, deeper managed offers in niches where they have expertise and data, where scale is less important than quality and trust.

Scale matters in a different way for firms of different sizes. The largest groups can run global delivery networks that make managed services profitable at volume. A smaller firm will not match that, but it can compete on specialism. A managed service that monitors one regulatory regime, one sector benchmark or one operational process for a defined type of client can be delivered by a compact team with good tooling, and it can command a premium because it is deeper than a generalist alternative.

Contract data from ISG show a market in which different forms of sourcing move in different directions. In the first quarter of 2026, combined global annual contract value reached a record $39.4 billion, up 29% year over year. Within it, the as-a-service market soared 44% to a record $28.2 billion, with infrastructure as a service up 57% to $23.1 billion and software as a service up 5% to $5.1 billion. These figures describe technology spending rather than consulting fees, but they show where money is moving.

Managed services were far more modest. ISG recorded managed services annual contract value of $11.2 billion, up 3% on the prior year, while IT outsourcing slumped 7% to $7.9 billion. Traditional outsourcing, built on labour capacity, is losing share to platform-based consumption. Buyers appear willing to pay for scalable, metered services, and less willing to commit to long contracts for people. An advisory firm that wants recurring revenue should note that the recurring revenue is flowing to platforms.

This is a vendor's index and it measures contract value, not profit, so it should be read as a directional signal. ISG's data cover technology and sourcing markets, not the advisory sector alone, and a quarter is a short window. But the pattern is consistent with the Accenture and Forrester evidence: buyers are favouring services that scale without a matching increase in headcount, and they are seeking pricing that tracks outputs.

For the chief executive, the implication is to design offers that look like the growing categories. A managed service with defined service levels, a subscription to a data product, or a platform that clients use themselves are closer to the as-a-service model than a team of advisers on a time sheet. They do not need to be technologically elaborate. They need to be repeatable, measurable and priced so that improvements in your productivity increase your margin, not your client's savings.

One further caution concerns recurring revenue itself. Subscriptions are attractive because they are predictable, but only if clients renew. Renewal depends on continued value, which requires product improvement, customer support and evidence of results. Firms accustomed to project selling often underestimate the effort of keeping a subscriber. Build renewal metrics from the beginning, and treat churn as a signal about the product, not about the sales team.

BCG reported 2025 revenue of $14.4 billion, up from $13.5 billion in 2024, with AI- and tech-focused services now over 40% of revenue, driven by 25% year-over-year growth in AI services. Deloitte reported FY2025 revenue of US$70.5 billion, up 5%, and said it has allocated over US$3 billion to generative AI investments through FY2030. These are firm-reported figures, and the definitions of AI-focused work are the firms' own, but the scale of the commitment is clear.

Investment on this scale is a statement about where the firms expect margin. They are building platforms, data assets and agent-based tools, and they are redirecting their people towards them. Such investment also raises the bar for smaller competitors. A firm cannot match a multi-billion-dollar programme, nor should it try. Its advantage lies in speed, focus and closeness to a defined segment, which allows it to build a narrow asset that the giants will not prioritise.

Some reporting suggests that pricing models are changing at the same time, with a rising share of fees tied to outcomes. Those claims come from secondary summaries of press reporting and firm characterisations, not from audited disclosures, and we therefore do not rely on specific percentages. The direction, however, is consistent with Forrester's view and with buyers' demands. Procurement teams are asking providers to share risk, and firms with an owned asset are better placed to accept that risk.

The implication is that a firm should decide where on the spectrum of risk-sharing it is willing to operate. A fixed fee for a defined deliverable is the simplest step. A subscription for a platform with a service level is the next. Outcome-linked fees, where part of the price depends on a measured result, carry the greatest reward and the greatest risk. Each step requires better measurement, stronger contracts and a balance sheet that can absorb variability. Move one step at a time.

The speed of change argues for a flexible stance. Firms should avoid locking in large, long-dated commitments to a single technology or vendor when the tools are evolving quickly. Prefer architectures that can swap models and components, keep proprietary value in data, workflow and domain logic, and measure results frequently. That way, improvements in underlying AI capability strengthen your product instead of making it obsolete.

Many firms have tried to productise and found that a tool that works for one client does not work for ten. In our view four conditions separate the successes from the false starts. First, the service must be repeatable: a stable problem with a similar solution across clients. Second, there must be an owned asset, such as code, data or a method embedded in software, that does real work. Third, the outcome must be measurable by both parties. Fourth, the firm must be willing to change how it prices and pays its people.

The first test eliminates a good deal of work. Bespoke strategy for a unique situation cannot be productised, and attempting to do so produces thin generic content. Recurring problems can: benchmarking, compliance monitoring, diligence processes, procurement analysis, pricing reviews and reporting. In those areas, the expertise of experienced practitioners can be encoded into rules, templates and models, and AI can take on much of the routine analysis. The resulting service is delivered faster and more consistently than a bespoke engagement.

