Cienna Designs
- Client
- Cienna Designs
- Role
- Google Ads / Web Development
- Year
- 2024
- Scope
- Google Ads / Web Development / Ecommerce
The problem
An Australian brand, women owned, making sustainable bohemian womenswear of the kind people buy to travel in. The Google Ads were running close to break even and the owner said as much: to be profitable she needed 250% or 300%, and she wanted to scale from there rather than keep paying for the privilege of standing still. The previous manager had never set up a branded campaign at all, having told her the brand was not well known enough for one to be relevant yet.
The approach
Audit before touching anything, because an account this close to break even is usually losing money somewhere specific rather than everywhere at once. Then rebuild what the audit found: tracking that could be trusted, keywords grouped by intent, audiences segmented, and the campaign structure reset around what the store actually sells rather than around the old account's history.
The work
Full account audit and a strategy built on it, conversion tracking audited so the numbers could be relied on, keyword research and grouping, audience segmentation by demographic and interest, and campaigns restructured across platforms. The Merchant Center feed optimised product by product and segmented by what was earning clicks and conversions. Creative assembled from the brand's own organic photography and video rather than from stock. A branded campaign set up, the one the previous manager had refused, which became one of the best performing campaigns in the account. A/B testing across landing pages, headlines and other assets.
The result
Cost per conversion fell from $25.99 to $9.59 and conversion value rose from $398 to $6.89K, taking the return from 104.54% to 1,289.51% on a budget that moved only from $381 to $534. The client had been close to giving up on Google Ads entirely and raised the budget instead. She was also told plainly that a month like that is not a monthly event, and that the budget should leave room for testing rather than assume it repeats.
Next case study
Tali & Tasi