Anybody planning an in-house agency starts with the same sum: what the agency costs, what an internal team costs, where the difference sits. The sum usually works out and still points the wrong way, because it does not describe the actual problem. The reason companies go in-house is rarely price. It is the speed at which a decision turns into a finished piece of work.
- An in-house agency buys proximity, it does not cut cost.
- The bottleneck is never production, it is the brief.
- What recurs belongs inside. What is rare and risky belongs outside.
- Without its own quality check, in-house gets faster and flatter.
Why companies build an in-house agency
The move to an in-house agency is well documented and older than it feels. The ANA has found that 82 per cent of its members now run an internal unit, up from 78 per cent in 2018 and 58 per cent in 2013. Almost nine in ten of those units report a growing workload. This is no longer a fashion, it is the normal case.
More interesting than the number is the reasoning behind it. In the companies I know, the trigger was almost never a budget discussion. It was a specific moment when a small change to an image took four days, although the work itself took ten minutes. Make that experience a few times and you want access. An in-house agency is first of all a way of buying that access.
The saving used to justify these projects rarely survives the arithmetic. An in-house agency is a fixed cost, while an external agency breathes with the volume. In a weak year that is a disadvantage, not an advantage. The case only holds when volume is reliably high and proximity to the product genuinely makes a difference. Both are worth testing before the first role is advertised.
What belongs inside and what does not
The dividing line for an in-house agency does not run between important and unimportant, it runs between recurring and one-off. Anything that arrives on a predictable rhythm and benefits from precise product knowledge is better placed inside. Anything that happens rarely, carries high risk and lives on a fresh pair of eyes belongs outside. The mapping is unglamorous and still rarely drawn cleanly, because it has an uncomfortable consequence: the more exciting work stays out.
Clear cases for inside
- Derivatives. Formats, languages and channel variants of an existing image.
- Seasonal rhythm. Everything that returns twice a year in the same structure.
- Product-close work. Copy and pictures that require knowledge of the material.
- Reaction. Anything that has to be out within a day.
Clear cases for outside
The campaign idea that carries a brand for two years rarely emerges from a team looking at the same products every day. Proximity is an advantage in execution and a burden in invention. The same holds for formats a company is attempting for the first time, where external routine is worth more than internal product knowledge. Pull both of those into an in-house agency and you overload it in its first season.
What comes from outside cannot be rebuilt inside
The cost and speed advantage of an in-house agency is real, and it is still not where the decision gets made. An internal team is faster and cheaper because it has templates, knows the approval path and does not start from scratch on every job. That holds for everything routine, and routine work is the larger share of the workload. It does not hold for the part that lives on a connection to the outside world.
In fashion this is sharper than elsewhere. Look at the same collection every day and you lose the feel for what actually needs explaining about it. An external agency or a good freelancer brings more than craft: they bring a different slice of the world, other clients, other scenes, other references. That slice is the real value, and it cannot be rebuilt internally no matter how strong the team is. Ideation is therefore the one task where I would deliberately not make an in-house agency the only source.
Alongside the classic agency there is a third model I consider underrated: the collective. Loose groups of freelancers who reassemble around each brief, sometimes around a director, sometimes around a photographer, sometimes around someone from an entirely different discipline. For companies that want neither a permanent agency retainer nor everything solved inside their own in-house agency, this is the most fluid option. It does demand more leadership, because the line-up does not hold together on its own.
- Inside. Recurring work, product-close work and anything that has to be out within a day.
- Agency. The idea that outlives a season, and formats the company is attempting for the first time.
- Collective. A line-up that changes per brief, when the impulse matters more than settled routine.
The three are not mutually exclusive. The mistake I have seen more than once is treating it as an either-or: a company builds an in-house agency and cancels its external relationships in the same move, because the budget now sits internally. A year later there is no access left to people who think differently, and that access cannot be restored at short notice. A small, permanently open budget for outside work is cheaper than trying to rebuild a network that has already gone.
The bottleneck sits before production
The most common misdiagnosis during setup is to locate the bottleneck in production. It nearly always sits earlier. An in-house agency produces faster than any external one, as soon as it knows what is meant. That is exactly where things jam, because internal briefs are worse than external ones. Facing an agency, the contract forces precision. Facing colleagues in your own building, you rely on them understanding it anyway.
That sounds like a detail and is the most expensive item in the whole model. A team going three rounds because the task was blurry is slower than an agency that delivers once. So I have made a habit of treating internal briefing discipline exactly as hard as external, written task and named decision-maker included. How to judge the result afterwards is something I set out under evaluating creative work.
The quality question nobody enjoys asking
An in-house agency carries a structural disadvantage that gets discussed reluctantly. It cannot be fired and it does not go back out to pitch. The corrective that competition creates around an external agency is missing entirely. When the internal team slips, you notice late, because nobody is drawing a comparison and because working together day to day is pleasant.
The substitute has to be built in deliberately, and before the unit reaches a certain size. Two things have worked for me: a fixed external reviewer who reads the work twice a year and has nothing to sell, and a regular exercise of laying our own output next to that of two or three competitors. Both are uncomfortable and both need cover from above, or they get cancelled the first time the calendar fills up.
What a staged build looks like
Anyone building an in-house agency should not start with a structure but with a single stream of work. The order that has proved itself is always the same.
- First. Bring one recurring stream with high volume and low risk inside.
- Second. Measure the time from brief to sign-off before anything else changes.
- Third. Only then add capacity, and add it where the measurement shows the jam.
- Fourth. Set up the quality check before the team grows past ten people.
The most common mistake when building an in-house agency is to begin at stage three. A team gets hired before the stream of work is defined, then spends six months looking for something to do. After that the discussion can no longer be held about value, only about utilisation.
An in-house agency also gets measured against the wrong number with some regularity. Count delivered assets and you get exactly that: plenty of assets. The more useful figure is the time from decision to publication, because it describes the reason the team exists at all. It is uncomfortable, because it also makes the waiting time inside approvals visible, and that is precisely where its value sits.
What I took away from it
My biggest misjudgment was treating an in-house agency as a question of organisation. It is a question of clarity on the client side. An internal team reflects the quality of internal leadership back faster and less filtered than any agency, because the contractual politeness that otherwise cushions things is absent. Lead vaguely and you see the result in the work within weeks.
What surprised me is the side effect on the remaining agency relationship. Once the routine work sits inside, the external collaboration improves, because it concentrates on the questions it was meant for. How I sort this in the wider picture is under brand leadership. More on this in the Brand Leadership & CMO section.