The second test is where many professional-services firms fall short, because a deck is not an asset. An asset must be something that continues to generate value without the same people redoing the work. A proprietary dataset that improves with each client, a platform clients log into, or an automation that runs a process on a schedule are assets. A slide library is not. Ask of any proposed product: if the author left tomorrow, would it still deliver?

The third and fourth tests are about commercial design. Define the outcome in terms the client already tracks, such as days saved, errors removed or cost avoided, and agree how it will be measured before the contract starts. Then align internal incentives, since partners paid on hours billed will resist a product that reduces hours. Changing partner compensation is the hardest part of the shift, and firms that avoid it usually find that the product loses to the old model inside the firm.

A final test is whether clients would pay for the asset on its own. If the answer is only yes when bundled with senior advisers, it is a tool for your team, which is useful but is not a product. Ask three or fthe firm's clients what they would pay for access, what they would need to trust it and what would make them stop. Their answers are the cheapest market research available and often change the design before money is spent.

Building assets costs money before it earns it. Software needs engineers, product managers and security, none of which are inexpensive, and data products need sourcing, cleaning and licences. Early revenue is small and uneven, while the costs are fixed. The largest firms can fund this from earnings. A mid-sized firm needs a plan for how it will pay for the first two years, whether from retained profit, debt or outside investment, and how much it is prepared to lose before it stops.

Culture is the second risk. Consulting cultures prize bespoke work, individual brilliance and client intimacy, while product cultures prize standardisation, roadmaps and release discipline. These are not easy to combine. Product teams sometimes feel like second-class citizens in a partnership, and advisers may resent being asked to feed a product with their experience. Leaders must give the product team authority, its own metrics and a protected budget, and must signal that the product is a strategic priority and not a hobby.

Cannibalisation is the third. A product that automates a service reduces the hours that service generated. The temptation is to protect the old revenue by under-investing in the new. The right response is to cannibalise deliberately, on your own terms, because a competitor or a client's in-house team will do it otherwise. Model the revenue trade: how many hours fall away, what the product earns in their place and how much margin improves. That analysis makes the case to the partnership and sets expectations about timing.

There are also quality and liability questions. A tool that makes errors at scale can create larger exposures than an individual's mistake, so testing, human review and clear limits on use are essential. Contracts should state what the product does and does not do, and professional indemnity and cyber cover should be checked. Data protection obligations rise when the firm holds client data in its own systems. None of this is a reason to avoid products, but it is a reason to treat them as engineering with controls.

Talent strategy needs to change as the asset grows. Product management, data engineering, design and customer success are unfamiliar functions in many advisory firms, and the people who do them expect different career paths from partnership track advisers. Decide early whether to hire, partner or acquire these capabilities, and give them real standing. Firms that treat technical staff as service providers to the partners rarely attract the people they need to build something durable.

The most reliable path is narrow. Choose one service that clients buy repeatedly, where your firm has real expertise and some proprietary data, and where outcomes can be measured. Build the smallest version that delivers value: it can be a managed service run by a small team with tooling, rather than a fully developed platform. Sell it to a handful of existing clients on a pilot basis, measure results, and improve it. Resist the urge to launch a suite of products at once.

Then decide how to price it. Early on, a fixed fee or subscription tied to a clear scope is easier for both sides than an outcome-linked fee. As measurement improves, a portion of the price can become variable. Keep a record of the effort actually required, so you can see whether the margin is improving as the tooling matures. If the product needs as many hours as the old service after six months, it is not yet a product, and the pricing should not pretend otherwise.

Govern it as a business. Appoint a product owner with accountability for revenue, cost and client satisfaction, set quarterly milestones, and agree stop criteria in advance. Report product performance separately from the consulting practice, so that it is neither subsidised invisibly nor judged by the wrong yardstick. Review the portfolio of ideas annually and be willing to close those that are not working. A discipline of stopping is as valuable as a discipline of starting.

Finally, tell the market what you are doing, and prove it. Publish evidence of results in the form of measured outcomes agreed with clients, not claims. Invest in the commercial infrastructure for recurring revenue: renewals, customer success and clear terms. A firm that moves steadily, with one asset at a time, will build a business that compounds, whereas one that pursues a dozen simultaneous initiatives will usually end up with a collection of prototypes. The framework and steps in this paper are designed to support that discipline.

It is also sensible to write down what success looks like at twelve and thirty-six months. At twelve months the test may be a handful of paying clients, a measured improvement in delivery effort and a clear view of unit economics. At thirty-six months it may be a meaningful share of revenue from recurring sources and a product that sells without senior partners in the room. Without such markers, product efforts drift, and enthusiasm fades when the first difficult quarter arrives.

Karter McKinley's AI & Technology implementation work supports leadership teams in identifying the services worth productising, building the first managed service or tool, and redesigning pricing and partner incentives around it. We work from your own engagement data and capacity, and we do not assume a particular platform or vendor.