Productised services are one of those concepts that gets talked about constantly in agency circles and understood very shallowly. Most agencies that claim to have "productised" something have really just written a better pitch deck for existing services, slapped a name on a recurring engagement, or created a process document they can reuse. That isn't productisation, it's standardisation with marketing copy. Real productisation is structurally different, and the agencies that have actually done it have built something closer to a software company's product than to a traditional agency offering.
The starting point is understanding what a product is, in the strict sense. A product is a thing with a clear and defined scope, where the buyer knows what they're getting before they buy, where the cost to produce the next unit is significantly lower than the cost to produce the first unit, and where the thing itself exists somewhat independently of the specific people delivering it. A car is a product. A piece of software is a product. A specific methodology you've codified, packaged, and can apply repeatably with predictable outcomes is a product. A custom strategy engagement that your senior partner happens to deliver in a similar-looking way each time is not a product, it's a service pattern.
The distinction matters because the economics of products and services are fundamentally different, and productisation is interesting to agencies precisely because product economics are so much better than service economics. A traditional service business has a hard ceiling on margin: every new piece of revenue requires new hours from new people, so growing revenue means growing headcount roughly proportionally, which means margins stay roughly constant. A product business doesn't have that ceiling: once the product is built, selling it to the second customer costs a fraction of what it cost to build for the first one, and the third customer costs less than that. Margins expand with scale instead of staying flat. This is why software companies trade at much higher multiples than agencies, and why agency leadership keeps looking longingly at the product model even though very few have actually cracked it.
The agencies that have genuinely productised something usually start by recognising that they have a specific capability that clients want repeatedly, that can be meaningfully standardised without destroying its value, and that the agency has enough depth in to make the standardisation worthwhile. This is harder than it sounds because most agency work fails at least one of those three tests. Brand strategy is high-value but too contextual to standardise. Media buying is standardisable but has been commoditised by the platforms themselves. Creative ideation is valuable and distinctive but resists productisation almost by definition. The sweet spots tend to be narrower and more specific: a particular kind of research methodology that produces consistent outputs, a particular kind of creative system that can be applied to different brands within a category, a particular kind of measurement framework that answers a question clients keep asking, a particular kind of operational capability that plugs into clients' existing workflows.
A few recognisable shapes of productised agency services are worth understanding.
The first and most common is the methodology product: the agency codifies a specific way of doing a specific kind of work, trains people to execute it consistently, and sells it as a named offering with defined inputs, outputs, timeline, and price. The classic example is the branding sprint, five days, fixed price, specific deliverables, specific process, but the pattern applies to many things: onboarding audits, category analyses, positioning frameworks, naming processes, customer journey mapping, AI readiness assessments. The product isn't the software; it's the methodology itself, made repeatable enough that different people in the agency can deliver it with consistent quality. The economic benefit is that the first time you do it, you develop the methodology; the tenth time, you're just executing it, which takes much less senior time. The margin expands as the team gets more efficient at running the same play.
The trick with methodology products is that they only work when the methodology is actually valuable and distinctive. Most methodology products are some version of "here's our process, it has four steps, each step is named", which is not a product, it's a slide. A real methodology product has accumulated enough institutional knowledge that someone trying to replicate it from scratch would find it significantly harder than buying it from you. Either because it's built on years of iteration, or because it requires specific expertise the client can't easily hire, or because it incorporates data or IP the agency has developed over multiple engagements. Without that depth, the methodology product is just a service with a better name, and clients eventually realise they can get the same thing cheaper somewhere else.
The second shape is the platform or tool product: the agency builds actual software, a tool, a platform, a dashboard, a system, and sells it to clients as a product with its own subscription or licensing model. This is what most people think of when they hear "agency productisation," and it's the hardest to do well. Building software is expensive, maintaining it is expensive, selling it requires different skills than agency business development, and supporting it requires a customer success function agencies don't naturally have. The agencies that have made this work have usually committed real capital (sometimes tens of millions of dollars) to building the thing, treated it as a long-term R&D investment rather than a side project, and either stood up a separate business unit with its own P&L or spun the product out entirely. R/GA's various product attempts over the years, AKQA's work in this space, Huge's tooling efforts, and various holding company venture plays are examples, with mixed results. The ones that worked usually worked because leadership made the unglamorous investments in ongoing engineering, customer success, and sales capability that product businesses require. The ones that failed usually failed because the agency tried to build a product with part-time engineers borrowed from client work, treated sales as account managers selling extras, and expected the product to run itself after launch.
