NewDoes ChatGPT recommend you? Get your free AEO Rank Tracker report.Check your brand

← All case studies

Case Study

How GOOD gets a 14.8x return from Google Ads

From January to October 2026, GOOD's Google Ads returned 14.82x on platform-reported sales value, with 670 online purchases, more than twice as many as in the previous ten months. The return rose 31% from 11.31x, and each sale cost 22% less, while spend rose 75%. Syte runs Google Ads for GOOD (goodclothing.co.za), a Cape Town womenswear brand founded in 2008 that makes its clothing locally and is known for its prints and artist collaborations. The account works because it protects the brand name cheaply and uses Performance Max to put the product range in front of new shoppers.

ClientGOOD
VerticalFashion: locally made womenswear
MarketSouth Africa
ChannelsGoogle Search (brand, generic, competitor), Performance Max
Period1 January – 31 October 2026
Google Ads return14.82x (up from 11.31x)
Purchases670 (up 123%)
Click-through rate4.6% (up from 2.6%)

The challenge

GOOD is a small, independent fashion label competing online with national chains and international fast fashion. Its customers are loyal and often search for the brand by name, but a local label can't outspend the big retailers on generic searches like "women's dresses". The job for Google Ads was to make sure everyone looking for GOOD found the GOOD store first, and then to reach new shoppers who'd like the clothes but hadn't heard of the brand, without the cost per sale climbing as spend went up. Fashion also moves in drops: each new collection needs to be seen while it's in stock.

Approach

We set the account up in three layers:

  • Brand search, kept tight. An exact-match brand campaign makes sure GOOD shows first for its own name, at a low cost per click.
  • Performance Max for the range. Performance Max uses the product feed to show GOOD's clothes across Shopping, YouTube, Display and Gmail to people likely to buy.
  • Collection launches as their own campaigns. New collections, such as Kaleidoscope, get a separate Performance Max campaign so they can be pushed while stock lasts, without disturbing the core account.
  • Measured tests at the edges. Generic and competitor search run on small budgets and are paused when they don't sell.

Channels used

  • Google Search — Brand: the cheapest sales in the account, from people who already know GOOD.
  • Performance Max — Core range: the main way new shoppers find GOOD's products.
  • Performance Max — New collections: short pushes behind new drops such as Kaleidoscope.
  • Google Search — Generic and Competitor: small tests for new customers.

What we changed

  • Start of 2026: the account ran on three campaigns bidding to a target cost per sale: brand search, Performance Max and generic search.
  • During 2026: rebuilt the account into new campaigns: an exact-match brand campaign, a core Performance Max campaign, a separate Performance Max campaign for the Kaleidoscope collection, and generic and competitor search tests.
  • The original three campaigns are now paused, and so is the new generic search test, which produced no sales.

Results (platform-reported)

MeasureJan – Oct 2026Comparison period (3 Mar – 31 Dec 2025)Change
Return on ad spend14.82x11.31x+31%
Purchases670301+123%
Cost per purchasen/an/a−22%
Click-through rate4.6%2.6%+2.0 points
Conversion rate1.6%2.2%−0.6 points
Google spendn/an/a+75%

More people clicked GOOD's ads in 2026, and more than twice as many bought. A smaller share of clicks turned into sales (1.6% against 2.2%), which can happen as an account reaches more new shoppers, but each sale still cost 22% less. Monthly returns from June 2026 were 5.5x in June, 14.3x in July, 20.3x in August, 17.0x in September and 17.9x in October.

By campaign, January to October 2026:

CampaignPurchasesReturn
Brand search (original)32514.6x
Performance Max (original)20516.7x
Brand search, exact match (new)6569.3x
Generic search (original)397.7x
Performance Max core (new)299.3x
Performance Max, Kaleidoscope collection65.8x

Methodology. Figures are platform-reported (Google Ads) for 1 January to 31 October 2026 unless stated. They are compared with 3 March to 31 December 2025, the previous period of the same length in the reporting dashboard. Purchases are the conversions Google Ads records, and return is Google's recorded purchase value divided by ad cost; it is not matched to the store's own sales records. No revenue or spend amounts are shown.

What's next

Brand search is a big part of this result. It is the cheapest source of sales, but those shoppers already knew GOOD. The harder job is growing sales from new customers, so the next step is to give the core Performance Max campaign more room, plan collection pushes around GOOD's drop calendar, and check the purchase tracking against the store's own orders so the figures are confirmed on both sides.

FAQ

Is Google Ads worth it for a small South African fashion brand?

It has been for GOOD. From January to October 2026 its Google Ads returned 14.82x on platform-reported sales value, with 670 online purchases, more than twice the previous ten months.

Should a fashion brand bid on its own brand name?

Usually yes. GOOD's brand search campaigns were its cheapest source of sales in 2026, and the new exact-match brand campaign returned 69.3x. Without it, other retailers can show above you for your own name.

Does Performance Max work for clothing brands?

It can. GOOD's original Performance Max campaign returned 16.7x across 205 purchases in 2026. It works best with a clean product feed and good images.

How should a fashion brand advertise a new collection on Google?

GOOD gives each new collection its own Performance Max campaign, such as Kaleidoscope, so it can be pushed while stock lasts without disturbing the core account.

Can a brand grow sales and return at the same time?

GOOD did in 2026: purchases rose 123% and return rose from 11.31x to 14.82x, while spend rose 75%.

Related

Talk to us

Tell us your growth target. We'll show you the fastest path to revenue.

No contracts · Free first call · 48hr response