Domino's · Abnormal AI
41 hours — time saved on manual email investigations and remediation
Vendor-reportedRetail & e-commerce · Fraud detection · Classical ML · In production
Fraud detection
Valvoline Inc. · Abnormal Security
Vendor-reportedRetail & e-commerceUnited StatesIn productionEnterpriseClassical ML
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by abnormal.ai (vendor self-report). Checked against the source page on 2026-09-06. 1 of 1 figures below appear on that page word for word.
The problem
Valvoline faced persistent email attacks, including spear phishing and invoice fraud, despite using multiple legacy security solutions, leading to significant manual effort for their security team.
What was deployed
- Vendor
- Abnormal Security
- Products
- AI Security Mailbox
- Technique
- Classical ML
- Build or buy
- Bought off the shelf
- Deployment
- SaaS
- Data used
- Valvoline's email environment data
- Scale
- 9,100+ Protected Mailboxes
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
2Legacy email security solutions replaced
“legacy solutions (SEG + API) replaced by Abnormal”
Quoted word for word from the source
Cost saved by stopping a recurring invoice fraud attack
“Saved by stopping a recurring invoice fraud attack”
Quoted word for word from the source
Analyst hours saved per month on email and solution managementper month
“Analyst hours saved per month on email and solution management”
Quoted word for word from the source
reducedTime spent by analysts on email-related tasksfrom 90% of each day for four to five analystseach day
“We'd had four to five analysts spending 90% of each day on email-related tasks; Abnormal allowed us to get away from that manual work”
Quoted word for word from the source
“Attackers already use AI for greater impact, so you need solutions that improve as attacks evolve. Abnormal’s been ahead of the behavioral AI game, and as attacks become more refined, Abnormal will be there to help us.”
Difficulties and limits
- Despite using three solutions (a SEG, Microsoft 365’s native security, and an API-based solution), email attacks still consistently reached inboxes
- Persistent spear phishing and invoice fraud, ranging from $5,000 to millions of dollars
- Cybersecurity awareness and training alone were not enough, since humans will always make mistakes
Sources
- abnormal.ai — vendor self-report
Similar deployments
Perry Ellis · Yotpo
1.4% — conversion rate
Independently reportedRetail & e-commerce · Customer support · NLP · Scaled
Amazon · Amazon
10% more — shopper spend
Independently reportedRetail & e-commerce · Process automation · Computer vision · Scaled · 2024
Zalando
increased by 40% — basket sizes
Independently reportedRetail & e-commerce · Personalization · Classical ML · In production
Walmart · Microsoft
22% — global ecommerce growth
Independently reportedRetail & e-commerce · Search & retrieval · LLM · In production · 2024
Kingfisher Group
more than 10% — e-commerce sales originating from product recommendations
Independently reportedRetail & e-commerce · Personalization · Classical ML · Scaled · 2023