Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS

AI & NLP applied to semantic targeting in Search Engine Marketing

Periodic Reporting for period 2 - OA (AI & NLP applied to semantic targeting in Search Engine Marketing)

Reporting period: 2021-04-01 to 2023-06-30

Problem:The problem addressed is the management of keywords in Paid Search ads, which requires a lot of tedious manual work, or programming complex scripts. Often keywords in the ads are managed less frequently than needed, causing poor performance of the campaigns (low volume of conversions, higher cost per lead, etc.) Until OA, there was no single technology capable of automatically analyzing million of multi-language search terms from Google for the selection of the perfect keywords and the creation of the right long-tail keywords combinations to maximize performance/cost ratios in SEM strategies.

Relevant impact in the EU society:
Our SaaS service allows companies doing Paid Search campaigns in GoogleAds to improve the performance of their campaigns and save time spent managing the campaigns. The unique selling points, compared to alternatives were:
Extremely easy to use, based on rules and not requiring scripts or coding
High performance improvements, up to 40%
End-to-end automation, covering everything from defining target metrics to uploading the terms in GoogleAds through the APIs.
Total visibility and control over the process

Objective: Our goal was to be the choice solution for advertisers (end advertisers managing their campaign internally of Advertising agencies) for the management of complex Paid Search campaigns on any network (Google, Bing. Amazon Ads)
WP1 - Project Management and dissemination


Task 1.1. Project Management (M1-M39):
We successfully achieved a high level of project coordination encompassing technical, administrative, legal, and financial management aspects. Administrative management was carried out efficiently, encompassing tasks such as organizing meetings, maintaining communication channels, and managing project documentation.


Task 1.2. Dissemination activities (M1 – M39)
To promote the project's objectives, outcomes, and impact to a wider audience, including relevant stakeholders and target groups we participated in conferences, fairs, and industry events and employed digital platforms and online communication tools to disseminate project information effectively. This included maintaining an informative project website, publishing articles and blog posts, and utilizing social media channels to reach a broader audience.


Task 1.3. Search for potential investors due to changes in equity terms and conditions (M6 -M39)
This task involved identifying and approaching relevant individuals, organizations, and venture capitalists who align with the project's objectives and can provide the necessary financial support.
We conducted market research to identify potential investors who had shown interest in similar projects. We prepared and delivered investment pitches that highlighted the project's unique value proposition, growth potential, and alignment with market trends.


WP2 - Development and integration of all functionalities in the product for Google Ads. (M1-12)
This WP aimed to finalize the developments dealing with all functionalities of the SaaS product for Google Ads. It achieved its goals by delivering a commercial product by Q1 2021.


Task 2.1. Development of Google Ads product functionalities (M1- M12)
We developed the following functionalities to automate the keyword management in Google Ads: Keyword Locator DSA, Keyword Traffic Driver, Search Term Booster, Keyword Cleaner Low Search Volume, Keyword Locator, Keyword Booster, Duplicated Traffic Cleaner Development, integrated the reporting of users' campaign metrics and developed a dashboard tracking Google Quality Score.


Task 2.2. Testing of Google Ads product functionalities (M4 – M12).
Each functionality was tested with real data and with several clients.


Task 2.3. Full integration with production systems (production stage) (M4–M12)
Once developed and tested, each functionality was integrated in production stage


WP3 - Development of the PaaS infrastructure (M13-39)
After market research concluded that there was no market demand for a PaaS service at this stage, focus was shifted to designing the future PaaS architecture, proof test the REST API publication of selected capabilities and develop additional functionalities that could be included in a future PaaS and be integrated immediately in the SaaS product.


Task 3.1.Detect services/products to solve digital marketing common challenges (M13-39).
Our research efforts were dedicated to identifying services and products that can effectively solve common challenges in digital marketing.


Task 3.2. Design and develop a public API for these services (M14 – M39).
We successfully signed the architecture of a future PaaS and created distinct endpoints, offering access to our keyword relocator service and various reporting services. The following 3 features were exposed through a REST API and integrated in the SaaS user interface for immediate use:
Task 3.3 New feature: Keyword relocator (M13-39).
We successfully designed, developed, and implemented the keyword relocator feature for the Google Ads product.
Task 3.4 New feature: Metrics Report Dashboard (M13-39)
Through the metrics reporting dashboard, clients can access real-time data and metrics related to their campaigns, such as impressions.
Task 3.5 New feature: Activity Report Dashboard (M13-39)
A user-friendly dashboard that presented activity-related information in a clear and organized manner. Clients can customize their dashboard to display the specific activities and tasks that are relevant to their campaigns and business goals.
On top of the 3 features mentioned above, other important features were added to the SaaS service, including: 1. New automation activity, Search Term Scalability, 2. Automation Report available through the UX, 3. Microsoft Ads (Bing): Search Term Analyzer, 4. New version of the service for smaller GoogleAds accounts “Lean product”, 5. White labeling: for advertising agencies using the SaaS tool in a transparent way for their clients. 6. New simplified onboarding of new clients, 100% self managed by client. 7. SaaS App Multi Language 8. Costs and Scalability improvements, 9. Patent application for Smart Locator.
WP4 - Ethics requirements.
We prepared a report with our Ethics advisor a report. We received the approval of the Ethics expert.
Progress beyond the state of the art: The solution implemented proprietary algorithms, leveraging NLP techniques, for which a Patent was filed, and delivered improvements up to 40% in performance of the campaigns, as documented by case studies in different sectors and countries
Expected results until the end of the project: The development and implementation of the new features showcased the potential to improve digital marketing campaigns and provide valuable reporting and analytics tools to clients. 26 paying clients in 9 different countries used the service during the project, including 2 out of the top 5 advertising groups worldwide.
Potential impacts: The project aimed to support SMEs and contribute to the growth and competitiveness of the EU and worldwide digital sector. The innovative solutions developed have the potential to enhance the effectiveness and efficiency of digital marketing strategies, benefiting businesses and fostering innovation within the industry.
Horizon 2020 agreement nº 946602
Logo OrbitalAds