How HelloFresh Optimizes 100% of Meta Ad Spend With Genie Agents on Databricks
Read the Portuguese translation →
Author: Databricks · Watch on YouTube
Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.
Original language: English
Original video: How HelloFresh Optimizes 100% of Meta Ad Spend With Genie Agents on Databricks · Channel: Databricks · 01:06 · Watch on YouTube
Topics: AI agents, Task and workflow automation, Data, databases and embeddings, AI governance and regulation
Summary
Section titled “Summary”HelloFresh describes how Databricks helped it address data silos that had slowed decisions about advertising spend. Genie and its agents now optimize all of the company’s Meta ad spend, while Unity Catalog has made data access easier to manage across the business. The company also reports substantially less time spent on maintenance and reconciliation, faster ad testing, and a one-click launch process. HelloFresh sees this as one of several use cases it can now unlock with Genie.
Key points
Section titled “Key points”- HelloFresh operates eight brands across three continents and describes itself as the world’s leading meal kit provider.
- Before the changes, data silos and manual data gathering could delay decisions about spending by weeks.
- Genie and its agents have gone from optimizing about two-thirds of Meta ad spend to optimizing 100%.
- Unity Catalog simplified how HelloFresh defines and shares data access across the business.
- Maintenance work that took 40 to 50 hours a week can now be done in one or two hours.
- Reconciliation has been reduced from a week to under an hour on Monday morning.
- HelloFresh says it is testing ads four times faster than before.
- A launch process that used to involve seven people now happens in one click.
- HelloFresh views this Genie use case as one of many it can now unlock.
Chapters
Section titled “Chapters”| Start | Chapter | Summary |
|---|---|---|
| 00:00 | HelloFresh’s data challenges | HelloFresh introduces its business and explains how data silos and manual data gathering slowed advertising decisions. |
| 00:33 | Optimizing Meta spend with Genie | The company describes scaling Genie and its agents to all Meta ad spend and simplifying data sharing with Unity Catalog. |
| 01:06 | Faster operations and future use cases | HelloFresh highlights faster reconciliation, ad testing, and launches, and says more Genie use cases can be unlocked. |
Quotes
Section titled “Quotes”We had data silos everywhere. — 00:00
Today, Genie and our agents are optimizing 100% of it. — 00:33
Work that used to take us 40 to 50 hours a week, our colleagues are now able to do within one or two hours. — 00:33
We’ve gone from a week of reconciliation to under an hour on a Monday morning — 01:06
What used to be a seven-person launch process now happens in one click. — 01:06
Concepts and entities
Section titled “Concepts and entities”HelloFresh Group, Databricks, Genie, Unity Catalog, Meta, agents
Full transcript (original language)
Section titled “Full transcript (original language)”00:00 [music] >> HelloFresh Group is a fast-moving consumer goods [music] company and the world’s leading meal kit provider. We are operating eight brands across three continents. >> [music] >> Databricks has empowered HelloFresh to get the most out of our data. Something that we just observed. We had data silos everywhere. We had people that were like really scrambling data together from everywhere and that was taking sometimes weeks until decision was made on how we [music] would actually spend money.
00:33 >> We started this with about two-thirds of our meta ad spend. Today, Genie and our agents are optimizing 100% of it. Unity Catalog transformed our ability to share data across the business because it dramatically simplified how we were defining access to data in the first place. >> Maintenance work for [music] keeping things up and running, taking care of the integrations, debugging certain topics when things were not going well. Work [music] that used to take us 40 to 50 hours a week, our colleagues are now able to do within one or two hours. >> We’ve gone from a week of reconciliation
01:06 to under an hour on a Monday morning [music] and we are testing ads four times faster than before. What used to be a seven-person launch process [music] now happens in one click. >> What excites me most about the future is this was the primary use case and also the most impactful one that we’ve been running so far [music] specifically on Genie. This use case is only one of many that we are able to unlock now. >> [music]