Revenue Increase Through Google Ads Rebuild – Fitness Apparel Case Study

How we rebuilt a stalled Google Ads account from scratch - fixing tracking, feed, and campaign structure - and grew monthly revenue by 10x in 90 days.

10x

Revenue Growth

61%

ROAS Growth

+89%

Conversion Rate

Challenge

When we took over the account, we inherited a setup that was significantly underperforming. Legacy campaigns were generating minimal revenue, and the account lacked the structure, data quality, and feed optimisation needed to scale.

The issues ran deep on three fronts.

1. Feed Not Optimized for Norwegian Search Behavior

Norwegian gym-goers search for training wear in their native language, but product titles were written primarily in English – leaving the feed invisible to the most commercially relevant local queries. Without the right local-language terms in the right fields, products were simply not surfacing for the searches that mattered most.

2. Weak Conversion Signal and Insufficient Conversion History

The existing setup imported GA4 conversions into Google Ads – a method that introduces delays, attribution gaps, and signal loss compared to native Google Ads tracking. Compounding this, the account had accumulated very few historical conversions, which meant Smart Bidding strategies couldn’t be activated on most campaigns – Google’s automated bidding requires a minimum conversion threshold to function reliably, and the account simply hadn’t reached it.

3. No Campaign Segmentation

All Shopping traffic ran through a single, unstructured campaign with no intent-based split, no return on ad spend (ROAS) target differentiation, and no mechanism to prioritize budget toward the highest-converting segments.

Solution

We approached the rebuild in sequence: fix what’s invisible first – tracking and feed – then restructure for intent, then let the algorithm scale what’s working.

1. Conversion Tracking Overhaul

Before any campaign went live, we replaced the broken GA4 app-based setup with a GTM implementation, configuring a server-side purchase conversion action as the sole primary goal in Google Ads – giving Smart Bidding a clean, reliable signal from actual purchase events. All other conversion actions were demoted to secondary status. Bot detection and cookie retention features were activated to protect data quality, and a remarketing tag was deployed to begin building retargeting audiences from day one. Without clean conversion data, no bidding strategy can learn.

2. Feed Rebuild and Norwegian Keyword Enrichment

Using one of our feed management solutions, in this case, DataFeedWatch (DFW), we rebuilt the product feed from the ground up. We analyzed Norwegian search behavior to identify the local-language terms fitness consumers actually use when shopping for gym wear, then mapped those terms to the relevant product categories and embedded them into product titles. Beyond titles, we enriched the feed with additional attributes that had been missing entirely: color, material, and product highlights – giving Google’s algorithm more structured signals to match products to the right queries. In Shopping and Performance Max, the product feed is the primary lever for ad relevance – this optimization directly unlocked impression share on high-intent queries the brand had previously missed entirely.

3. Intent-Based Campaign Architecture

We paused all legacy campaigns and rebuilt from scratch. Shopping traffic was split into two dedicated campaigns – high-intent and low-intent – with a deliberate negative keyword strategy controlling traffic routing: high-intent Norwegian fitness terms were excluded from the low-intent campaign to ensure purchase-ready queries only triggered the right campaign.

Alongside Shopping, we launched a feed-only Performance Max (PMax) campaign, a controlled, feed-first setup anchored exclusively to Shopping inventory.

4. Phased Bidding: Manual CPC First, tROAS When Ready

With a newly rebuilt account and no purchase history, Smart Bidding had no signal to learn from. Rather than forcing automated bidding into a data vacuum, we launched Shopping on Manual cost-per-click (CPC), setting bids at the product-type level to accumulate performance data with full control. After sufficient conversion volume had built up – approximately two weeks – we transitioned both Shopping campaigns to target return on ad spend (tROAS) bidding, with initial targets set conservatively to gather more data. Targets were progressively raised in subsequent weeks as account performance improved.

5. Progressive Budget Reallocation

As data accumulated, we dynamically shifted the budget from Shopping to PMax based on where the conversion value was actually generated. The structure was built to respond to results, not to stay fixed.

Results

In 3 months, the rebuilt account delivered compounding growth across every core metric. Monthly revenue grew 10x compared to the pre-management baseline – a direct result of moving from a near-inactive account to a structured, data-driven setup. Conversion rate (CVR) improved by +89%, from 1.76% to 3.33%. ROAS accelerated by +61% in a single month, with the Performance Max campaign reaching 3+ ROAS by month three – while ad spend grew by only +26% over the full period.

ROAS improvements don’t come from bid adjustments alone. They come from giving the algorithm what it needs – clean conversion data, a relevant feed, and a campaign structure that matches intent. Build the foundation right, and the numbers follow.