The Multi-Channel Law Firm
Legal work is splitting into productised and bespoke - how will firms respond?
For decades, law firms organised delivery around one primary model: bespoke legal services, marketed on expertise, delivered by highly trained professionals. That model still defines most firms’ identity, culture, and economics.
But the ground is shifting. Clients are asking about outcomes, not inputs. AI and automation are making it possible to systematise entire processes, not just discrete tasks. New entrants are launching AI law firms built entirely around productised delivery.
In my mind, the direction is clear: legal work is going to be routed, in a much more deliberate way, via distinct delivery channels based on shape, size, and client. High-stakes matters get senior judgment, expertise and bespoke delivery, while repeatable work gets speed, certainty, and predictable costs.
I think the net result is the rise of a new operating model – the “Multi-Channel Law Firm”.
The Two Primary Channels
Legal work is splitting into two core delivery modes. (There are probably more than two but these are the primary channels that define the model.)
The Productised Channel handles repeatable work as a defined offering. Fixed scope, fixed SLAs, quality designed into the process. Success measured by outcomes, turnaround, reliability. The value proposition is speed, transparency, predictable economics.
To be clear, productised doesn’t mean no humans. Most productised services in BigLaw will be process + people + technology. But they’re priced and structured like products, rather than bespoke engagements.
The Bespoke Channel handles higher-stakes, higher-judgment work. Differentiation comes from expertise, strategy, navigating ambiguity. Value comes from experience, pattern recognition, and judgment through uncertainty.
The bespoke channel isn’t disappearing. But if you believe more work gets automated over time (and it does – more on that below) then more is going to flow through the productised channel. That leaves firms with a decision about where to place their bets. And I think most will settle on a multi-channel model.
Why Now
Automation has been possible for a long time but two things are converging.
Clients are more deliberate buyers. A new generation of general counsel, backed by legal ops teams, are making clearer decisions about what to keep in-house and what to outsource. When they outsource, they come with a point of view about the “how” and the “outcome”. They’re asking: do I need your best judgment, or do I need this done to an acceptable standard, at scale, within 24 hours, at a fixed price, every time?
More legal work is structurally automatable. McKinsey, in a study of agentic AI across industries, estimated in November 2025 that 70% of legal work has automation potential (among the highest of any industry). That number will make lawyers skeptical and we’re of course very far from 70% being automated today. But something has changed. Earlier legal tech automated discrete tasks or individual workstreams, or was exclusively deterministic: document assembly, extraction, etc. The current wave is combining generative AI with workflow systems to automate entire processes end-to-end.
What’s happening:
Contract review services applying client playbooks at volume, with turnaround measured in hours, including human in the loop.
Due diligence workflows that route documents through AI review, flag exceptions, escalate only genuine judgment calls
Regulatory compliance programs that ingest rule changes, map them to client obligations, generate responses, package for sign-off
History tells us that in any industry, what can be automated eventually is automated. It starts with high-volume, low-complexity work but, over time, works its way up the value chain. There is no reason legal will be any different.
Where This Already Happened
There are clear signs of this shift in other parts of professional services already.
Consulting. McKinsey, BCG, Bain run partner-led strategy work alongside implementation arms built on standard playbooks. Accenture derives most revenue from managed services and platforms while keeping a high-end advisory layer. Clients seem to get it - they pay for judgement and design where it matters and expect productised delivery where it doesn’t.
Audit and accounting. The Big Four have long separated professional judgment from industrialised processing. They operate global delivery centres for audit prep, tax compliance, regulatory reporting. Standard methodologies, senior partners focused on risk and sign-off. No client expects partners to manually execute high-volume audit work. They expect systems and outcomes.
The New Competition
2025 has been defined by a new category of law firm – the “AI Law Firm” (or the “AI-First Law Firm”, or “NewMods”, depending on who you ask). I started building a running list here.
Crosby and Tacit offer contract review with rapid turnaround and humans in the loop. Avantia has a productized process for Funds. Landfall is trying to productise patent work. A new firm launches every week. These firms aren’t competing on breadth, headcount, expertise or directory rankings. Their focus is speed, throughput, transparency, cost – and outcomes.
Right now, they’re still relatively small. But they’re growing fast - proving productised legal delivery is viable and that clients will pay for it. They’re also showing clients that work can be delivered effectively through systematised channels - and that expectation doesn’t stay contained to these new entrants.
So, how does BigLaw respond?
Option 1, opt out
Some firms will choose to double down on bespoke work and let others handle the systematisable stuff. This doesn’t mean ignoring AI, but it means integrating it into existing processes and team structures, and focusing on being the best in the world for human judgment and expertise in a given practice area. This is simple, clean, and avoids some of the structural complexities of a multi-channel firm. But it’s also fraught with risk. Some of the work is going to go to other firms – maybe it’s less profitable work, maybe it isn’t. Either way, this strategy means giving up data/intelligence (on all those contracts, disputes, regulatory matters the firm didn’t handle), relationship touchpoints, and cross-sell opportunities.
