PERFORMANCE MARKETING · GOOGLE ADS
Google Ads restructuring to support increased investment
During my time as Performance Marketing Manager at Computrabajo, I managed Performance Marketing strategy across multiple markets and an annual global budget of more than €3 million. This case study looks at restructuring Google Ads to better understand demand and prepare the account to scale.
The issue was not just the campaigns. It was how demand was categorised.
The account had brand and generic campaigns, but search terms were not categorised and negative keywords were not managed precisely enough to prevent all overlap. This made it difficult to understand what demand was driving conversions and how the search, keyword, ad group, ad and landing page related to one another.
The challenge was to understand how demand was distributed across an account that had grown over time, then rebuild its structure so it could be interpreted more clearly.
First, I needed to understand what was behind the data
Before proposing any changes, I carried out a thorough review of the search-term history. I analysed what users were searching for, which queries converted, what demand categories existed, and which searches overlapped or were not represented in the account structure.
- Search
- Keyword
- Ad group
- Ad
- Landing page
- Conversion
- What users were searching for and what search intents existed.
- Which terms generated conversions and what demand they represented.
- Where searches were being mixed or overlapping.
- Which searches were not sufficiently represented in the structure.
A telling sign of a broader problem
Brand and generic campaigns were already in place. However, categorisation and negative keywords did not control precisely enough which searches could trigger each campaign. One of the clearest signs emerged when I reviewed a generic campaign that appeared particularly efficient: some of its conversions came from searches related to Computrabajo.
The conversions were real, but that campaign’s CPA did not reflect generic demand alone. This finding was not the whole problem; it was evidence that greater control was needed over how searches were categorised and distributed.
01
Generic campaign with a strong CPA
02
Some conversions came from brand searches
03
Demand was not classified precisely enough
The figure was not wrong; it was necessary to understand what type of demand lay behind it.
Before optimising performance, I needed to understand how demand was distributed and build a structure that made it possible to interpret it correctly.
From demand analysis to a new account structure
Based on the analysis, I decided to reorganise the entire account. The four stages show the work involved in moving from search terms to a more relevant and controllable structure.
01
Analyse
Review the historical data to identify intent, conversions, overlap and under-represented demand.
02
Categorise
Group searches by term and intent, then define campaigns and ad groups around related keywords.
03
Refine
Expand the negative keyword lists to control which searches could trigger each campaign, according to intent.
04
Rebuild
Reorganise the account with more campaigns and ad groups, and expand the keyword coverage.
The level of detail had a purpose: each ad group brought together closely related keywords, so the ad could better match the search that triggered it. The aim was not to create campaigns for their own sake, but to gain control, improve relevance and make performance easier to interpret.
This was not about moving campaigns. It was about rebuilding the structure.
The restructuring was the visible outcome of broader work analysing, categorising and refining demand.
- 01Search terms
- 02Categorisation
- 03Keywords
- 04Negative keywords
- 05Campaigns
- 06Ad groups
- 07Ads
Each layer had to fit with the next.
I expanded the keyword coverage, created new campaigns and additional ad groups, and worked extensively on negative keywords to control which searches could trigger each part of the structure. Each ad group brought together closely related terms to improve the connection between the user’s search and the ad message.
The new structure had to show that it could support growth
01 · Prepare for the peak in demand
The new structure went live before Christmas, a quieter period, giving it time to adapt before the most important month for acquisition.
02 · Enter January with the account stabilised
CPA initially increased after the structural change. As January arrived and volume grew, performance began to stabilise and the account was able to support higher acquisition volumes.
03 · Make the most of the opportunity
January was one of the periods with the highest demand of the year. Conversion volume was high enough for senior management to decide to increase investment and make the most of the account’s performance.
+17%
budget compared with the amount originally planned
↑ CONVERSIONS
more conversions than in January of the previous year
The increase in conversions should be interpreted in light of the fact that investment levels were not the same as the previous year.
A stronger foundation for scaling
The new structure made it possible to work with higher levels of investment while keeping sight of the demand the account was capturing.
- What demand are we buying?
- Which campaigns generate each type of demand?
- Where does it make sense to increase investment?
- What performance are we seeing in each category?
More than a Google Ads restructuring
01
Question an apparently strong result and analyse the demand behind it.
02
Turn search intent into a more relevant and controllable structure.
03
Connect performance analysis with investment and scaling decisions.
What I would do differently today
The granular structure made sense at the time because it provided greater control over how searches were organised and triggered. I would not replicate it in the same way today: Google Ads has evolved towards automation systems and learning models that need more room to work with signals and data volume.
For that reason, I would now use a simpler structure and leave more room for automation. I would keep campaigns separate where there are genuine differences in budget, objectives, bidding, intent or measurement, but avoid adding complexity simply to create the appearance of greater control.