Agency vs. In-House: Who Should Run Your Affiliate Programme?

# Agency vs. In-House: Who Should Run Your Affiliate Programme?

"It depends" is the honest answer to this question, and it's also a useless one on its own. Every guide on this topic eventually lands there, lists a few pros and cons, and leaves the actual decision to you. This one is going to try to do better than that — with a real number to anchor the decision, not just a feeling.

The number that actually matters

Start here: how much revenue does your affiliate channel currently drive, in a typical month?

Practitioners who work across many programmes point to a rough threshold — somewhere around $80,000 a month in affiliate-attributed revenue is where in-house management starts to become the more cost-efficient option. Below that, an agency is usually cheaper and faster, because you're paying for a shared team's existing relationships and infrastructure instead of building your own from nothing. Above that, the cost of a dedicated in-house hire starts to make more sense against what the channel is actually generating.

That's not a rule to follow blindly. It's a starting point that turns a vague debate into a specific question you can actually answer about your own business.

What each option is actually buying you

An agency — sometimes called an OPM, outsourced program management — brings a team that already has publisher relationships, platform expertise, and a track record across other clients. You're not building anything from scratch. You're renting access to infrastructure and relationships that took someone else years to build. Retainers typically run somewhere between $3,000 and $15,000 a month depending on scope, usually plus a performance component.

An in-house hire gives you someone who works only on your programme, sits inside your team, knows your product and your customers without translation, and builds institutional knowledge that stays with the company. A full-time affiliate manager typically costs somewhere in the region of $80,000 a year before the tools, tracking software, and fraud detection systems they'll also need — costs an agency would otherwise already have built and shared across clients.

Here's the part worth sitting with: neither of these is better in the abstract. They're better at different things, for different-shaped businesses, at different stages.

Two real situations worth thinking through

A direct-to-consumer skincare brand once moved from an agency to an in-house model and immediately ran into a problem nobody had budgeted for: fraud detection. The agency had tooling and pattern-recognition built up across dozens of clients. The in-house team had to build that capability from a standing start, and spent months exposed before catching up.

The reverse pattern shows up often too. A brand starts on an agency retainer somewhere in the $2,000-$15,000-a-month range, and as the channel scales, the math flips — a full-time in-house hire, typically $70,000-$100,000 a year fully loaded, becomes the cheaper option relative to what the channel is now generating. That's when brands tend to make the switch, sometimes building out a full internal team as affiliate revenue keeps growing.

Same decision, opposite outcomes, both correct for the business at the time they made the call. That's really the whole point: this isn't a question with a universally right answer, it's a question with a right answer for where you are right now.

The questions worth asking yourself directly

Rather than a generic pros-and-cons list, ask these in order:

1. Do you know what "good" looks like for this channel yet? If you don't have enough internal expertise to tell whether a proposed commission structure or partner mix is actually smart, you're not ready to manage it alone — you need someone who already knows, at least for now. 2. Is your channel revenue predictable, or still finding its shape? A new or underperforming programme usually grows faster with an agency's existing relationships behind it. A mature, stable channel is where in-house control starts paying for itself. 3. What do you actually want more — control, or speed? If direct access to whoever runs your programme, same-day decisions, and total alignment with your brand voice matter most to you, that points in-house. If you want to move fast using relationships and tooling someone else already built, that points agency. 4. Could you tolerate single-person dependency risk? In-house often means one person, or a small team, holding all the institutional knowledge. If that person leaves, you feel it immediately. An agency spreads that risk across a team that doesn't disappear when one person changes jobs. 5. Do you actually have $80K+ a month in affiliate revenue, or are you assuming you'll get there? Be honest here. Building in-house infrastructure for a programme that isn't at scale yet is one of the most common ways this decision goes wrong.

Where most brands actually land in 2026

The honest current answer, based on how this plays out across most companies right now, is hybrid: an internal owner who understands the business and sets direction, paired with agency execution for the parts that benefit from existing infrastructure and relationships — tracking, partner recruitment, day-to-day optimization. Full agency outsourcing tends to work best early, when speed and existing relationships matter most. Full in-house tends to work best once the channel is mature, predictable, and big enough to justify dedicated headcount. Most brands spend real time somewhere in between, and that's not a failure to decide — it's usually the correct answer for that stage.

The actual point of this exercise

The question isn't really "agency or in-house." It's "what does my programme actually need right now, and am I honest enough about our current stage to admit it." A young, unpredictable programme pretending it's ready for a fully in-house build is setting itself up for the fraud-detection scramble the skincare brand went through. A mature, scaled programme still paying agency retainers out of habit is quietly leaving money on the table.

Where your own programme actually sits against these questions is worth working through directly, rather than guessing.