Option 2, opt in
Doing this requires navigating a challenging path.
Accept Cannibalisation
This is the uncomfortable part – which is the very reason the winners will be the ones who tackle it now. Some productised offerings will replace work currently handled via bespoke delivery. A contract review that took 20 hours might take four in a productised channel, at half the revenue.
But avoiding this just hands the market to competitors who won’t hesitate. Better to cannibalise your own business – the old example about the iPhone cannibalising the iPod.
Perhaps the easiest win is to start by targeting new clients or adjacent work where there’s limited legacy billing to protect. Prove the model, prove the demand, then expand into existing relationships where you’ve built trust and can have honest conversations about value.
Decide Who Owns the Product & Sales
Productised delivery needs someone who owns the product - not the legal output but the end-to-end system that produces it. This person designs intake, defines quality gates, manages SLAs, tracks metrics, and continuously improves the system.
Is this the Partner? Or is it a Delivery Lead, Product Owner, Director of Legal Engineering, Chief Product Officer? I suspect some Legal Engineers will be Partners on the productised side of the business (and or course some will be Partners on the bespoke side too).
And who is in charge of selling these products? Law Firm BD has always been differentiated from “Sales” (in fact, most law firms regard “Sales” as a dirty word) but in this new world of selling products, things like outbound, inbound, lead qualification, pipelines, demos - one can envisage a much more sales-based approach, which may require bringing new skillsets into the firm.
Redesign Incentives
Compensation models that primarily reward bespoke hours don’t create the right incentives for productisation. Partners have little economic reason to route work through a productised channel if it means lower billings and higher relationship risk (because they are less directly in control of the work product), even if it’s better for the client and ultimately more profitable for the firm.
To execute on a productised strategy, firms will need to reward total client value and profitability. This means crediting partners for origination and relationship management even when delivery happens through productised channels, and building bonus structures around client retention, margin improvement, and channel adoption.
Move From Effort to Outcomes
Productised work gets sold and measured on deliverables, quality, and reliability rather than hours spent. This means rethinking pricing, engagement letters, risk sharing, billing systems, SLAs and client reporting. If we’re agreeing to a 24-hour turnaround time, we need a system to route work to the right agent or associate and enforce deadlines systematically. It also means pricing risk differently. In bespoke work, scope creep can sometimes benefit the firm. In productised work, it’s a system design failure that needs to be fixed in the next iteration. Expect to see those Engagement Letter precedents getting amended in 2026!
What are the next steps?
How do firms build out these new productised delivery capabilities? I see a few options:
Option 1: Repurpose the managed service unit. Many firms already have nearshore or offshore centres, primarily staffed with paralegals handling compliance, volume work (e.g., GDPR or leases), and doc review. These were built primarily as labour arbitrage plays - but much of the productised delivery infrastructure is in place. Repurposing a legal service unit would be about evolving it from supporting discrete projects to owning entire productised offerings. This may mean investing in the team with new skillsets, different metrics, different client-facing positioning – maybe even rebranding.
Option 2: Acquire an AI-first law firm. A handful of new entrants have built entire practices around systematised delivery (client intake, workflow automation, fixed pricing, the full stack). I’m aware of a couple of these AI-first law firms already having received investment from BigLaw. The next step would be for an AI-first law firm to be acquired by a BigLaw firm. After that, there’s a question about whether it continues to operate under an independent brand (keeping the main law firm brand for the bespoke work), or gets merged into the main brand.
Option 3: Set up a new unit under a separate brand. Some firms will want separation (different positioning, different pricing, potentially different regulatory structure). They might decide to build, rather than buy, and set up an entirely new “AI Legal Service Unit” or even an “AI Law Firm within a Law Firm”, operating under a separate brand. Maybe they will partner with other organisations on this – consulting, or software for example – to provide a complete business solution, of which legal is only one part.
More than two channels?
While there are two primary channels, I think these will further splinter, which is why I think of it as a “Multi-Channel Law Firm”.
For example, within the productised channel, look out for:
- Self-serve offerings
- Managed outcomes with SLAs
- Variations of the “Lawyers on Demand” model more akin to complete managed legal services – outsourced legal operations
Within the bespoke channel, there’s:
- Premium advisory
- Formal senior sign-off layered over systematised work
- Faster-response yet still with specialist support
- New services like system/workflow design
The Takeaway
I actually think all of this adds up to a huge opportunity for firms.
Yes, there may be difficult conversations about cannibalization, and what work we realistically want to keep – but there are also opportunities to go after whole new categories of work, and build entirely new businesses around it. At a time like this, the biggest risk is standing still.