Case Study Spotlight: Creating Google Cloud Marketing Mix Models That Deliver for FTD

Subscribe to our monthly newsletter to get the latest updates in your inbox

As one of the world's most trusted online floral retailers, FTD took an industry-leading approach to prepare for a cookieless future. It took a giant leap forward by using advanced data engineering to consolidate its data across its many online and offline touchpoints across radio, linear, connected TV (CTV), social, search engine marketing (SEM) and programmatic display channels into Google Cloud and BigQuery. 

Ultimately, FTD wants to use its data to understand how to reach its target audiences and deepen relationships with them using their preferred channels in the most cost-effective and efficient way possible. To accomplish this, its team needs access to the insights through an intuitive, prescriptive interface.

So, the company turned to Adswerve to help it build a new privacy-centric attribution model that could use FTD’s data to help its marketing team determine the most effective media mix. The model needed to meet today’s (and tomorrow’s) privacy standards, understand its customers’ preferences and show how to optimally allocate its marketing budget across the company’s many online and offline touchpoints.

Developing Custom Marketing Mix Models

With off-the-shelf Marketing Mix Modeling (MMM) products, you can’t see how the models work or customize them, and they often take an enormous amount of time and effort to implement. But using Google Cloud gave Adswerve the power, speed and flexibility to seamlessly tailor models to FTD’s needs.

Using Google Cloud technology, Adswerve data scientists built a multifaceted, durable modeling solution to determine which media touchpoints will impact FTD’s revenue the most.

FTD’s data is in an advanced state, which made it easy for Adswerve to review and validate its Google Analytics dataset. Adswerve created, trained and optimized a model using Python to help FTD understand how its media spend affects sales and to learn more about its spend allocation across media.

Adswerve’s data scientists also used machine learning (ML) to predict which channels will reach select audiences and allow the FTD team to target its messaging to the right customers at the right times without oversaturation.

Once the model was up and running, the Adswerve team developed channel attribution, adstock effects and diminishing returns visualizations for the FTD team to see the insights the model gleaned from the data.

Finally, Adswerve developed a proprietary budget tool that offers FTD weekly model-driven insights with recommended budget allocation. The easy-to-use app is available to FTD’s entire marketing organization and secured using Google Cloud. 

How the Model Impacted FTD’s Bottom Line

Adswerve’s Google Cloud modeling solution delivered value right from the start. When it first tested the modeling solution, FTD determined its television campaign spend wasn’t optimal. Since then, the marketing team has reduced its TV campaign budget by 30% while reaching the same audience. FTD was able to reallocate the savings into other channels that had more opportunity for growth.

Moving forward, this new, durable technology will drive FTD’s data-informed decisions about media purchasing. Every member of its marketing organization can easily access and seamlessly use Google Cloud’s machine learning models and insights using Adswerve’s app to make quick, revenue-boosting media decisions.

Interested in creating advanced modeling solutions for your brand or agency? Let’s talk.