What is a creative system?

A creative system is strategy, production and optimisation for paid social run as one continuous loop, with the weekly decisions about what to pause, remix and replace made from the account's own data, and it is what the medium now requires. Social is the largest advertising channel in the world. WARC put global social ad spend at $247.3 billion in 2024, up 14.3% in a year and past paid search for the first time, and GWI has adults in Europe spending twice the weekly hours on short form video that they spend on television, Gen Z five times. For most consumer brands it is the biggest line on the media plan. It is also the only major channel still largely fed with creative built for a different medium, by an operating model built for a different medium, and the distance between those two facts is the largest unlock most brand teams have available to them.

The creative arriving in the channel tends to be one of two things, a campaign asset adapted down from television thinking, one hero idea stretched across formats until it fits a feed it was never designed for, or content volume, a hundred get-ready-with-me videos commissioned in the hope that quantity finds quality. One is a single bet placed before any data exists, the other is noise, and neither is a system. The third option is the subject of this essay, and the reason it exists is a short piece of history.

How social changed the conditions

The campaign model was one of the great pieces of commercial machinery and it was right for its conditions. Concentrate the budget behind one big idea, pretest it because the bet is large, negotiate the media months ahead, flight it, measure afterwards. Every step rested on an assumption about how media worked, negotiated prices, captive audiences, scarce production, and social changed four of the mechanics underneath without anyone deciding to.

Pricing moved to an auction. Broadcast media was priced before the creative existed, so the quality of the work could not change what a rating point cost, and procurement's split between working media and non-working production made complete sense. On social every impression is won in real time, and Meta describes its auction as ranking each ad on three things, the bid, the estimated action rate, meaning the probability the person takes the action the campaign is optimised for, and ad quality. Two of the three are properties of the creative. An ad people watch wins impressions it would otherwise have lost and wins them cheaper, an ad people skip pays more for what it still gets, and the creative became the variable that sets the media price. Why creative is your largest working media lever carries that argument.

Curation became algorithmic. The feed shows each person whatever they are most likely to keep watching, from any source, so an ad is ranked against the friend's holiday, the creator they follow and the news, not against the other ads in the category. The competitive set is the best thing in the feed, and the bar for looking like something worth a second of attention rose with it.

Formats became skippable. The ad break held people in their seats, so planning could treat exposure and attention as the same thing and price every exposure as if it were watched. Most feed placements can be swiped away after a second or two, the swipe is the default, and for the first time inattention is a recorded event. Amplified Intelligence, which sells attention measurement, puts the memory threshold at about 2.5 seconds of active attention and finds more than 80% of ads never reach it. A skip is not nothing. A branded first second can refresh a memory that already exists, which is the reminding job, but it cannot build an association that is not there, and building is the job growth depends on. The served impression stopped being the unit that matters and the watched one took its place.

Production became something anyone could do. For most of advertising's history polish carried a signal, because only a real company could afford a film crew. Then the phone in every pocket became a studio, billions of people grew fluent in made-for-feed video, and polish stopped signalling quality and started signalling ad, which the feed has taught people to pass. The same collapse in production cost is what makes a portfolio of twenty to thirty distinct ads affordable at all, so the break and the opportunity are the same event.

None of this was anyone's mistake. The model was a correct answer to conditions that no longer hold, and operating models outlive their assumptions, particularly when they live in org charts, rosters and approval chains.

The definition

A creative system is Strategy, Production and Optimisation operating as one continuous loop. Three parts, none optional, none separable.

Strategy translates brand goals into a portfolio of distinct creative hypotheses. Not one big idea, twenty to thirty of them, each built for a specific group of category buyers and a specific category entry point, the situations, occasions and feelings that connect a buyer to the category in Jenni Romaniuk's framing. The portfolio is the unit of planning, not the ad, and the platform enforces that. Meta's own guidance now describes creative diversification as "the best lever to find the most relevant audiences", because its delivery system matches each ad to the people who respond to it, so what is in the frame decides who sees it. Diverse creative is diverse targeting, and Why you need a creative portfolio, not more content explains the mechanics.

Production makes ads that look like the medium they run in. Television ads are shot like television, radio ads sound like radio, and social ads need to look like social content, real people in real environments, or the production style triggers the skip before any message lands. Native does not mean off-brand. The distinctive assets, the message and the associations are carried in a format the feed does not reject, and consistency lives there, in the brand, while the variation lives in the execution, because different openings stop different people. What high quality creative means on paid social argues the standard from effectiveness backwards.

Optimisation is what happens after launch, which is where data-driven marketing lives. Every week the portfolio faces three decisions, which ads to pause, which to remix, and which net-new concepts to brief, and an ad is retired because it was not watched, never because it is old. The market tells you what works, the system acts on it, and the learning carries into the next brief. Why data-driven creative only starts after launch describes the loop, and How to measure attention on social with data you already have describes the instrument it runs on.

Strategy without production is a deck, production without optimisation is content, optimisation without strategy is reporting, and the loop is the product.

