How Ysabel Mora established a robust foundation for future growth by integrating Robyn open source Marketing Mix Model as a key measurement tool.
The Challenge
Ysabel Mora lacked a clear understanding of the return on ad spend (ROAS) generated by each of its paid media marketing channels. The evolving digital landscape, characterized by increasingly stringent privacy regulations and the subsequent limitations on tracking technologies, further complicated the accurate measurement of campaign effectiveness. Discrepancies between data reported by ad platforms and the company's existing analytics tools created significant challenges in drawing reliable conclusions and making informed strategic decisions.
Furthermore, according to the “last click” attribution figures used as the main measurement approach, Ysabel Mora exhibited a strong reliance on paid search campaigns utilizing its own brand as keywords. This reliance necessitated a deeper understanding of the intricate interplay and cross-channel influences between the rest of paid media initiatives and branded search campaigns. Also, Ysabel Mora needed to gain a comprehensive strategic understanding of the business baseline and the relative significance of paid media channels within the overall trajectory of business growth.
Finally, the initial project assessment identified several critical measurement gaps. These included the need to quantify the impact of owned media channels, specifically email marketing campaigns, on overall business performance. Additionally, the project aimed to determine the significant influence of promotional activities, such as sales and discounts, on sales volumes and overall marketing effectiveness.
The Solution
The Meta Marketing Science Team partnered with Kraz to execute a Measurement Project aimed at addressing the faced challenge. Marketing Mix Modeling (MMM) was identified as the optimal analytical framework to address the project's primary objective: determining the Return on Ad Spend (ROAS) of each paid media channel. Furthermore, the MMM approach offered a comprehensive understanding of key marketing performance indicators beyond ROAS, including:
Establishing the Business Baseline:
Defining the underlying sales performance in the absence of any marketing activity, providing a crucial benchmark for evaluating the impact of marketing initiatives.
Quantifying Email Marketing Impact:
Measuring the contribution of email campaigns to overall sales, providing valuable insights into the effectiveness of this owned media channel.
Assessing the Influence of Promotional Activities:
Analyzing the impact of sales, discounts, and other promotional activities on sales volume and overall marketing effectiveness.
To gain a deeper understanding of the intricate relationship between paid media channels and branded search volume, the Meta Marketing Science Team and Kraz decided to develop a secondary MMM model.
This new model employed branded search volume as the target variable, utilizing the same dataset as the primary sales-focused model, but excluding variables related to paid branded search campaigns. This approach enabled the isolation and quantification of the impact of other paid media channels on the generation of branded search volume.
This approach enabled Kraz to quantify both the direct and indirect sales impact of paid media campaigns for Ysabel Mora:
Direct Sales Impact:
This refers to the measurable sales generated directly by paid media channels.
Indirect Sales Impact:
This encompasses the influence of paid media campaigns on increasing branded search volume, which subsequently drives additional sales through paid search channels."
The MMM methodology has demonstrated its robustness and resilience to evolving tracking technologies and evolving privacy regulations. This inherent stability positioned MMM as a viable and reliable long-term measurement solution for Ysabel Mora. Consequently, the project paved the way for the potential adoption of MMM as a standardized measurement tool within Ysabel Mora's ongoing analytical toolkit.
The impact
After implementing Robyn, the open source MMM code, Ysabel Mora realized:
Enhanced Return on Ad Spend (ROAS):
The optimized budget allocation across existing channels resulted in an 11% increase in ROAS.
Direct and indirect impact of paid media:
The analysis successfully measured the combined, or "double effect" of the primary paid media channels on sales (direct and indirect impact on sales). The direct impact measured the effect of paid media in sales whereas the indirect impact measured the impact of paid media in branded search volumes (that also impacted on sales in a second step)
Measuring the overestimation of the impact of branded paid search:
The overestimation of the ROAS of Paid search campaigns was detected, identifying the part of the branded search caused by the paid media campaigns. Specifically, the model concluded that paid media channels generated 37% of the branded search volumes that were monetized into sales via specific branded search paid media campaigns.
There are two key figures of this case:
