The creative strategist is the most important hire in DTC advertising right now

For about ten years, winning at paid ads meant out-fiddling everyone else. You stacked audiences. You split tested bids. You built account structures with more layers than a wedding cake and defended them in meetings like holy texts.

That job is gone.

Now, before I make the case, let me kill the stat you’ve seen a hundred times. “Over 70 percent of performance comes down to the creative.” I went looking for the study. There isn’t one. The number appears to come from someone adding creative (47 percent) to reach (22 percent) in an old Nielsen study and rounding up, plus a Google executive saying “70 percent” in a trade interview with no methodology attached.

The closest real research is NCSolutions, who modelled about 450 campaigns and put creative at 49 percent of short-term sales lift. That’s still the biggest single factor, bigger than reach, targeting and recency put together. But it’s about half, not 70. It’s packaged goods, not DTC. And it measures short-term sales lift, not “paid media performance”.

Good. The argument doesn’t need the fake number.

Here’s the definition, in plain words. A creative strategist is the person who reads the ad account like a scoreboard and turns what it says into instructions for the people who make the ads. They sit between the numbers and the camera. They don’t usually edit the video. They decide what the video has to do.

That role barely existed as a job title five years ago. Now it’s the one brands fight over.

What actually broke

Two things, in order.

First, Apple’s App Tracking Transparency landed with iOS 14.5 on 26 April 2021 and gutted the data platforms used to aim ads. Roughly 96 percent of US users left tracking off. Meta’s CFO later told investors it would cost the company about $10 billion in 2022 revenue. That is not a tweak. That is a business model taking a hit it never recovered from.

Second, the platforms rebuilt delivery around machine learning. Performance Max arrived in November 2021. Advantage+ shopping campaigns followed on 10 August 2022. Both take the steering wheel off you and hand it to a model.

That model has to aim at someone. So it reads the ad. The visuals, the words, the pace, the mood. It uses those signals to guess who will care, then goes and finds them.

That’s not a metaphor and it doesn’t need a percentage to be true. In an automated buy, the creative is the targeting. Our own testing frameworks guide covers the mechanics if you want the wiring diagram.

Which left a hole in most marketing teams. On one side, media buyers who could talk about cost per acquisition all day and had no view on why an ad felt cheap. On the other, designers who made beautiful work and had never heard the phrase “contribution margin”. Foreplay’s guide to the role describes the same split. Neither side could fix a losing ad, because neither side could see the whole problem.

The creative strategist is what the market built to fill that hole.

What the job looks like on a Tuesday

The honest version: a lot of admin wrapped around a few sharp decisions.

Every day takes about a fifth of their time, per 3Search’s breakdown of the role. They open the dashboards. Meta Ads Manager, plus something like Motion or Triple Whale that shows creative performance visually instead of as a wall of numbers. They’re hunting for outliers in both directions.

Three numbers do most of the work. Hook rate (did people stay past three seconds). Hold rate (did they stay past fifteen). Click-through rate.

They’re also watching for fatigue, and this is where most teams are working off a myth. The common line is that ads “start to decline after four to seven views”. Meta published the actual research in 2023 and it says something harder to live with. There is no threshold. Conversion likelihood falls on a smooth curve from the very first repeat, and by the fourth exposure it’s down roughly 45 percent. Meta looked specifically for a warm-up period where repetition helps. There isn’t one. Their words: clicks and conversions “become monotonically more expensive with repeated creative exposures”.

The average person sees the same creative 4.2 times a month. Over 19 percent of impressions go to someone who has already seen that ad five times or more.

So fatigue isn’t an event you catch. It’s a tax you’re paying from day one, and the only way to stop paying it is new work.

They also spend real time just looking at ads. Meta Ad Library, TikTok Creative Center, the feeds themselves. Not to copy. To spot a format or a sound before it peaks.

Every week is where the output happens. Monday is usually a post-mortem with the media buyers: kill the losers fast, work out what the winners had in common. Then the briefs.

The brief is the job. Everything else is preparation for it. A good one names the exact person it’s aimed at, the psychological bet being made, the required hook, the pacing, and the number that decides whether it worked. LaunchCodex describes this as the core of performance creative: every element engineered against a business outcome, then tested.

