It starts innocently. Somebody on the team is decent with the software, a client asks for a visual, and that person produces one over a weekend. It goes well. The next project asks for three.
Two years later that person spends half their week on renderings, is no longer doing the work they were hired for, and has quietly become a single point of failure. When they take a holiday, the visuals stop. When they leave, the firm discovers that nobody documented how any of it worked.
This is the most common way small and mid sized design practices end up with a visualization problem. Not a decision, a drift.
The real cost of the in house version
Firms usually compare an hourly rate against an external quote and conclude that in house is cheaper. The comparison is missing several things.
The first is the opportunity cost. An engineer or architect producing images is not producing the work that gets billed at a professional rate. In firms I have seen do the arithmetic properly, this is usually the largest number and the one nobody tracks.
The second is hardware and software. Workstations capable of production rendering, the licences, the asset libraries and the render time all sit in an overhead line that never gets attributed to the visuals.
The third is the failure mode. One person holding a skill that the practice sells is a risk, and it is a risk that materialises at the worst moment, which is usually the week of a competition deadline.
When keeping it in house is right
There are good reasons, and they are about iteration rather than cost.
If the images are part of the design process rather than a deliverable at the end, in house wins. A designer who can test a massing option at four in the afternoon and look at it before leaving is doing something an external studio cannot match on turnaround.
If the work is highly repetitive and technical, such as the same product family rendered in variants, a small in house setup pays for itself.
And if the practice sells visualization as a service in its own right, obviously.
When it stops making sense
The tipping point is usually a capacity problem rather than a quality one.
Competition weeks are where it shows. Everything is due at once, the one person who can render is also the person who should be resolving the facade, and the firm is choosing between a good scheme and good images of it.
The second trigger is a client asking for something outside the in house skill set. An animation, a VR walkthrough, a verified view for a planning submission. Each of those is a different discipline with its own tooling, and building any of them internally for one project is a bad trade.
The handoff is where teams lose the benefit
Handing it over badly produces worse results than doing it badly in house, and the failure is almost always in the brief.
A model sent without a description of what matters comes back with the wrong emphasis. The studio has to guess which elevation is the hero, which material the client is precious about, and what the image is for. Guessing takes iterations, and iterations are where the schedule goes.
The teams that get value out of 3D architect outsourcing tend to have a repeatable brief: the model in an agreed format, a marked up plan showing camera positions, a note on the time of day and the season, the material schedule with actual product references, and a statement of what the image is being used for.
Planning, investor, marketing and construction images are different products and saying which one you want removes most of the back and forth.
The three things worth agreeing up front
Who owns the model. If the studio builds from your files, be clear about what comes back and in what format. This matters two years later when the client wants twelve more views.
How many revision rounds are included. Unlimited revisions do not exist. Two or three named rounds with a defined scope is normal and manageable.
What happens when the design changes. It will. A rate for updating an existing model is cheaper to agree before the change than after.
The quiet effect on hiring
There is a staffing consequence worth naming.
Practices that outsource the production work tend to hire differently. They stop looking for the hybrid designer who is also a renderer, which is a rare and expensive combination, and hire for the discipline they actually need, with a coordinator who can write a brief and review what comes back.
That second role is underrated. A person who can look at a returned image and say the mullion spacing is wrong on the third bay saves more time than any amount of software skill, and that person is usually already in the office.
What a good external studio asks you
There is a quick test for whether the studio you are talking to knows the work.
A weak one asks for the model and a deadline. A strong one asks what the image is for, who signs it off, whether the design is frozen, which elevation matters, what the client is sensitive about and whether there is a material schedule with real product references. Those questions are tedious to answer and they are the reason the first draft comes back close rather than wrong.
If nobody asks them, expect to spend the saved time in revisions.
The summary a principal can act on
Track the hours your team spends on visuals for one quarter, priced at the rate those people bill at. Compare to what the same output would cost externally.
If the numbers are close, keep it in house, because control and turnaround are worth something.
If the numbers are not close, and in most practices they are not, the decision is not really about renderings. It is about what you want your qualified staff doing on a Tuesday afternoon.
