Advertising agency design is not one job. It is at least three jobs that use the same software and almost nothing else in common. Brand design decides what a company looks like for the next five years. Campaign design decides what an idea looks like for the next quarter. Performance creative decides what a single ad looks like for the next eleven days before it is replaced. Hiring one person, or one agency, and expecting all three at the same standard is where most Singapore SMEs go wrong.
The short version, if you only read this paragraph. Brand work is slow, expensive and infrequent. Campaign work is a medium sized project with a clear beginning and end. Performance creative is a factory, and it should be judged on throughput and learning rate, not on how any individual asset looks. In our experience, the businesses that get the most out of design are the ones that stop treating these as a single line item.
This article covers how design sits inside an advertising agency, how the three types differ, how creative gets briefed, tested and iterated, what a creative retainer costs in SGD versus project work, and the argument that in 2026 the number of creative variants you can produce matters more than the polish on any one of them.
What advertising agency design actually means in Singapore
Singapore's agency market has an unusual split. On one side sit branding studios that produce genuinely excellent identity work and then hand you a PDF. On the other sit performance shops that can produce forty ad variants a week but whose output looks like it was assembled from a template library. Very few operations sit comfortably in the middle, which is why so many local businesses end up with a beautiful brand book that their Meta ads bear no resemblance to.
The practical consequence for a Singapore SME is that you need to be explicit about which job you are buying. Asking a studio for "design support" is how you end up paying brand rates for social assets, or paying performance rates and being disappointed that the result does not feel premium.
There is a second local factor worth naming. Singapore campaigns frequently run in more than one language and across more than one market. If the same campaign runs in Singapore, Malaysia and Indonesia, every asset multiplies. A design system that handles that gracefully is worth more than a beautiful one-off, and this is a question worth asking any agency before you engage them.
The three types of design and how they differ
| Type | Time horizon | Typical output | Judged by | Typical SGD range |
|---|---|---|---|---|
| Brand design | 3 to 5 years | Identity, system, guidelines, templates | Consistency and distinctiveness | 12,000 to 60,000 project |
| Campaign design | 1 quarter | Key visual, adaptations, master assets | Idea clarity and coherence | 6,000 to 25,000 project |
| Performance creative | 1 to 4 weeks per asset | Ad variants, statics, short video, UGC edits | Cost per result and learning rate | 3,000 to 12,000 per month |
| Always-on social | Continuous | Feed posts, carousels, stories, reels covers | Consistency and posting cadence | 1,500 to 5,000 per month |
| Web and landing pages | 1 to 3 months | Page design, components, responsive states | Conversion rate | 4,000 to 30,000 project |
Brand design
Brand design answers a question that has nothing to do with this quarter: what should this company look and sound like so that it is recognisable and coherent everywhere it appears. Logo, colour, type, photographic direction, iconography, tone of voice, and crucially the rules that let other people apply it without you.
The output that matters is not the logo. It is the system. When we audit a brand handover, the first thing we look for is whether the guidelines actually tell a junior designer how to build a 1080 by 1350 Instagram post at 9pm on a Tuesday. Most do not. They show three moodboard pages and a logo exclusion zone diagram, and then everyone improvises for the next four years.
Campaign design
Campaign design takes a strategic idea and gives it a visual form for a defined period. A key visual, a headline treatment, a set of adaptations across formats. This is closer to traditional advertising craft, and it is genuinely hard to do well because it has to be distinctive without abandoning the brand system.
Campaign design fails in Singapore for a predictable reason: the campaign is designed for a hero placement, usually a large horizontal format, and then squashed into vertical mobile placements at the last minute. We recommend designing the vertical mobile version first and adapting upwards, because that is where the majority of impressions will land.
Performance creative
Performance creative is a different discipline that happens to look like design. The unit of work is the variant. The goal is not the best possible ad, it is the fastest possible route to knowing which message and which visual format work, and then producing more of that.
Performance creative teams work with a testing matrix, not a moodboard. They will produce four hooks by three visual treatments by two calls to action and let the platform sort them out. The individual assets often look unremarkable in isolation. The system produces results. This is the work we run through our social media ads service, and it is deliberately structured as a pipeline rather than a series of projects.
How creative actually gets briefed, tested and iterated
Bad briefs are the most expensive thing in design. A designer given "make it modern and clean, similar to Apple but for accounting software" will burn two weeks producing work that gets rejected for reasons nobody can articulate. Here is the structure we use.
The brief
A usable creative brief for advertising work fits on one page and contains eight things.
- The single message. One sentence. If you cannot write it, the design will not fix it.
- The audience and their current belief. Not demographics. What do they currently think that we need to change.
- The proof. The specific reason to believe. A number, a guarantee, a comparison.
- The placement and format list. Exact aspect ratios and platforms, decided before design starts.
- The mandatories. Logo, legal, offer terms, promotional dates.
- What this is competing against. Screenshots of what appears next to it in the feed.
- The success measure. Cost per lead, click through rate, view through rate, brand recall.
- The variants required. How many hooks, how many visual routes, how many sizes.