The third shape is the data product: the agency has accumulated data or insights from working across many clients or many markets, and packages that data into a sellable offering. Research firms have done this forever, syndicated studies, tracking reports, industry benchmarks, and agencies are increasingly doing versions of it. The product might be a regular report on what's working in a specific category, a benchmark dashboard that compares a client's performance against anonymous industry averages, a trend tracking service that spots emerging patterns before they hit mainstream awareness. The advantage is that the data compounds, the more clients use it, the more data the agency has, the better the product gets, which is a genuine network effect. The challenge is that the agency has to be disciplined about data rights, privacy, and anonymisation, and has to resist the temptation to let the data product cannibalise the custom research work that pays the bills. The agencies that do this well are careful about which work feeds the product and which stays bespoke, and they negotiate data usage rights explicitly with clients upfront rather than hoping nobody notices what's being aggregated.
The fourth shape, increasingly common and interesting, is the operational service product: the agency builds a capability to run a specific operational function for clients as an ongoing service, with clear pricing and scope. This isn't a methodology (which is typically project-shaped), and it isn't software (though software often enables it). It's a productised operation. Think: a personalisation engine that a brand can subscribe to, where the agency runs the whole thing and charges a monthly fee scaled to usage. Or a creator marketing program that comes with defined inputs, outputs, and pricing tiers. Or an always-on creative production pipeline that delivers a set number of assets per month for a fixed monthly fee. The product is the operational capability itself, and the client is essentially outsourcing an ongoing function to the agency on product-like terms rather than traditional agency retainer terms. This fits particularly well with the always-on modular delivery model — the operational backbone that makes always-on work can itself become a productised offering that multiple clients subscribe to.
The fifth shape is the licensed IP product: the agency develops some piece of intellectual property, a framework, a creative asset, a sonic identity system, a set of templates, a playbook, and licenses it to clients either as part of an engagement or as a standalone transaction. This is less common but increasingly viable, especially as the work becomes more systematised. The challenge is that most agencies don't think of themselves as IP owners and don't have the legal infrastructure, pricing models, or commercial conversations to license things well. The ones that have figured it out tend to have specific high-value IP that's genuinely differentiated and has been built up over years.
A few honest observations about what separates the agencies that have genuinely productised from the ones who just claim they have.
Real productisation means saying no. A productised service has a defined scope. When a client asks for something outside that scope, the answer has to be either "that's not what this is, here's the custom engagement that could do it" or "sure, but it's a different product with different pricing." Most agencies can't say no to client requests, their entire business is built around being responsive and flexible, so their "productised" offerings get stretched and customised until they're indistinguishable from the bespoke work they were supposed to replace. The agencies that have actually productised something have the discipline to protect the product's scope even when it means turning down revenue, because they understand that the economics only work if the product stays standardised. Agencies that let the product get customised for every client are running bespoke work with a product label, which has all the costs of custom work without any of the product benefits.
Productisation requires different talent than custom work. The skills that make someone great at bespoke, high-context agency work, sensitivity to nuance, ability to improvise, deep relationship building, creative judgment, are partially opposed to the skills that make a product successful: standardisation, process discipline, reluctance to customise, focus on scalable systems. The agencies that have productised well have usually brought in different kinds of people to run the product side: product managers, operations people, sometimes engineers, sometimes people from SaaS backgrounds. Trying to have the same team run both the custom work and the product usually ends with the product getting neglected because the custom work is more urgent and more familiar. The agencies that have succeeded have almost always had to create a separate function, with separate leadership, to protect the product from the gravity of the service business around it.
