Lululemon
AI-driven paid media (Performance Max) focused on new-customer acquisition
In 2024-2025 in Canada, Lululemon used Google's AI-driven Performance Max campaigns (brand excluded) to lower its acquisition cost, improve its ROAS by 8%, and raise the share of revenue from new customers from 6% to 15%, a case documented by Google.
Key points
- Nonbranded Performance Max campaigns driven by Google's AI, in Canada, to win new customers.
- ROAS improved by 8% and acquisition cost down on nonbranded campaigns.
- Share of revenue from new customers rose from 6% to 15%.
- Measured via MMM, experiments, and attribution; case published by Google, still active in 2026 (Level B evidence).
Objective
Bring down acquisition cost on nonbranded campaigns in Canada and increase the share of revenue coming from new customers, by handing Google's AI the distribution and optimization of ads on Search and YouTube from a single campaign, with measurement solid enough to separate the incremental from the already-acquired.
The deployment
Lululemon a revu son dispositif Google Ads autour de trois chantiers : restructurer les campagnes shopping, construire un moteur d'acquisition de nouveaux clients, et solidifier la mesure. Le coeur du dispositif est Performance Max, une campagne unique ou l'IA de Google choisit les encheres, les audiences et les placements et diffuse sur Search, YouTube et le reste du reseau. Pour orienter l'algorithme vers l'acquisition plutot que vers des clients deja gagnes, l'equipe a travaille sur des campagnes nonbranded (marque exclue) au Canada et injecte la donnee first-party ainsi que la distinction nouveaux clients vs clients existants. En parallele, elle a mis en place une trifecta de mesure (Marketing Mix Modeling, experiences, attribution) pour verifier que la baisse de cout et la hausse de revenu nouveaux clients etaient reelles. Le cas a ete documente par Google et a recu une distinction aux Google Search Honours ; la strategie a ensuite ete etendue a l'international.
Results Proof B
Quantified case published by Google (Think with Google) on a named and quantified Performance Max deployment (ROAS +8%, share of new-customer revenue from 6% to 15%, acquisition cost down in Canada), corroborated by established ecommerce press (Digital Commerce 360, July 2026) that covers the same case citing Google. Quantified platform case study equals B.
How it works
Documented architectureThe stack in detail
- plateforme Google Performance Max A single campaign where Google's AI optimizes bids and placements and distributes ads across Search, YouTube, and the rest of the Google network. Used here on nonbranded campaigns (brand excluded) to go after new customers.
- plateforme Google Ads Ad account operating the restructured shopping campaigns and the acquisition engine.
- outil Measurement trifecta (MMM + experiences + attribution) Combination of Marketing Mix Modeling, experiments (geo/incremental tests), and attribution to measure the real impact and steer optimization toward new customers.
How it runs, concretely
For ops teams-
1Restructure the shopping campaigns marketing
Reorganize and clean up the Google Ads campaign structure to give the AI a sound base.
-
2Build a new-customer acquisition engine marketing
Create nonbranded campaigns (brand excluded) and set a new-customer acquisition objective, feeding in first-party data.
-
3Distribution and optimization IA
Performance Max chooses bids, audiences, and placements and distributes across Search, YouTube, and the Google network from a single campaign.
-
4Measure and recalibrate equipe data
Cross-reference Marketing Mix Modeling, experiments, and attribution to verify incrementality and redirect budget toward what really brings in new customers.
First-party data and conversions tagged as new versus existing customers. Without this clean conversion signal and the new/existing distinction, the algorithm optimizes for volume and falls back on already-acquired customers instead of chasing incremental ones.
How your customers perceive this type of use
Sourced studiesLe pricing algorithmique est le terrain le plus inflammable : 68% des consommateurs disent se sentir leses quand les marques utilisent le pricing dynamique et 80% jugent plus dignes de confiance les marques aux prix constants (Gartner, 2024). L'equite percue varie selon le secteur : le pricing dynamique n'est juge juste que par 33% a 40% des repondants selon qu'il s'agit de concerts ou de cinemas (YouGov, 17 marches). Le prix personnalise par les donnees individuelles est le plus rejete : 47% des Americains s'y opposent fermement (Consumer Reports, 2024).
Acceptance conditions
- La constance des prix comme signal de confiance : 80% jugent plus fiables les marques aux prix stables (Gartner 2024)
- Le secteur conditionne l'equite percue : le pricing dynamique est mieux tolere pour les cinemas (40% le jugent juste) que pour les concerts (33%) (YouGov 2024)
Red lines
- Le pricing dynamique percu comme abus : 68% se sentent leses (Gartner 2024)
- Le prix individualise a partir des donnees personnelles : 47% d'opposition ferme (Consumer Reports 2024)
- Les frais caches et hausses imprevues, vecus par 79% des consommateurs sur un an et associes a la perte de confiance (Gartner 2024)
Sources: Gartner 2024 · YouGov 2024 · Consumer Reports 2024
How to replicate
Inference, not sourcedData prerequisites
- Clean first-party data usable in advertising
- A new-versus-existing-customer distinction in conversion tracking
- Reliable tracking of conversions and revenue by campaign
Org prerequisites
- A performance marketing team able to run Performance Max day to day
- A measurement function that can build MMM, experiments, and attribution
- A media budget dedicated to nonbranded acquisition
Possible stack
- Google Performance Max or equivalent (Meta Advantage+)
- Marketing Mix Modeling
- Attribution / geo experiment platform
The plan, step by step
- Step 1Clean up and restructure the existing shopping campaigns to give the AI a clean base.Deliverable: A simplified, readable account structure.
- Step 2Tag conversions to distinguish new customers from existing ones, and connect the first-party data.Deliverable: An acquisition signal usable by the algorithm.
- Step 3Launch a nonbranded Performance Max campaign with a new-customer acquisition objective.Deliverable: An acquisition engine in production on Search and YouTube.
- Step 4Set up the measurement trifecta (MMM, experiments, attribution) to validate incrementality.Deliverable: A reliable read on acquisition cost and the share of new-customer revenue.
- Step 5Reallocate budget toward what actually brings in new customers, then extend to other markets.Deliverable: An acquisition setup scaled internationally.
First step: Separate new customers from existing ones in conversion tracking, then launch a nonbranded Performance Max campaign with an explicit new-customer acquisition objective.
The same pattern, proven in other industries
AAA Auto Club Enterprises
AI-driven search campaign that customizes ad text and reaches long queries beyond the purchased keywords
ADAC Autoversicherung
full-funnel paid campaign driven by AI (Performance Max) optimized for new customer acquisition
adidas
AI-driven audience prospecting and bidding (ARO) across the full funnel
Sources
- S1 Lululemon's AI performance marketing strategy Interested party archive pending
- S2 Ecommerce Trends: How Lululemon is using AI Established press archive pending
- S3 How Lululemon's AI-Driven Marketing Strategy Redefined Customer Acquisition Secondary archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.