FOXNUT AI

Comparison

Agency vs in-house: which one should you choose?

Build in-house when the function is permanent and close to how you win. Hire an agency when the need is bounded, urgent or needs senior judgment.

By the Foxnut team · Updated

Should you use an agency or build in-house?

Build in-house when the capability is permanent and close to how the company wins. Hire outside when the problem is bounded, ahead of what your team can learn in time, or needs senior judgment more than headcount. Most teams asking about agency vs in-house are really asking two questions at once, a staffing question and a timing question, and the comparison only helps when both sides are allowed to win.

We are Foxnut AI, a senior-only practice in Bengaluru and Paris, and we sell external work through the studio’s consulting work, which puts us on one side of this comparison. There’s a link to reach us at the end of this article. Let’s first work out which side you are actually on.

We have kept this page two-sided anyway, because nearly every version of it on the open web is written by an agency and resolves accordingly. The sections where in-house wins are not fine print, and the refusal section at the end names the situations where we are the wrong buy.

How do agency and in-house costs compare?

They compare on shape more than on price. A single price comparison flatters whichever side hides its overheads better, so the table compares the two on structure.

AxisAgency or externalIn-house
Cost shapeA fee you can stopSalary plus loading, management time and ramp - and you cannot stop it quickly
Speed to competenceDays to weeks; pattern brought from other companiesMonths of hiring plus months of ramp
Context depthThin at the start, and it resets if the team changesCompounds daily, and stays
Ceiling on seniorityHigh - you rent judgment you could not afford to employCapped by what the role can pay, full time
Continuity riskThe vendor’s churn becomes your problemOne resignation can zero the function
What you own after a yearWhatever the contract says you keepThe people, the process and everything they learned

When should you choose an agency or in-house?

Each side wins real situations and fails in a characteristic way. A version of this page that cannot name both failure modes has picked a side.

When should you build the function in-house?

Build in-house when the function is permanent, the volume is steady and owning the learning matters more than speed of arrival. In-house wins on context, on cost per unit at sustained volume, and on the compounding that happens when the same person watches the same numbers for two years.

Failure mode: the ramp. A bad hire in a function you cannot yet evaluate costs a year, and a single-person function concentrates everything in one resignation.

When should you hire an agency or outside help?

Hire outside when the need is faster than recruitment, spikier than a salary, or more senior than the role could pay full time. Outside help wins on pattern exposure, because it has seen your problem elsewhere, and on being something you can stop.

Failure mode: the structure. You are one account among many, the delivery team can change without your consent unless the contract forbids it, and at the end you may own a dependency and not a capability.

Does an agency actually care about your business?

Structurally, an agency cares about your business in proportion to what your account represents to it. Agencies are not cynical. Portfolio economics decide where attention goes when two clients need the same people in the same week.

That is not an argument against external help. It is an argument for reading the structure before you sign. Care you can verify looks like named people in the statement of work, a scope with a defined end, and a handover that leaves your team running the result. Care you cannot verify looks like enthusiasm in the sales conversation.

What do you lose by keeping it in-house?

You lose pattern exposure and calibration. An in-house team sees one company’s data forever, and has no way to know whether its conversion rate, its costs or its playbook are normal, good or quietly broken. An outside practitioner is calibrated by the market against many companies.

You also lose surge capacity and the ability to rent a level of seniority the role could never pay for permanently. The loss on the other side is just as real: buying outside means the learning walks out at the end of the engagement unless the handover is contractually real.

What does this comparison usually get wrong?

It is written by one side. Agency vs in-house pages are authored almost entirely by agencies, which is why the in-house column above is allowed to win rows and why this page discloses its author’s position twice.

One of our founders has sat on the other side. Swati Thakur built an in-house account-based marketing function from zero at a Paris B2B company before selling external work. Our view of what in-house is actually for, owning the compounding and not saving the fee, comes from that seat and not from a vendor’s.

Her answer to a commercial leader weighing her old role against outside help is that it depends entirely on the situation. A one-time need, a SOC 2 compliance audit for example, is a consultant’s brief, and hiring for it makes no sense. An industry with standing regulatory risk is the opposite case: build the in-house team, because the exposure never ends, and augment it with an expert consultant where the depth runs out.

The other standing error is comparing sticker prices. An agency fee against a salary is the wrong sum on both sides: the in-house number is missing its loading, management and ramp, and the agency number is missing the internal time an agency still consumes. The comparison that decides well is the one above: cost shape, ownership and continuity, priced honestly on both sides.

How do you decide which side you are on?

Use Swati’s rule: a bounded need with a fixed job to do is a consultant’s brief, and standing exposure that never ends is a hire.

If it is a brief, the usual worry about outside help is the junior team that turns up after the sale. We do not have one. Everyone who works with you is senior, we built and launched our own products before we advised anyone, and the work ends with your team running the result.

You can start with a free audit call and no pitch deck. If you are not sure which of those describes your situation, tell us what it is and we will say plainly whether it is ours to take or a signal to build the role instead.

The part most pages leave out

When not to choose Foxnut AI

Situations where another option is the better call, and where we say so in the first conversation rather than the fourth.

  • The capability is core to how the company wins and will be exercised every week for years. In-house. A permanent capability belongs on payroll; renting it indefinitely is the most expensive way to end up owning nothing.
  • The strategy is settled and what the team needs is hands: execution volume, not judgment. An in-house hire, or an agency retainer if the volume swings. An external senior studio is the wrong shape for pure execution and would charge judgment rates for production work.
  • The role needs someone in the room five days a week, absorbing the context nobody writes down. In-house, without qualification. No external engagement replaces a person who lives inside the company's daily context.
  • You want named references from your exact category before signing anything external. An established agency with public case studies, or an in-house hire whose track record you can check directly. Foxnut AI has no publishable named client case study yet, and we say so here and not in procurement.

Foxnut AI takes on work like this from Bengaluru and Paris. If you want to see how we do it before you talk to anyone, here is which engagement shape fits, and which does not.