Client:
A construction company specializing in multi-apartment residential projects in Kyiv and Kyiv region. The company is managing multiple projects at different stages of development, assigning a separate marketing team to each project.
Expected outcomes from Contextual Ad Campaigns:
- Higher conversion volume
- Reduced number of fake and non-target leads and calls
Campaign period:
- Campaign Setup: Jan. 2025
- Campaign Management since Feb. 2025 (ongoing)
Ad Campaign Performance Over 3 Months
The Issue of Spam Leads and Fake Conversions
Over time, it became clear that Google Ads’ AI algorithms occasionally expanded beyond high-quality audiences, resulting in low-value traffic such as click bots and fake leads. As a result, user data was distorted, leading the system to develop an inaccurate target audience model.
Consequently, automated bidding strategies started losing accuracy, while the number of spam conversions from paid search grew exponentially.
In fact, the solution is quite simple — importing offline data through GCLID. However, in our case, a specific constraint made this approach unworkable: the use of third-party call tracking prevents click parameters from being passed to the CRM and fed back into Google Ads.
Google Ads Conversion Data Import Flow
How to Address the Issue Without Importing Data?
Unfortunately, there is no simple solution. However, after reviewing all options we managed to identify the weak points in the system — what could be fixed and what had to be removed altogether.
Targeting display campaigns
Use a combined approach to display campaigns: audience combined with manually selected ad placements. Although reach and clicks may drop, this combination will effectively protect against spam and improve traffic quality.
Avoid using optimized targeting, especially when advertising high-ticket services or operating in a very specific niche where the system may not accurately identify the target customer profile:
Combine keyword match types
Account structuring based on the SKAG approach (one keyword per ad group) and relying on multiple match types is an outdated practice which may negatively impact search campaign performance.
When duplicate keywords are created across different match types, the data used for smart bidding becomes fragmented. As a result, this may negatively impact performance, and in our case roughly double the time required for automated bidding strategies to retrain.
Use portfolio bid strategies and a shared budget
By combining search campaigns into the group with a shared target and budget, you provide automated bidding strategies access to consolidated performance data, helping improve the quality of future conversions:
To strike the right balance between conversion volume and quality, the media mix should be built around search campaigns, the most controllable format.
Optimal Budget Allocation Formula:
- Search campaigns — 70%
- Performance Max — 20%
- Display campaigns — 10%
Under such an allocation, AI algorithms still handle a substantial share of campaign optimization, while overall account management remains completely under your control:
Project Participants:
Project Manager: Oleksandra Nesina
PPC Specialist: Oleksandr Shylin