Point six is the one most often missing and the one that changes the work most. A designer who has seen the feed your ad will appear in makes different decisions to one who has only seen your brand guidelines.
The test
Once the variants exist, testing has to be structured or you learn nothing. We test one dimension at a time at the top of the funnel and hold everything else constant.
- Round one tests the hook. Same visual treatment, four different opening messages. This is where the biggest differences appear, usually by a wide margin.
- Round two tests the visual route with the winning hook fixed. Product-led versus person-led versus text-led.
- Round three tests format. Static versus motion versus user generated style.
- Round four tests the call to action. Usually the smallest effect of the four, which is why it goes last.
The order matters. Testing calls to action first, which is what most accounts do because it is easy, wastes the budget on the dimension with the least variance.
The iteration
Iteration is where retainers earn their money. Once a hook wins, you do not simply run it until it dies. You produce five siblings of the winner, varying only slightly, and feed them in as the original fatigues. We have seen a single winning concept sustain a Singapore account for five months through sibling production, where the original asset alone would have exhausted its audience in about three weeks.
Copy and design are inseparable in this loop. A hook is a piece of writing that a designer sets. If your agency separates these into different departments with a handover between them, iteration slows to a crawl. We keep copywriting and design in the same working session for exactly this reason.
Retainer versus project: what each costs in SGD
The single most common commercial question we get is whether to buy design as a retainer or as projects. The honest answer is that it depends on whether your need is continuous.
Project work makes sense when the deliverable has a clear end state. A brand identity. A website. A campaign key visual with a fixed adaptation list. You are buying a defined artefact, and you should pay for it as one. Singapore project rates for identity work typically run SGD 12,000 to SGD 60,000 depending on scope and the size of the applications list. A campaign key visual with adaptations sits closer to SGD 6,000 to SGD 25,000.
Retainers make sense when the need is continuous and the volume is predictable. Performance creative is the clearest case. If you are running paid social year round, you need new assets every two to four weeks forever, and buying that as a series of small projects means paying a project overhead every time.
Singapore creative retainer bands, based on what we scope and deliver:
| Monthly retainer (SGD) | Typical monthly output | Best suited to |
|---|---|---|
| 1,500 to 2,500 | 12 to 20 static assets, light adaptation | Organic social only, single market |
| 2,500 to 4,500 | 20 to 40 assets including 2 to 4 short videos | Paid social at modest spend, one market |
| 4,500 to 8,000 | 40 to 80 assets, 6 to 10 videos, monthly concept round | Multi-channel paid, 2 markets or 2 languages |
| 8,000 to 15,000 | Full creative pipeline, weekly test rounds, dedicated team | Spend above SGD 30,000 a month, multi-market |
A caution on the low band. Below roughly SGD 1,500 a month you are buying template population, not design. That is a legitimate thing to buy if you know that is what you are buying, but do not expect concept work or testing discipline at that level. If ongoing feed content is the actual need, our content creation service is scoped for that specifically rather than dressed up as strategic design.
One more commercial note. Photography and video production are usually not included in a design retainer and are quoted separately, because the cost drivers are completely different. A shoot has crew, location, talent and equipment costs that have nothing to do with design hours. We handle those through our production team, and for physical goods specifically through product photography. Get a shoot right once and it feeds a design retainer for six months.
A common mistake we see
The mistake is spending the entire creative budget on one hero asset.
It is an understandable instinct. Design is expensive, so you concentrate the money on the most important thing, get it perfect, and run it everywhere. Ten years ago this was correct. The media plan had few placements, production was slow, and one great execution genuinely carried a campaign.
In reality, the economics have inverted. Platform algorithms now do most of the targeting work, which means the creative is the variable you actually control. Audiences see far more advertising per day, so fatigue arrives in weeks rather than months. And the ad account rewards the account that gives it more options to choose between.
We have run the comparison directly on client accounts. One perfectly finished hero asset, produced over three weeks, against twelve competent variants produced in the same three weeks from the same concept. The variant set wins on cost per result almost every time, and it wins by a margin that is not close. More importantly, the variant set tells you why it won, because you can see which hook and which format carried it. The hero asset tells you nothing except whether it worked.
This does not mean polish is worthless. There are two places where it still pays.
Brand assets that persist. Your logo, your website, your recruitment materials, your pitch deck. These are seen thousands of times over years and cheapness compounds. Spend properly here.
Category signalling. If you sell premium services, visibly cheap creative undermines the price. The fix is not to spend three weeks on one asset. It is to build a template system that makes all forty variants look considered by default.
That is the resolution. Put your craft budget into the system and the design language, then run the variants through it at volume. A common mistake we see is teams doing the opposite, improvising a system per asset and then wondering why forty pieces of content took four weeks.
Web design is the exception that proves the rule. A landing page is seen by every single person your ads reach, so the return on getting it right is enormous. We treat website design as a craft project even for clients whose ad creative we run as a factory.
