Insights

In-house creative vs agency creative for eCommerce

Neither model on its own gets you to volume at a low cost per ad. The setup that actually scales is a briefed, coached in-house team producing at volume, with outside strategy driving the scripts and angles.

Matthew Feng · Published 7 August 2026 · 4 min read

Two models, two different failures

At $5M+ the question is not "who makes prettier ads." It is "how do we put enough distinct, on-brand concepts in front of the algorithm every week to keep finding winners." Both the agency route and the pure in-house route fail that test, but for opposite reasons.

Agency creative: high cost per ad, low volume

A creative agency produces beautiful work. The problem is the unit economics. When you pay per finished ad, the cost per creative climbs into the thousands, and that price forces scarcity. The familiar retainer is four ads a month, maybe eight. That is nowhere near enough to keep a growing account fed.

The scarcity has a second cost. When every ad is expensive, everyone gets attached to it. A precious ad is the enemy of fast learning, because testing is supposed to be mostly wrong. There is also a structural gap: the creative agency is not accountable for performance, and the media agency ignores the creative inputs that decide whether the ads can work at all. You pay both and still starve the account.

In-house creative: cheaper volume, but blind

An in-house team fixes the cost problem. Your own people can shoot and edit dozens of ads a month at a fraction of the agency price. Speed and brand consistency improve because the team lives inside the brand every day.

What in-house teams usually lack is strategy at the top of the funnel. Left alone they produce volume, but it is volume of the same idea: ten variants of one static, a handful of hooks the team already believes in. Platforms treat near-duplicates as one thing and they fatigue together. High output, narrow range, slow learning.

Cost per ad and volume are not a trade-off you have to accept. The winning model buys both by separating the thinking from the making.

Scoring them honestly

• Cost per ad. Agency: high. In-house: low. In-house wins.
• Volume. Agency: capped by the retainer. In-house: high if the team is set up for it. In-house wins.
• Brand consistency. Both can be strong, but in-house holds the brand more naturally day to day.
• Speed. Agency: slowed by rounds and invoices. In-house: fast once briefed. In-house wins.
• Strategy and angles. Agency: strong. In-house: usually the weak point. Agency wins.

Read that list back and the answer is obvious. You want in-house economics and speed, with agency-grade thinking on the scripts and angles. That is a hybrid, not a choice between two columns.

The hybrid model we run

Scale Digital does not produce your ads. We brief the strategy into your in-house team, coach them to produce at volume, then buy the media against what they make. It has three moving parts.

1. Brief the scripts and angles in

Before anyone films, the concept is written down: the hook, the angle, the promise, the proof, the format, the call to action. We reverse engineer what is working in your market and translate it into scripts specific to your brand. The brief is the product. The finished ad is just the brief made visible. This is the strategy layer the agency charged a premium for, delivered as instructions your team can execute.

2. Coach the team to produce at volume

Because the brief did the hard thinking, production becomes fast and cheap. Your team makes dozens of distinct concepts a month, not four. You are not paying a premium per asset, so you can afford to be wrong most of the time, which is the entire point of testing. You keep in-house cost and speed while gaining the range in-house teams usually miss.

3. Feed winners back into new briefs

When an ad wins, we interrogate why, then write the next round of briefs around that insight. A winning hook becomes five new hooks. A winning angle becomes a new format. The system compounds because every result teaches the next brief, and losers get cut without ceremony because they cost almost nothing to make.

Why this is the model that scales

Media buying only works on top of a system that never runs dry. Keep the account supplied with fresh, on-brand concepts and it can absorb budget without fatiguing. This is the through-line behind results like Copini at 890% on a 10x MER, Freya Meds growing US revenue from $300k to $3m a month, and b.box up 52% year on year. Creative volume was the ceiling, and the hybrid raised it.

So the honest answer to "in-house or agency" is neither, on its own. Take the cost and speed of in-house, add outside strategy on the scripts and angles, and let the media buying work against a pipeline that keeps producing winners.