The trick is leaving room. Brief it too tight and the creator reads your script like a hostage. Brief it too loose and you get a mood board that costs you $4,000 and converts nobody.

Every month is the audit. Are we too reliant on one format? If 90 percent of spend is going to talking-head UGC, the algorithm has stopped exploring. Time to force statics, carousels, founder pieces, anything that opens a new door.

You’ll see confident benchmarks quoted here, usually something like “6 to 10 new concepts a month”. I couldn’t trace that one to any source, so I’m not going to repeat it at you. The volume numbers further down are real and they’re better anyway.

Every quarter is the money conversation. Production budget against media budget. You’ll hear a rule of thumb that 10 to 20 percent of media spend should go into making and testing creative. Worth knowing where that comes from: it’s inherited from television, where roughly 90 percent went to the buy and 10 percent to the shoot. It was never an optimum anybody measured, and in digital it’s openly contested. Nike reportedly runs closer to 80 percent production.

Treat it as a sanity check, not a law. If you’re spending $100k a month on ads and $3k on making them, you’ve found your bottleneck without needing a benchmark to tell you.

The research bit most people skip

Most “creative research” is scrolling the Meta Ad Library and copying whatever a competitor is running.

Think about what that gets you. You’re copying an ad you can’t see the results of, made for a different offer, aimed at people who have already seen it. You’re arriving second to a party that’s already loud.

The better method goes to the customer instead. Dara Denney’s approach, which she’s walked through publicly, is built on reputation analysis and review mining. Skip the competitor ads. Go read support tickets, Amazon reviews, Reddit threads, the comments under your own posts. Pull out the exact words people use for their problem and the exact words they use for the relief.

Then you stop targeting demographics and start targeting arguments. Not “women 25 to 34”. If you’re selling a coffee alternative, you make one ad for the person who cares about taste, one for the person doing the maths against their daily flat white, and one for the person who just wants it faster in the morning. Three different bets, fed to a machine that’s looking for exactly that kind of variety.

The testing framework

The goal isn’t a perfect ad. Nobody makes a perfect ad. The goal is a machine that finds winners on purpose instead of by luck.

The structure that works is the sandbox: one new variable, one controlled ad set, fixed budget, fixed targeting. Change the hook and the copy and the call to action all at once and you’ve learned nothing, because you can’t tell which one moved.

Test in roughly this order, because that’s the order of impact:

• The hook. First three seconds. Biggest lever by a distance
• The format. Static against video against carousel. Polished against rough. This decides where the ad even gets placed
• The angle. Feature or benefit. Aspiration or problem-solution
• The copy and CTA. Real, but small. Don’t start here

You’ll find blog posts claiming a hook swap moves hook rate by 15 to 20 percentage points. Given that published hook rates sit around 18 to 28 percent, that would mean doubling the metric or erasing it. Nobody has run that test publicly. It’s a made-up number and I’d rather tell you that than pass it on.

The win rate nobody wants to hear

Now the part that reframes everything.

Motion analysed 578,750 unique creatives across 6,015 ad accounts and $1.29 billion in realised spend, covering September 2025 to January 2026. Motion sells creative analytics software, and the data is from their own customers, so read it knowing who benefits. It’s still the biggest public dataset we have on this.

Their definition of a winner is deliberately narrow: an ad that spends at least ten times the account’s median ad spend, and at least $500 total. Note what that measures. It measures whether Meta’s auction decided to feed the ad. It is not a profit test. Motion says so explicitly: outcomes are not tied to ROAS, revenue or conversions.

The overall hit rate is about 5 percent.

Roughly one in twenty things you make will be the thing.

And here’s the detail that stings if you’re small. The win rate goes up with account size, from around 4 percent at micro accounts to around 9 percent at enterprise. Bigger brands don’t just ship more, they hit more often. Whatever advantage they have compounds.