Always-on, at account level

A campaign is one bet placed before the data exists, and a system is a portfolio of bets that improves weekly while the data flows. On a medium that tells you within days what is working, set-and-forget creative means paying for the feedback and ignoring it, so a creative system runs as an always-on portfolio at the level of the connected ad account rather than as a sequence of campaigns with start and end dates, because the platform's delivery system does not treat campaigns in isolation and neither can the people optimising it. Campaigns still exist inside it, launches and seasons are real, but the learning carries across them.

The first ninety days of a portfolio are calibration, not a verdict. Twenty to thirty hypotheses go live, the early numbers are mixed because mixed is the design, and the account becomes intelligent in the third month, when the surviving portfolio is concentrated in validated territory and the next briefs are written from evidence. The phases compress with scale, and the brand payoff lands over quarters and is read on its own clock.

How results are read

A system that runs on data needs a discipline for reading it, and ours is published in full in How to read paid social results. Results are read in one order. Effectiveness first, did the brand move, read from lift studies against a holdout and from tracking, on their own clock. Performance second, did the ads earn attention, read weekly from stop rate, hold rate and watched rate per ad on one platform and one threshold. Efficiency third, what did it cost, read from CPM, which is media efficiency, and from the all-in cost per watched impression with the fee included, which is marketing efficiency, both with the campaign objective stated beside them. Effectiveness goes first because achieving the right goal matters more than saving on the wrong one, and a watched rate is never a proxy for brand lift. If effectiveness was not measured, we say what was measured and stop.

Two distinctions keep the reading honest. Attention is not engagement. Likes, comments and shares are actions a minority takes, skewed to fans and heavy buyers, and Ehrenberg-Bass is right to dismiss them, whereas watching is the absence of an action, not skipping, measured across every impression including the light buyers who will never follow you, so a watched second with the brand's assets on screen is a branded second, and watched rate across a portfolio is branded reach. And we read attention, we do not buy it. The creative is optimised for attention and the platform objective stays set for the brand's goal, reach for brand building, because setting the objective to video views sends the system looking for people who watch anything, and reach narrows. The goal is the brand's, the objective is only the platform setting. Why every social ad has two jobs sets out what each second of the ad is doing, reminding in the opening with the distinctive assets on screen, building in the body with the watched view, and which metric reads each.

Two variables, one of them managed

A paid social account has two variables, media and creative, and the industry manages one of them. Media has a strategy, a setup, a weekly rhythm, a measurement standard and an owner, correctly, and the media team's remit does not change in any of this. Creative, on a medium where it sets the price of every impression and decides who the ad reaches, arrives as a deliverable, and between campaigns nobody is set up to make the weekly decision about it. The creative agency works at campaign cadence, the media agency holds the data and no creative mandate, the in-house team makes the cutdowns, and the client holds every contract and assumes, reasonably, that between excellent partners the loop must be running. It falls into the seam, and the seam has no owner because the org chart predates the medium. Why creative needs managing as actively as media makes the full case, and Where a creative system fits with your agencies maps the roster around that seam. The short version is that a creative system holds the layer between the brand platform and the media account, translates the platform into a portfolio, surfaces creative signals to the media team, and supplements every partner rather than replacing any. No media buying happens inside it.

The strategy that layer runs on is not new. How to apply brand science to social media works through Byron Sharp's penetration arithmetic, light buyers who sit in no retargeting pool and follow nobody in the category are the growth audience, and reaching them is what advertising is for. The medium changes one piece of the mechanics, on a skip medium an unwatched impression is not reach in the memory sense, so the portfolio and the weekly loop are how you get the reach Sharp wants.

What changes when the loop runs

The effect shows up in two places, attention and the media bill, because on auction-priced media they are the same lever.

Ads people watch cost less to deliver. In one measured account, same budget, same media plan and targeting, across three weekly cycles, the brand's existing creative was buying a completed view at $0.50 and the social-first portfolio reached $0.09. Part of that gain is the loop retiring weak work, each week's pause decisions removed the ads that were spending without being watched, and the rest is the auction rewarding the ads that were. The improvement was creative.

The brand outcomes move with it, and these are the outcomes brand builders are paid to deliver. Vaseline saw three times the brand lift of the brand's existing creative against the same brief and media spend. Whiskas saw five times the brand favourability lift of its existing creative on the same media. KitKat, a brand everyone already knew, saw two times the purchase intent of its existing creative in lift testing. Mental availability, association, intent, achieved not by spending more but by making the existing spend work harder. The Advertising Research Foundation's 2026 validation work is the right caution, attention translates into lift only alongside media weight, reach and frequency, so the loop does not replace the media plan, it decides what the media plan delivers.

The better campaign

The counter-position deserves its strongest form because most of the industry holds it. The campaign model does not need replacing, it needs upgrading. Adapt faster, produce more assets, coordinate the agencies more tightly, run the quarterly review monthly, and the existing roster covers the medium. Creative agencies hold this view because it protects the integrity of the idea, media agencies because it protects the boundary of their remit, brand teams because it protects continuity, and all three concerns are legitimate.