Case study: a Singapore F&B group stuck on one hero asset
One Singapore food and beverage group we worked with had a well-executed brand refresh from a local studio and a genuinely good hero key visual for their quarterly promotion. They ran that key visual, resized into four formats, across Meta and Google Display for eleven weeks. Media spend was around SGD 9,000 a month.
Performance for the first three weeks was decent. Cost per purchase sat at SGD 18.40. By week seven it had drifted to SGD 31.20, and by week ten it was SGD 41.80 on flat spend. Frequency had climbed above 6 in their core audience. Their read on this was that the market was saturated. It was not. The creative was.
We kept the brand system and the concept entirely intact and changed only the production model. From the same shoot assets and the same key visual, we built a template kit with locked type, colour and layout rules, then produced 14 variants in the first fortnight: five hook variations in the headline, three visual crops, two motion versions, and four user generated style edits from existing footage. Nothing new was shot. Total additional design cost for that first round was under SGD 3,000.
Over the following eight weeks on the same SGD 9,000 monthly spend, cost per purchase settled at SGD 19.60, roughly back to launch efficiency. Frequency in the core audience fell to 2.8 because the account had more options to rotate. The clearest finding was in the hook test: the variant naming a specific price point outperformed the brand-led headline by a factor of just over two on cost per purchase, which is a piece of information the single hero asset could never have surfaced.
They now run a rolling creative retainer at the SGD 4,500 to SGD 6,000 band and refresh the variant pool every three weeks. Their brand still looks like their brand, because the system holds it together.
Field Notes
Observations from our own creative delivery in Singapore. Treat these as planning benchmarks.
- Creative fatigue window: 14 to 28 days for paid social in Singapore at moderate spend. Above roughly SGD 15,000 a month in a single market it compresses towards 10 to 14 days. Plan production capacity against that, not against your campaign calendar.
- Hook variance dwarfs everything else. Across our test rounds, the spread between the best and worst hook on the same visual is routinely 2x to 4x on cost per result. The spread between call to action wordings is usually under 15 percent. Test in that order.
- A template system pays for itself in about 6 weeks. Building a locked kit takes us roughly 12 to 20 design hours. It typically cuts per-asset production time by 50 to 70 percent, which recovers the investment inside two production cycles.
- Sibling assets extend a winning concept by 8 to 16 weeks. Producing five close variations of a winner is the cheapest performance gain available in a creative retainer.
- One production day feeds roughly 3 to 6 months of design output when it is shot with a variant plan rather than a single hero shot in mind. This is the highest leverage decision in the whole process and it is made before the camera is switched on.
- Realistic monthly output at SGD 4,500: 40 to 60 finished assets across formats, including 6 to 10 short videos, plus one concept round. If a quote at that level promises 200 assets, you are being sold template population.
Frequently asked questions
Should we hire an in-house designer or use an agency? It depends on volume and variety. If you need 40 or more assets a month in a consistent style, an in-house designer at Singapore market salary is usually more economical and faster. If you need concept work, multiple disciplines, and periodic production, an agency spreads that cost across specialists you could not hire individually. Many of our clients run both, with an in-house designer executing and us handling concept, testing and production.
How many creative variants do we actually need per month? As a working rule, one new concept every four to six weeks and five to eight variants per concept in market at any time. At paid spend below SGD 5,000 a month you can go lighter. Above SGD 20,000 a month you will need considerably more, because fatigue arrives faster at higher frequency.
Does design or targeting matter more now? Design, by a clear margin, for most accounts. Platform targeting has consolidated into broad automated audiences, which means the creative is doing the qualifying. This is one of the biggest shifts of the last few years and it is why creative budget deserves a larger share than it used to.
What should be in a proper brand handover? Editable master files, a font licence you actually own, a colour system with digital and print values, at least six built templates for the formats you use most, and written rules for the situations the guidelines do not cover. If you are evaluating agency partners on this, our guide to choosing a content and SEO agency covers similar due diligence questions.
How do we know if our creative is the problem? Look at frequency and at the trend in cost per result on flat spend. If frequency is climbing above 4 in your core audience and cost per result has drifted up 40 percent or more with no other changes, it is creative fatigue rather than market saturation. If your ads have never had a hook test, that is where to start regardless. Paid search creative follows different rules again, which we cover on our search engine marketing page.
Where to start
If your creative is beautiful and your results are drifting, the answer is almost never a better hero asset. Build the system, then run volume through it. If your creative is fast and cheap and your brand feels thin, the answer is to invest in the system rather than in individual assets. The system is the thing that carries both.
We build creative that has to survive a media buy, which means a design language strong enough to hold together across forty variants and a production model quick enough to keep feeding the account. Have a look at our digital marketing services for how we scope creative alongside media, read more about how we work, or talk to us about what your account is actually short of.
Natasha Tan is the founder of Digital Marketing Singapore, a full-service SEO and digital marketing agency based in Singapore. With hands-on experience across SEO, paid media, and content strategy, she works directly with Singapore businesses to build organic visibility and generate consistent leads. Natasha specialises in the Singapore market — including local search behaviour, PDPA compliance, and government grant navigation for SMEs.