You can’t do much about the rate. The only lever you have is volume. Here’s what accounts at each spend level actually launch per week:

| Monthly ad spend | Average new creatives/week | Top 25% | | :– | :– | :– | | Under $10K | 2.80 | 4.83 | | $10K to $50K | 4.10 | 8.09 | | $50K to $200K | 6.67 | 15.95 | | $200K to $1M | 11.24 | 31.11 | | $1M+ | 18.85 | 54.64 |

Source: Motion, Creative Benchmarks 2026

Important: that’s a description of what people do, not a target Motion is setting. They say so themselves. Use it to place yourself, not to feel bad.

Read the bottom row anyway. At $1M a month, the top quartile is shipping 54 creatives a week. That’s not a creative team. That’s a small factory.

What all that volume buys has a name: creative liquidity. Enough different ads live that the algorithm can spread your money around without any single ad becoming a single point of failure. It’s the difference between scaling and holding your breath.

The tools

Running 50 to 100 tests a month in a spreadsheet is how people quit.

Native reporting won’t save you either. Ads Manager wasn’t built to isolate creative variables, and the attribution muddies what you’re looking at. So strategists lean on visual tools. Motion and Triple Whale’s Creative Cockpit let you stack statics against video, or see the exact second people drop off, and then show that picture to an editor who was never going to read your pivot table. That’s the real value: it makes the argument visible to the people who make the ads.

For briefing and swipe management, Foreplay does the same job in the other direction.

Then AI, where the honest picture is smaller than the hype. IAB surveyed 104 US advertising executives in late 2025 and found 83 percent say their company has deployed AI somewhere in the creative process, rising to 85 percent for social advertising specifically. That’s US ad executives at large advertisers, not DTC brands, and 104 people is a small room. You’ll see that 85 percent quoted as “85 percent of DTC advertisers”. It isn’t.

The useful applications are boring and enormous anyway. Feed a model two years of support tickets and get back a ranked list of what people actually complain about. Take one winning script and generate twenty hook variations. Set up monitoring that pings you the moment cost per sale crosses your profit line instead of finding out on Friday.

What AI hasn’t done is have the taste. It’ll give you twenty variations. Picking which three are worth making is still the job.

The part that separates good from expensive

The best creative strategists are fluent in money, not just metrics.

Platform ROAS is the most confidently wrong number in your business. It’s self-reported by the company selling you the ads, and it’s been getting less reliable every year since 2021.

The alternative most operators land on is MER: total revenue divided by total ad spend. Crude, honest, and it can’t be gamed by attribution windows. aMER goes one better by counting only new customer revenue, which tells you how well the creative is bringing in strangers rather than re-selling to your email list. Cody Plofker, then running marketing at Jones Road Beauty, is the operator most associated with pushing this into DTC practice.

Then go one step further and subtract your cost of goods. Now you’re looking at profit on ad spend. An ad that sells a mountain of your worst-margin product looks like a hero in Ads Manager and a problem in your bank account. A strategist who understands margin will kill that ad in favour of one that sells less and earns more.

That’s the difference between someone who makes ads and someone who runs a growth engine.

Three brands worth studying

Jones Road Beauty launched in October 2020, the day Bobbi Brown’s non-compete with Estée Lauder expired, and passed $100 million in revenue in its third year.

The TikTok story is the one worth knowing, and it needs its date attached. In early 2022, a customer asked how to contour her nose. Bobbi Brown filmed a 29-second reply: why would you want to contour your nose, your natural face is the most beautiful face. In the two weeks after that post, TikTok’s share of the brand’s revenue went from 2 percent to over 30 percent. The team put Spark Ads behind the organic content and took daily TikTok spend from $1,000 to over $4,500.

You’ll see that 30 percent quoted as a permanent state. It wasn’t. It was a two-week spike off a viral moment, which is a more useful lesson anyway: the win wasn’t the video, it was someone recognising what they had within days and putting money behind it.

Koala, the Australian furniture brand, went the other way and built it in-house. Mi3 reported in 2021 that creative, media and tech were handled internally by some 40 staff, after the business axed its CMO role and restructured.

Their “Never Uncomfortable” work, which ran in March 2020, is the case study I’d point most Australian brands at. Instead of listing mattress features, they set physical comfort against deeply uncomfortable human moments. Sub-par bedroom performance. Unexpected pregnancy. Coming out to your parents.