The trouble is the cadence. The medium produces decision-grade creative data daily and the campaign model's feedback loop runs quarterly however hard everyone works, because the loop's speed is set by the structure and not by the effort inside it, so a weekly medium run on a quarterly loop spends most of every quarter flying on stale information. More assets without distinct hypotheses is one ad with ninety-nine invoices, because the platform consolidates delivery behind similar ads. Faster adaptation of a hero film is the same ad meeting the same skip in more formats. Tighter coordination between a creative team and a media team does not create the function neither of them owns. The campaign model was not wrong, its assumptions expired, and an operating model cannot be upgraded past the assumptions it was built on.

That gap is what Creatively Squared exists to close. It is the paid social creative system for brand growth, neither agency nor software, a managed technology solution combining social-first creative strategy, native production and AI-powered optimisation to deliver 2-5x improvements in attention, conversions and brand outcomes, where those multiples are measured against the brand's own existing creative on the same media, in lift studies and in the account's own data, and stated only where they were. The unlock is not more spend and not more content, it is treating creative as the system it has become, on the channel where it now decides what every media dollar buys. Media creates the opportunity to see your ads. Creative determines whether they work.

What to do with this

You do not need a vendor to find out whether you are running a system. Pull last quarter from Ads Manager and answer four questions. How many distinct creative concepts ran, not formats or cutdowns, concepts, and if it is under ten the portfolio is too narrow for the platform to do its matching. When an ad underperformed, what happened next and how fast, and if the answer is the next planning cycle, the loop is not running. Did anything learned this quarter change what gets briefed next quarter, because learnings that live in a wrap report mean the system has no memory. Would your ads be watched in the feed without the logo, because ads that only make sense as ads are being skipped before the message arrives. How to audit your own paid social account turns those into a worksheet with the numbers to pull. Then read Why social lowered tolerance for ads, not attention, which starts the argument for the loop with what a skip actually is.

Questions

How is a creative system different from a creative agency?

An agency sells its people's time and craft, usually organised around campaigns with a brief at the start and a wrap at the end, and the best ones are superb at the idea. A creative system sells the running of a portfolio, strategy, production and optimisation operating continuously on one connected ad account, with the weekly pause, remix and replace decisions made from the account's own data and carried into the next brief.

The agency keeps the brand platform and the hero work. The system translates that platform into twenty to thirty social-first executions and manages them week to week, which is the function that falls between an agency working at campaign cadence and a media team with no creative mandate.

How many ads do we actually need?

Enough distinct hypotheses for the platform to do its matching, which in our experience is twenty to thirty per period, each built for a specific group of category buyers and a specific category entry point. Distinct means different situations, people and openings, not aspect ratios or cutdowns of one film, because the delivery system matches each ad to the people who respond to it and consolidates spend behind similar ones.

The number scales with media, a larger plan reaches more kinds of people and needs more ways in, and it is refreshed by the loop, which retires an ad because it stopped being watched, not because it is old.

Does a creative system replace our media agency?

No, and it should not try. The media agency keeps bidding, placement, audience settings, budget allocation and campaign setup, all of which still matter, and no media buying happens inside a creative system. What the system adds is the creative half of the account, the portfolio flowing in and the creative signals flowing out, which ad is tiring, which group of buyers is responding, read weekly and handed to the media team to act on.

Better creative lowers the media agency's CPMs under a reach objective, which is the shared outcome, and Where a creative system fits with your agencies sets out the boundary in detail.

Sources

• WARC Media, "Social overtakes paid search to become the world's largest channel by ad spend", 15 May 2024. Global social ad spend of $247.3 billion in 2024, up 14.3%. Subscription research publisher, no media interest.
• GWI, 2024 to Q1 2025, all adults and Gen Z, Europe, weekly hours by video type, as carried in Meta, "Demystifying attention", 2025. Short form at twice the weekly hours of TV, Gen Z five times. Survey data, presented by Meta, which sells the media.
• Meta Business Help Centre, "About ad auctions". Ranking on bid, estimated action rate and ad quality. Platform documentation, Meta sells the media.
• Engineering at Meta, "Meta Andromeda", 2 December 2024, and Meta for Business, "Demystifying Creative Diversification", April 2025, as quoted by Jon Loomer. Ad diversity and advertiser outcomes, creative diversification as the best lever to find relevant audiences. Platform, Meta sells the media.
• Amplified Intelligence, "Why does the attention-memory threshold matter?", 27 June 2022. The 2.5-second threshold and the share of ads that miss it. Amplified sells attention measurement.
• Advertising Research Foundation, Attention Measurement Validation Initiative Phase 3, 9 June 2026. Attention translates into lift alongside media weight, reach and frequency. Member-funded industry body.
• Byron Sharp, How Brands Grow, 2010, and Jenni Romaniuk, Building Distinctive Brand Assets, 2018. Penetration, light buyers, category entry points, consistent assets with varied executions. Ehrenberg-Bass Institute, sponsored by large advertisers.
• One measured account, same budget, same media plan and targeting, anonymised. Cost per completed view from $0.50 to $0.09 across three weekly cycles, part of the gain from the loop retiring weak work. Our own account data.
• Vaseline, Whiskas and KitKat, brand lift testing against each brand's existing creative on the same brief and media. Cleared proof points from our own accounts.