The pricing is often harder than the production. A product should be priced based on its value to the buyer, not its cost to produce, that's what gives products their margin advantage. But agencies are trained to price based on cost (hours × rate + margin), and the muscle memory is hard to shake. The agencies that have priced their products well tend to have gotten comfortable with the idea that a product that takes them two days to deliver might be worth $50,000 to the client, and that charging $50,000 for two days of work isn't gouging, it's capturing value that took years to build the capability to deliver. The agencies that can't get past the hours-times-rate instinct usually price their products too low, undervalue the work, and leave the margin on the table that was supposed to be the whole point.
Most productisation attempts fail on the sales motion. A product has a different sales conversation than a service. Services are sold through relationship and customisation ("tell us about your problem and we'll figure out the right approach"); products are sold through clear scope and predictable outcomes ("here's exactly what you get, here's what it costs, here's how it compares to the alternatives"). These are different conversations with different skills, and most agency business development people are trained for the first kind. When the same BD team tries to sell products, they default back to the service conversation, listening, customising, proposing a bespoke version of the product, which defeats the whole purpose. The agencies that sell products well either train their BD team in product sales or bring in different people to handle product conversations.
The product-service tension is permanent. Even the agencies that have productised successfully usually still do custom work, and the tension between the two parts of the business never fully resolves. The custom side sees the product as a distraction that takes attention and investment away from the relationships that drive most of the revenue. The product side sees the custom work as a threat that constantly tries to pull the product out of its scope and eat its margin. Managing this tension is a leadership job, not a process question, and the agencies that handle it well have leadership that is genuinely committed to the dual model rather than quietly hoping one side will win over the other. The agencies that handle it badly let one side dominate, which either kills the product (service takes over) or hollows out the custom practice (product takes over and the senior talent leaves).
The scale economics only kick in after real investment. The productisation pitch is usually "we build the product once, then sell it many times, and the margins are great." What's rarely mentioned is that the upfront investment to build the product properly, real product design, real engineering if needed, real operational infrastructure, real sales enablement, real marketing, is often in the millions of dollars and takes years to recoup. The agencies that have tried to productise on a shoestring, using spare time from client work, have almost all failed. The ones that have made real investments have sometimes succeeded, but the payback period is usually longer than the leadership that initiated the investment is willing to wait. This is one of the reasons holding-company-owned agencies find productisation particularly hard: the quarterly reporting cycle punishes long-payback investments, and the pressure to hit this year's margin targets makes it very difficult to sustain the kind of patient capital that product businesses require.
Not every agency should productise. The industry discussion often implies that productisation is a goal every agency should pursue, but the honest answer is that it's only the right move for some agencies, and only for some parts of their offering. An agency whose core value is senior creative judgment, whose best work is deeply customised to each client's situation, and whose relationships are built on a specific partner's trust is probably not well-suited to productisation, trying to force it will dilute the thing that made the agency valuable in the first place. An agency whose work has become more systematised, whose outputs can be meaningfully standardised, whose best capabilities could be extracted from specific individuals into repeatable processes, is a better candidate. Knowing which kind of agency you are is a strategic question, not a capability question, and getting it wrong is expensive in both directions.
The deeper strategic point about productised services is that they represent a specific answer to the labor-cost squeeze that's affecting the whole industry. If your business model depends on selling senior time at high hourly rates, and senior time is getting more expensive while clients are resisting rate increases, you have a structural problem that will eventually catch up with you. Productisation is one way, not the only way, but a real one, to decouple revenue from hours and build a business where the marginal cost of serving the next client is significantly lower than the revenue that client brings in. The agencies that have done this successfully are often quietly more profitable than their custom-only competitors and have meaningfully different valuation multiples in any M&A conversation. The agencies that haven't are either making the business work through exceptional custom craft at premium prices (which is viable for a small number of genuinely elite firms) or slowly watching their margins compress as the labor arbitrage gets harder to sustain.
The broader lesson is that productisation isn't really about products, it's about building something that scales independently of the people delivering it, which is the fundamental move from service economics to product economics. That move is hard, it requires patience and investment most agencies don't have, and it fails more often than it succeeds. But for the agencies that get it right, it's one of the few paths that leads to a fundamentally different kind of business than the one the agency started out as, and the ones that have taken it are often the ones that have turned out to be most resilient as the industry keeps restructuring underneath them.