And they broke the rule everyone repeats about digital video needing to be under 15 seconds. The hero cut ran 2 minutes 30, with 30 and 15 second versions beneath it. It held attention by cutting every 2 to 2.5 seconds between live action, animation, stills and typography. Koala’s creative director told B&T the campaign delivered a 20 percent lift in positive response rate and a 41 percent view rate. The production company’s own case study adds 1.3 million views across Facebook and YouTube combined.

Six years old and still the best argument in the Australian market that length was never the problem. Boredom was.

Savannah Sanchez runs the agency version at scale. Motion’s write-up of her production system says 200-plus ads a week across roughly 50 clients including Fabletics and Athletic Greens, though another page on the same site says 100-plus, so take the exact figure lightly. The system is the point: film Monday and Tuesday, edit Wednesday and Thursday, deliver Friday. Two to three hours of research per concept, one to two on each brief, then it moves.

The part worth stealing: when a hook structure wins for one client in one category, they abstract why it worked and brief it into a completely unrelated category. One test, many payoffs.

What the job pays in Australia

This is where I have to be straight with you about the data.

“Creative strategist” barely exists as a title in Australian salary datasets. SEEK has no salary page for it. Indeed’s Australian career page for it returns a 404. The only Australian-specific figure I could find is Glassdoor’s, and it rests on 43 self-reported salaries.

So here’s what there is, with the sample sizes attached, all in AUD:

| Role | Figure | Source | | :– | :– | :– | | Creative strategist, AU | ~$95,500 average total; base range $70,000 to $110,000 | Glassdoor AU, n=43, Aug 2026 | | Performance marketing manager, AU | ~$115,800 average total | Glassdoor AU, n=44, Mar 2026 | | Digital marketing manager, Melbourne | $113,300 / $144,200 / $154,500 (25th / 50th / 75th) | Robert Half 2026 Salary Guide | | Digital marketing manager, Sydney | $119,900 / $152,600 / $163,500 | Robert Half 2026 Salary Guide | | Creative strategist, US | ~US$130,509 average total | Glassdoor US, n≈495, Aug 2026 |

Two things to note. The recruiter guides quote base salary excluding super, so add 12 percent for a package number. And the spread between sources for the same job title is roughly $40,000, because self-reported data skews junior and recruiter data skews senior and metro. That gap is the honest story. Anyone quoting you a single precise number for this role in Australia is guessing.

The Advertising Council Australia runs a proper salary survey, 3,914 salaries across 111 agencies, but it’s restricted to participating agencies. If you’re in one, go read it. It beats everything above.

There’s no degree for this. Most people come in sideways: media buyers who developed taste after spending millions, or editors and designers who got curious about what their work actually earned. You build one pillar deep, then cross-train hard into the other.

The ceiling is higher than the title suggests. Margin literacy, consumer psychology, running a testing system, briefing humans. That’s roughly the CMO skill set.

What you can do right now

Count your output. How many genuinely new creatives went live last month? Not variations. Compare it to the Motion table at your spend level. If you’re at the bottom of your tier, nothing else on this list matters yet.

Reset the win rate expectation before you fix anything else. One in twenty. If your team feels bad about that, they’ll start playing safe, and safe ads die faster than bad ones. Make the number explicit so nobody’s ashamed of it.

Stop waiting for fatigue to show up. Meta’s own curve says the decline starts at exposure two, not exposure five. Build the refresh cycle on a calendar rather than on a dashboard alarm.

Isolate one variable this week. Same audience, same budget, one thing different. Most accounts I look at have never run a clean test, which means every conclusion in the doc is a guess wearing a suit.

Spend one hour on reviews instead of the Ad Library. Your own reviews, your competitors’ one-star reviews, the Reddit threads. Write down the ten phrases customers actually use. Those are your hooks. Highest-return hour in the whole job and almost nobody does it.

Put margin next to ROAS in your creative report. Per concept if you can. The first time an ad with a great ROAS turns out to be selling your worst-margin SKU, the conversation in your business changes permanently.

The old job was working a machine that let you steer. That machine’s gone. The new one reads your ad and decides for you.

So the only real question left is whether the thing you’re feeding it is worth watching.

One in twenty is.

If you want a look at your creative volume, your win rate and what’s actually driving the account, book a growth call with Thrive.