Press release: Nexoya study: Artificial intelligence boosts the performance of digital marketing campaigns by 77 percent

When human experts optimize multichannel marketing campaign budgets, they underperform AI-based tools by up to 19 percent. Compared to non-optimized budgets, artificial intelligence boosts performance by up to 77 percent.

Zurich and Berlin, September 28, 2021 – What effect does the use of artificial intelligence have in optimizing campaign budgets compared to conventional approaches? Project leader Dr. Markus Meierer from the University of Zurich investigated this question together with the startup nexoya. The result of the investigation: AI beats humans in the optimization of campaigns. The focus was on the effects of continuous optimization of digital marketing channels over an extended period of time. Several series of tests compared the performance of different digital advertising campaigns that ran simultaneously across many digital channels and were constantly optimized. Split in campaigns that are optimized automatically with the help of artificial intelligence, campaigns controlled manually by marketing experts, and a control group where no changes were made. The goal of the campaigns was to increase traffic to a website. Therefore, the number of website visitors was defined as the most important KPI to be optimized. Compared to the control group without budget adjustments, using an AI-based tool for budget optimization achieved 77 percent more visitors. If a marketing expert controlled the budgets, AI performance was 19 percent higher.

In digital marketing, advertising campaigns often run across many channels simultaneously. As a result, managing the activities is complex and time-consuming: Marketing teams need to keep track, compare results across channels, and allocate budgets efficiently. AI-based solutions can relieve marketers of this work by collecting and preparing all the data and suggesting the optimal budget
allocation.

Researchers at the University of Zurich wanted to determine whether and under what conditions AI-based tools can do this job better than humans. In collaboration with Flavia Wagner from nexoya, a corresponding study was conducted. The startup has developed a platform that bundles all organic and paid marketing metrics as well as third-party data on weather or exchange rates. In this way, marketing can evaluate and compare the performance of individual campaigns and assets at the click of a mouse. Thanks to the automated optimization of ongoing campaigns, AI algorithms provide concrete suggestions for cross-channel budget allocation.

Realistic conditions and scientific methods
“Together with nexoya, we conducted an extensive randomized experiment. The collected data shed light on the causal effect of budget allocation optimization techniques on digital marketing campaign performance. Furthermore, it was investigated whether AI-based optimization outperforms traditional approaches,” explains Markus Meierer, project leader at the Chair of Marketing and Market Research at the University of Zurich.

The study’s uniqueness: The focus was on the effects of continuous optimization of digital marketing channels over an extended period of time. Three test series were carried out for the study, each of which ran for 28 days. It was crucial that the test series depicts a realistic scenario: The marketing campaign had to be conducted exactly as it would typically run. An awareness/branding campaign was run, with all test groups using identical advertising materials. The ads were played out across multiple channels, including Facebook, Instagram, Google Display, Google Search Ads, Bing Ads, and LinkedIn. Twice a week, both nexoya’s AI software and the marketing expert optimized the budgets in each campaign. Budget allocation in the control group remained unchanged throughout the experiment.

“Optimizing budget allocation using AI increased marketing KPIs by up to 77 percent in the test series compared to constant budgets. Compared to a human expert, it showed a 19 percent increase in performance. The AI-based optimization, which proved to be the most powerful method, is based on an easy-to-implement software solution that can be quickly deployed to almost any environment and performs budget adjustments fully automatically,” says Marco Hochstrasser, CTO and co-founder of nexoya.

Images & Illustrations

When human experts optimize multichannel marketing campaign budgets, they underperform AI-based tools by up to 19 percent. Compared to non-optimized budgets, artificial intelligence increases performance by up to 77 percent. (Image: nexoya)
Download image: https://www.nexoya.com/wp-content/uploads/2021/09/UZHxNexoya-Study-Illustration-2_DE.png

The result of the study by Flavia Wagner and the University of Zurich: AI beats humans in optimizing campaigns. (Image:
nexoya)
Download image: https://www.nexoya.com/wp-content/uploads/2021/09/UZHxNexoya-Study-Illustration-1_DE.png

About nexoya
The startup Nexoya ltd with the headquarter in Zürich, is since 2018 developing a platform for automatic analysis for digital marketing of mid-to large size corporates – with high usability and high data privacy focus. On the basis of Machine Learning, the SaaS Platform “nexoya Marketing Analytics” optimizes multi-channel marketing campaigns. It aggregates all relevant metrics out of various channels such as google, instagram or facebook, allowing you to have well-arranged reporting and monitoring of your KPIs. On the basis of the gathered data and thanks to Artificial Intelligence (Predictive Analytics) nexoya optimizes your campaign budget thereafter. With that it helps Marketing teams to reduce their operational efforts, simplify automated data-driven decisions and reduce your ad-spend by up to 30%. Nexoya is only gathering non personalized data and is hosting strictly in certified datacenters in Switzerland. Customers are well-known businesses such as Emmi Group, Generali, CSS or the house agency. www.nexoya.com

Press contact
Sabrina Ortmann, nexoya
Phone: +49 (0)30 54909240
E-Mail: press@nexoya.com