The cost threshold where an in-house office starts to make sense
Staff salaries, office costs, and the overhead of building your own vetting and inspection capability from scratch mean an in-house presence generally only becomes cost-competitive with an agent's fee at fairly high, consistent order volumes — well above what most mid-sized buyers are placing annually.
What you give up by going in-house
An agent's existing factory network, accumulated negotiating leverage across many clients, and established inspection processes take years to replicate — a new in-house office starts from zero on all three, even with experienced local hires.
A middle path: a dedicated account with an agent
Many agents, as a client's volume grows, offer a dedicated account manager or team essentially functioning as an extension of the buyer's own operation — capturing much of the in-house office's responsiveness without the fixed overhead cost.
When in-house genuinely wins
For buyers at very large, consistent volume — effectively running a full-time sourcing operation — an in-house office can eventually offer more direct control and lower marginal cost per order than an agent relationship, but this threshold is higher than most buyers expect.
What the fixed cost actually consists of
An in-house office is not one salary. It is the salary of a sourcing manager who knows the factories, the salary of someone who can inspect, and both of their employer costs, plus office rent, a travel budget, and the time it takes before either person is genuinely productive. Expect that last component to be the largest: someone new to the region needs a year of walking factory floors before their judgement is worth having, and they will be reassessing supplier reliability from scratch while doing it. None of this is unarguable — it is simply the full number, and it is the number that should be compared against the agent's fee.
What is genuinely hard to outsource
Some things really do need to sit inside the business. Handling a supplier dispute requires ongoing context that only accumulates over years. Knowing that a factory's pricing is soft at certain times of year, that a particular supervisor handles a process better than anyone else, or that a material will behave badly in a coastal climate — that knowledge is the asset an in-house office is for, and it is real. The question is whether the business needs that depth yet. A buyer running two categories and four suppliers a year is unlikely to be getting enough benefit from it to justify the fixed cost.
How to move the decision forward
The most useful test is to look at the cost of being wrong in either direction. Underusing an agent costs a percentage on orders you were running anyway, which is measurable and modest. Underinvesting in your own capability while your volume grows costs the price of mistakes you did not catch — and that is typically larger and harder to see. That asymmetry usually favours the agent for longer than buyers expect, which is why the volume threshold for going in-house is consistently higher than the first instinct suggests.
A sensible path is to move to a dedicated account structure as volume grows, where the agent assigns named people to your orders and the arrangement starts behaving much like an outsourced department. Reserve the fully in-house decision until you have enough predictable volume that someone could be employed full-time purely on your account, and until the categories involved need more local knowledge than a fee can buy.