We've all heard the promise of perfect targeting: the right person, with the right message, at the right time. Whether it was custom audiences, lookalikes, CRM-based segmentation or multi-layer interest stacking, many brands adopt their competitive edge that lives inside the targeting menu.
Meta's new Andromeda optimisation system signals somewhat of a turning point with the algorithm now prioritising creative variation and quality above almost everything else. Targeting still matters but it is no longer the lever that differentiates one brand from another. Meta now claims their machine learning actively reduces the impact of granular audience choices.
How Creative Outperforms Targeting
Meta's latest updates, including Advantage+ Creative Suite and automated audience expansion, are designed to increase the algorithm's surface area. The more creative variants the system receives, the faster it can learn which concepts, messages and visual cues drive action.
Meta now believes that creative quality drives data quality drives performance. Previously Meta would ask which audiences you plan to target, whereas today they ask what creative variations should be fed into the machine.
The implications? With fragmented media, rising CPMs and increasingly opaque platform reporting windows, creative is becoming the de facto performance engine.
In fact, early independent tests show that campaigns built with creative-first inputs (i.e. multiple distinct visual styles, messages, formats) see lower CPMs and higher relevance scores under Andromeda. For example, one guide noted up to an 8% improvement in ad quality scores when moving from segmented targeting to creative-diversity-led campaigns.
This means that even brands without vast audience data now have a lever they can reliably pull: creative variation.
Moreover, the creative engine now incorporates signals such as hook strength in the first 3 seconds, on-screen product visibility, caption length, and engagement patterns (watch time, saves, shares). These become proxies for "who responds best" rather than explicitly defined audiences.
The ROI Gap Between Creative Leaders and Creative Laggards
Across hundreds of ADMATIC audits and meta-analysis of platform changes, one pattern is constant:
Brands with strong creative systems outperform those with targeting-first systems by 20-50% on ROAS. Why? Because the algorithm recognises what humans often overlook, and that is that users respond to behavioural resonance, rather than demographic labels.
One compelling image, one angle that reframes a problem, one storytelling device that creates an emotional micro-shift, tends to shape the consumer action more than a hyper-segmented audience.
In practical terms, this means brands that invest in high-signal creative libraries (multiple angles, UGC-style content, hero/hero-plus formats, early hooks) get more mileage for each dollar spent.
What This Means for Media-Mix Modelling and Multi-Channel Strategy
Consider a creative-led world in which the need for better measurement frameworks is accelerated.
When creative becomes your competitive edge, you must understand:
- Which creative themes drive upper-funnel lift?
- Which concepts convert at the lower funnel?
- How creative interacts with channel mix (CUTV, video, retail media, search)?
- What happens when a creative concept is pushed across multiple platforms?
This is where media-mix modelling becomes essential because creative performance is not isolated to Meta. A high-performing Meta concept can also succeed across YouTube, BVOD, CUTV and programmatic video.
Andromeda's impact is not limited to Facebook & Instagram. Because Meta's system uses creative as the core signal, brands that align their entire media ecosystem (TV, streaming video, social, display) around top-performing creative themes unlock compounding effects. For example, a hero video that performs on Meta can feed into CUTV buys, create lookalike DSP audiences, and unify the message across screens.
Consequently, MMM (media-mix modelling) now could include creative effectiveness as dimension.
Creative effectiveness becomes a multiplier in the model.
Additionally, the shift toward creative-first means that classic optimisation levers like changing audience splits or tweaking bids are less impactful. Instead, budget reallocation questions become: Which creative library should get +$ spend? and which creative cluster do we scale or pause?
The Creative Operating System
Consider something we coin the Creative Operating System. This is a data-led ecosystem that includes a measurement-first framework that integrates:
- Creative clustering
- Audience behaviour mapping
- Algorithmic learning cycles
- Multi-channel creative propagation
- MMM-driven budget optimisation
It's how brands can move from creative assets to creative intelligence.
Key components of the Creative Operating System at work include:
- Creative libraries: structured sets of assets grouped by funnel stage, persona, or message theme.
- Creative diagnostics: using platform signals to identify which hooks, formats, visuals produce strongest user engagement and conversion lift.
- Signal-feed loops: ingesting backend data (CRM, sales, onsite behaviour) to feed into meta-/platform-creative optimisation (for example, feeding high-lifetime-value customer segments back into creative testing).
- Cross-platform propagation: once a creative cluster wins on social, amplify it via CUTV, programmatic, retail-media networks, and integrate into media-mix modelling.
- Budget dynamics: shifting spend based on creative performance across channels, not merely channel performance alone.
At ADMATIC, we're helping clients audit their creatives and refine their Creative Operating Systems to future-proof their presence in a creative first world. Now is the time to get your brand Andromeda-ready!