DCO (Dynamic creative optimization)

How DCO works

Let's take a drink brand as an example. They sell a low-sugar sports drink for active women over 50, mostly from their own website. The drink brand buys ads on Meta, TikTok, and YouTube. Instead of building twenty finished ads by hand, the drink brand uploads the pieces separately: five headlines, four videos, three descriptions. Dynamic creative optimization (DCO) is the platform mixing those pieces into ads, then spending more on the mixes that sell.

For a Meta Advantage+ campaign (Meta's automated campaign type) that pool is 5 × 4 × 3 = 60 possible combinations. Not all of them get meaningful spend. The platform explores, then concentrates. The drink brand does not pick which person sees which mix. That is the trade of dynamic creative optimization (DCO): coverage and assembly, in exchange for control.

Let's take a drink brand as an example. They sell a low-sugar sports drink for active women over 50, mostly from their own website. The drink brand spent $8,000 over three weeks on dynamic creative optimization (DCO). The mix of Still training at 58 plus a kitchen pour video took about $3,200 of that $8,000 spend (40%) and bought purchases at $18 each. The mix of Electrolytes for athletes plus a studio bottle shot took about $400 of the $8,000 spend (5%) and bought purchases at $42 each. The other combinations shared the rest. Dynamic creative optimization (DCO) found a horse. It did not explain whether the headline or the video was doing the work.

Dynamic creative optimization (DCO) also acts as a fatigue buffer. Creative fatigue is when the same people see the same ad so often that results fade. More combinations mean the same person sees fewer identical impressions, which softens creative fatigue without a new shoot.

DCO vs dynamic product ads

Dynamic creative optimization (DCO) varies assets. DPA (dynamic product ads) varies products from a catalog feed. A dynamic creative optimization (DCO) ad can show one drink-brand pouch with ten headlines. A dynamic product ad (DPA) can show ten shoes from a catalog with one headline. The confusion costs teams weeks of pointless debate. If you sell many products and want the item the person viewed, that is dynamic product ads (DPA). If you have one hero product and many ways to talk about it, that is dynamic creative optimization (DCO).

How to read DCO results

Asset-level breakdowns in dynamic creative optimization (DCO) are directional at best. Delivery is not evenly split, so an asset that got 2% of impressions did not lose a fair fight. Treat the report as a shortlist, then confirm the winner in a clean creative test where only one thing changes.

Let's take a drink brand as an example. They sell a low-sugar sports drink for active women over 50, mostly from their own website. The drink brand took the winning headline and the winning video out of dynamic creative optimization (DCO) and ran them as two separate ads against a control, $1,500 each. The kitchen video won with either headline. The headline was not the lever. Dynamic creative optimization (DCO) could not have told the drink brand that, because it never isolated the two.

• Keep the asset pool coherent so any combination still makes sense. A price headline on a brand film will still get assembled.
• Do not mix two offers into one dynamic creative optimization (DCO) ad. The winning combination becomes unreadable.
• Retire weak assets rather than adding more on top of a bloated pool.

Common DCO mistakes

Calling a dynamic creative optimization (DCO) launch a test is the usual one. Let's take a drink brand as an example. They sell a low-sugar sports drink for active women over 50, mostly from their own website. The drink brand's first write-up said the kitchen mix won the test. It won delivery. Those are different sentences. Advantage+ (Meta's automated campaign type) on top of dynamic creative optimization (DCO) then made it worse: two automation layers, one result, no one sure which layer to credit.

• Calling a dynamic creative optimization (DCO) launch a test and reporting asset splits as results.
• Feeding it leftover assets, then concluding the audience is tired.
• Running dynamic creative optimization (DCO) and Advantage+ automation at once, then trying to attribute the change to one of them.