Marketing ops · Production

Burberry harnesses Labelbox and Databricks to curate strategic marketing assets

The problem

Burberry uses thousands of images from multiple sources for global marketing campaigns and must classify them precisely to drive the right action by the right audience. Training its object detection and classification models required labeling massive volumes of images — a task that was not feasible to do manually at the company's scale.

First attempt

Burberry first tried an open source image annotation tool that lacked an interface to read from their data sources, required storing all images locally on a data scientist's machine, and could only save progress in a JSON file — making it impractical at the company's scale.

Workflow diagram · grounded in source
1
Image import from Amazon S3
trigger
“The lightbulb moment was when we saw how easily we could import our images to Labelbox from our Amazon S3 bucket via API”
2
Image annotation via Labelbox
ai_action
“Image annotation projects that used to take months now take just hours”
3
Edge case detection and label correction
validation
“finding edge cases using Labelbox's Model product that boost model performance. This is accomplished by having an interface to find and fix label errors in the data that will most impact model results”
4
CRANE engagement scoring
ai_action
“Marketing users can now upload candidate marketing images into our proprietary in-house-built Content Ranking Engine (CRANE) app and score them against pre-trained models that predict the engagement these images will get in upcoming camp…”
5
Image ranking and insights output
output
“Stakeholders can see a ranking of the images they've uploaded and gain insights that help them make better decisions for the current marketing campaign”
6
Revenue tracking for continuous improvement
feedback_loop
“We're not only helping our stakeholders choose the best images to use in campaigns but also tracking the revenues generated by the emails they send. This information influences each batch of imagery we produce and leads to continuous imp…”
Reported outcome

Image annotation projects that previously took months now take hours, and insight generation was reduced from two months to a two-hour self-service process.
Burberry saved 10 headcount compared to building in-house, achieved a 70% improvement in time savings for generating insights from images, and has seen 4 years of decreasing total cost of ownership with Databricks.

Reported metrics
Headcount saved vs. building in-house10
Time savings for generating insights from images70%
total cost of ownership (TCO) decrease period with Databricks4 years
Image annotation project timemonths to hours
Show all 6 reported metrics
headcount saved vs. building in-house10
time savings for generating insights from images70%
total cost of ownership (TCO) decrease period with Databricks4 years
image annotation project timemonths to hours
insight generation time before Labelboxtwo months
insight generation time after Labelboxtwo hours
Reported stack
LabelboxDatabricks Lakehouse PlatformAmazon S3CRANEDatabricks
Source
https://labelbox.com/customers/burberry-customer-story/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Image annotation projects that previously took months now take hours, and insight generation was reduced from two months to a two-hour self-service process.

What tools did this team use?

Labelbox, Databricks Lakehouse Platform, Amazon S3, CRANE, Databricks.

What results were reported?

Headcount saved vs. building in-house: 10; Time savings for generating insights from images: 70%; total cost of ownership (TCO) decrease period with Databricks: 4 years; Image annotation project time: months to hours (source-reported, not independently verified).

What failed first in this deployment?

Burberry first tried an open source image annotation tool that lacked an interface to read from their data sources, required storing all images locally on a data scientist's machine, and could only save progress in a…

How is this marketing ops AI workflow structured?

Image import from Amazon S3 → Image annotation via Labelbox → Edge case detection and label correction → CRANE engagement scoring → Image ranking and insights output → Revenue tracking for continuous improvement.