Context, not proximity, decides whether in-house beats the old agency.
Bringing work in-house has an attractive savings argument. The business stops paying an agency's margin, and a brief no longer has to travel to another organisation. The people making the work can speak to the people running the business. Those are sensible reasons for the move. They leave another decision open: how the work will be organised once it arrives.
What follows the move receives far less attention. Within a year or two of launch, a significant number of in-house teams have rebuilt the structure they were created to replace: an intake form, a ticket queue, an internal client relationship managed by someone with an account-management title, a chargeback model, and a separation between the people who think and the people who make. The agency has been brought inside the building with its overheads intact and its main advantages left at the door.
The rebuild happens gradually, through a series of reasonable responses to rising demand.
Requests collide, so a form appears. Its submissions need sorting; a coordinator takes that on. Finance wants the costs assigned to departments. Requesters acquire the title of client, and the creative side acquires account management. The old agency shape has returned through a series of problems that seemed quite sensible to solve.
Eventually there is a client-agency boundary inside the building. The external agency's commercial discipline hasn't necessarily made the move with it.
Much of the bill accumulates in hours and in the work itself, which can make its cost difficult to see.
'Landing page, Friday' can fit neatly into the form. The reason for Friday, the last failed attempt and the stakeholder likely to object remain with the requester. Production recovers them through meetings and revisions, or supplies its own guesses. The job has arrived; much of its context is still somewhere else.
For illustration, allow four person-hours of clarification per request across 400 requests. The annual total is 1,600 person-hours. This is illustrative arithmetic, before revision time. It gives a way to examine the effort used to recover information already known when the job began.
More scope is something an external agency can price. Inside, refusing a senior stakeholder's additional request is harder, particularly when it has no visible cost to them. The new work waits beside these extra rounds. The team can lose the speed advantage it was brought inside to provide.
The external version of 'no' had a mechanism behind it. A scope document, a rate card and a change order turned an extra request into a conversation about trade-offs. Inside, the mechanism disappears and only the awkwardness remains. The discipline can return without the invoice. Publish the team's capacity for the quarter, show what each new request displaces, and let the stakeholder choose what moves. The cost is paid in visible trade-offs rather than chargebacks, so the conversation stays about the work.
Proximity doesn't automatically bring the makers into the thinking. Strategy remains with marketing leads and production receives requests. The business conversation may be taking place just down the corridor while the creative team learns about it only through the next form.
Look at how much of the coordinator's job is spent repairing missing context. Look at that work before appointing another. Three signals make it easy to see: the share of requests that come back for clarification, the time a job waits before a maker speaks to the person who asked, and the part of the coordinator's week spent chasing information instead of routing work. An agency structure once helped two separate businesses deal with each other; bringing the work inside deserves a fresh examination of what the boundary is for.
Inside the organisation, the starting point can be the problem colleagues share. Their borrowed scope discussions and forms may be separating people who need more of one another's knowledge. The external agency needed an arrangement for different business incentives. The internal team has a chance to design around its common objective.
A stable team included in planning can carry a brand or business problem across projects. It knows the previous request and what followed it. Deadlines remain, but the next brief no longer needs to introduce the business to the people expected to make the work.
Audience research belongs where the whole team can use it. Stakeholders also need a defined place in decisions, with a point at which the decision is settled. Otherwise, a continuing problem can still dissolve into an unlimited succession of comments and revisions.
Cost allocation, where finance needs it, is set at the level of the team and the objective and kept well away from individual requests. The moment each request carries its own price, the organisation has rebuilt the transactional relationship it set out to escape.
None of this makes every queue a mistake. Some work is high-volume, repeatable and well understood: resizing assets, localising an approved template, updating a price. Context adds little to a job like that, and a simple intake path serves it well. The trouble starts when the queue becomes the default route for everything, so that a problem needing judgement is handled like a banner resize.
A useful split sorts work by how much context it needs. Routine production can run through a light request path with a clear turnaround. Problems that touch positioning, a launch or a customer journey go to the stable team, with the people who own the outcome in the room from the start. Make both paths visible to stakeholders, so the quick route never becomes a back door into the slow one. Review the split every quarter, because yesterday's judgement call becomes tomorrow's template.
Bringing work in-house carries a second risk the rebuilt agency tries, clumsily, to solve. A team that works on one brand indefinitely can lose the outside perspective that an agency carried by default. Agency teams see several categories, several clients and several sets of failures each year. An in-house team sees one, in depth, and depth can harden into insularity.
A pitch-style presentation or awards target can imitate agency life without supplying its outside experience. Rotating staff between internal 'accounts' may leave them looking at the same business through a different window. Ask what perspective the arrangement actually adds.
Stability still needs an opening for fresh questions. Bring an outside specialist into a particular problem, or invite colleagues from another function to critique the work. Research beyond the category can expose another assumption. The familiar team keeps its accumulated context while giving new questions a way in.
A queue grows and demands another coordinator. Research, once shared, can serve the next brief as well as this one. That is an important difference in what the team gets for its effort. Learning can accumulate where the next project will be able to use it. A finding kept in a shared, searchable place informs the briefs that follow; the same finding in one person's slide deck leaves when they do.
Bringing work home changes the address. The hive still needs an arrangement that carries context to its makers. Follow one request through: which information arrived at intake, and which had to be recovered later? The next structural change can begin with what that journey reveals.
