The production-budget-percentage question is the wrong one

The spread of cited splits is wide enough to be a genuine warning sign rather than a range to average. Traditional TV-era guidance put production near 10% of budget; some digital-native takes argue for 30% or more; a widely-cited industry figure lands near 12.5%. The Association of National Advertisers, after reviewing this exact question, concluded there's no agreed working-to-non-working ratio that holds across categories.

That's not because every business is a unique snowflake. It's because a fixed percentage of media spend was never actually measuring the thing that determines whether a creative budget is adequate. It was measuring convention, and convention varies by who's telling you the number — an agency's production percentage recommendation is not independent of how much production revenue that agency earns.

Ad creative fatigues on a real, roughly predictable calendar. At meaningful spend, most creative loses effectiveness within two to four weeks as an audience segment is repeatedly exposed to it — this isn't a controversial claim, it shows up consistently across paid social accounts regardless of category.

That fatigue window is the actual constraint on production budget. If a given ad set needs a new concept every three weeks to avoid decline, the minimum viable production spend is set by how many ad sets you're running and what a concept costs to produce — not by what fraction of total media spend feels proportionate.

This reframes the entire question. Instead of asking 'what percentage should production be,' the right question is 'how many fresh concepts do I need to ship every fatigue cycle, and what does that cost' — which produces a dollar floor, and only afterward, incidentally, a percentage.

Take an account running five active ad sets, needing a new concept per set every three weeks, at $800 per concept — a realistic cost for a scrappy but genuinely tested creative operation. That's five concepts every three weeks, or roughly $1,333 a week, near $5,780 a month in production spend as a hard floor, independent of whether the account is spending $30,000 or $150,000 a month on media.

At $30,000 monthly media spend, that floor is about 19% — squarely inside the range agencies cite. At $150,000 monthly spend with the same five ad sets, the same floor is under 4% — well below most cited ranges, and correctly so, because five ad sets don't need more creative just because more dollars are flowing through the same five auctions.

The percentage only moves back into a 'normal' range when ad-set count scales with spend, which is common but not universal. A account running the same handful of evergreen ad sets at rising spend needs a production budget that tracks ad-set count and fatigue cadence, not one that auto-inflates with the media line item.

Stop budgeting production as a derived percentage of whatever the media number turns out to be. Start from testing cadence: how many ad sets are actually running, how fast does creative fatigue in this specific account (measured from actual performance decay, not assumed), and what does a concept genuinely cost to produce at the quality bar the account needs.

That calculation will sometimes land near a conventional percentage and sometimes won't, and the accounts where it lands far outside convention are exactly the ones where following a generic split would have either starved the account of fresh creative or wasted budget producing more than the ad-set count could ever use.

The uncomfortable part for any agency, including this one, is that this framing removes the easy answer a prospective client wants in a first call. 'What percentage should we budget for production' has a satisfying one-line answer. 'How many ad sets do you run and how fast does your specific account's creative decay' requires actual account data before anyone can answer honestly — which is exactly why the honest answer is worth more.