Case Study
Sotheby’s: building the growth stack for a 275-year-old auction house
Douglas Cirillo was Assistant Vice President, Director of Digital Product and Growth Stack at Sotheby’s from 2017 to 2020. He directed the rearchitecture of Sotheby’s digital and growth stack, including the migration from Adobe Experience Manager onto a composable, data-first stack, running the function with one direct report and a vendor team of about 20, directing $5 to $6M in spend. Organic traffic rose 167%, search click-through rose 48.7%, consignment search placement doubled, and consignment ad spend fell 50%.
The challenge
Search was not carrying its weight. The customer data underneath was fragmented across systems, so search, personalization, and measurement were all working from fragments.
How I led it
With one direct report, I could not mandate change. I had to earn it. We started with one team, built with them, shipped visible wins, and let other teams ask to join.
What I directed
Data foundation first. I directed the implementation of Twilio Segment as Sotheby’s customer data platform, connecting customer interaction data to the BigQuery warehouse the team called Alpine, and rebuilt the martech architecture around it in about seven months. We moved the content platform from Adobe Experience Manager to Brightspot. Gangverk, as my vendor, built the auction engine and catalogue platform under my direction. The auction platform’s code name was Viking. Conductor ran organic search intelligence; Algolia powered on-site and in-app search and recommendations; Optimizely ran experimentation; Zaius handled customer data activation. Amplitude came in as the analytics layer, trial first, then enterprise. At scale, Segment was configured across 23 systems, with about 400 million API calls a month.
I ran it through vendors, not headcount: one direct report, about 20 vendor staff, and clear ownership of what got built and in what order. Years later I joined Gangverk, the firm I had directed as a client.
The results
Organic traffic up 167%. Search click-through up 48.7%. Consignment search placement doubled, and consignment ad spend fell 50%. Sessions rose 90%, and unique visitors rose 119%. The model and the stack still stand and have evolved since I left.