Delivery Hero was scaling rapidly and acquired multiple brands over the years, which led to high complexity in their marketing organization. While high-level budgets and targets are defined by the central HQ team, campaign optimization decisions are taken by the brands and their local teams. This created varying needs for data and insights across the organization.

The central team needed a unified view of all brands, regions, and respective channels to assess how brands are pacing against budgets and targets. This view needed to be flexible enough for the central team to dive deep, uncover what’s working across brands and geos, and share best practices. Local and regional teams on the other hand, including channel marketers and strategists, required granular insights to steer their campaigns and budgets effectively and optimize signups and order metrics. 

To provide these insights, Delivery Hero had heavily invested in a state-of-the-art data tech stack, which involved hundreds of marketers, analysts, and strategists across various regions and brands. Over time, this setup led to significant challenges:

Dependency on the data team: Marketers required actionable insights to effectively steer their campaigns and budgets and optimize signups and order metrics. To do this, they often had to add new data segments, perform ad-hoc analyses, and integrate new channels. However, their BI setup required them to submit tickets for the data team to prioritize. This could often take weeks or even months. As a result, onboarding new team members, launching reports for new channels, or quickly executing experiments became difficult. Many hypotheses were left unexplored because marketers grew frustrated with the long wait for BI support.

Limited data accessibility: As marketers often couldn’t wait weeks for their tickets to be worked on, they resorted to Excel and the ad platforms for decision making. Thereby they typically lacked access to MMP, MTA, incrementality, and internal data. If they didn’t want to spend hours downloading and preparing data on spreadsheets, they had to rely on ad platform data only. This led to a decrease in decision quality and high manual workloads for performance marketers. 

Costly reporting maintenance: Delivery Hero’s traditional BI stack required significant time and effort to maintain. This created a lot of work for the data team. Instead of working on strategic projects, they had to build new dashboards, update queries, or add new custom dimensions. As a result, they struggled to find sufficient time for strategic projects like MMM, incrementality tests, and other products such as bid and cost automations.