SlideShare a Scribd company logo
Copyright © 2015 Splunk Inc.
Splunk at Sephora
2
Evan Wayne
Vice President
Store Systems and IT Operations
3
Agenda
Sephora Overview
My Background
Improving the In-Store Experience
Splunk for IT Ops and Sephora.com
What’s Next
Key Takeaways
4
The Science (and Background) of Sephora
Founded in 1970 in France
One of the largest U.S. cosmetics retailers,
largest in France
– 1900 stores in 29 countries, 380 in the US
– Largest store – Sephora.com
Owned by LVMH Moët Hennessy Louis Vuitton
North American Headquarters in San Francisco
5
My Background and Role
Joined Sephora in April 2013
IT team supports in-store, online and corporate
– Mobile POS on iPod, POS, COLOR IQ, Store Digital
– Sephora.com & M.Sephora.com
– Security and IT infrastructure (network, data center)
I guarantee you have a great shopping experience
My idea for a Splunk T-Shirt
– “Splunk guarantees we’ll check you out!”
6
COLOR IQ
7
MPOS is a Strategic Initiative
Implemented MPOS to:
– Have more places for customers to purchase
– Save floor space
– Reduce lines at the registers
Benefits of MPOS
– Improved customer experience
– More revenue per square foot
8
MPOS Should Not be a POS
Many errors occurred during year 1
– Coding issues
– Card-swipe failed
– Network hoping
– Battery Issues
No visibility into problems
– Log data was localized to each retail location
– “Needle in a haystack”
18% of all transactions failed
9
Why Splunk
Be the first to know before your business
Centralized view of distributed store systems
Ability to correlate issues to quickly determine root cause
Give more people the ability to assess problems
10
Splunk for In-Store Customer Experience
Mobile POS/POS/Color IQ/mPlus
Uses
– Monitor application health & performance
– Centrally analyze, troubleshoot device and store
issues
 POS server and register failures
 Type of error (how often down to device/store)
 Charging/Battery/Screen Dimming Issues
 SSID hopping (we’re in malls!)
 Card-swipe failed
– Track usage and version upgrades
– IOS Console Logging via Splunk
11
MPOS Dashboard
MPOS Transactions - Top Stores MPOS Error Codes – Start Up
12
The Good, The Bad and The Beautiful
The good - fast time to value
– Time to actionable data – 2 Weeks
The bad –discovered a lot of issues after 2 weeks
– Splunk uncovered about 40 new problems with our MPOS
The beautiful – Splunk helped crush the problems
– Failed transaction rate fell by 95% (<1% failure rate)
– MPOS store penetration doubled
– Revenues through MPOS doubled
The bonus – Quickly & personally determine problem
validity when a store calls
13
Splunk for Sephora IT Ops
Sephora US (retail, ecommerce, supply chain & corporate)
Network & Security
– Security Information & Event Management (SIEM) tool
– Corporate firewalls – proactive alerts on excessive session count and repetitive denies
– Internet traffic – WAN acceleration, web filters and Advanced Persistent Threat (APT) solutions
Enterprise Services
– Single Sign on (PING Identity)
 All corporate and retail applications migrating towards SSO strategy
Enterprise Service Bus
– Application logging and correlation for over 150 interfaces
1
14
Splunk for Sephora.com
Sephora.com/ Loyalty Program/ Mobile Web/ Mobile App/ Beauty Insider
– (Engineering, DevOps, Production Support)
50gigs of logs per day
• Real-time dashboarding for:
• Session count
• Order Per Minute (OPM)
• Exception reporting/Traffic Monitoring
• Application health (Apache -> back end systems -> DB)
• Integration check points (JCPenney, Intl Shipping, giftcards)
• Proactive monitoring for site optimization & trend analysis
• Fraud detection in real time (orders per second/# of failed logins)
• Mobile traffic patterns cross-channel
1
15
Splunk Infrastructure at Sephora
3 Splunk Search Heads*
3 Splunk Index Servers
2 Forwarders
16
Top Takeaways
Splunk extends beyond internal IT to the customer experience
Splunk has value for both pre-sales and purchase use cases
Don’t need a large # of users to get a huge benefit from Splunk
The size of that benefit can be in the millions of dollars
17
AHA! Moment
“Splunk has helped us move from ambiguity to
transparency”
18
What’s Next
Retail & Supply Chain
– Physical registers (NCR devices) in stores
 Example: Are reboots manual or forced/abnormal
– New POS Solution
– Scan guns – Retail Inventory (1500 devices) and Supply Chain (1200 devices)
– All retail IQ solutions (skincare, fragrance, lip) and Online Reservations
Sephora.com
– Order Management System (OMS – IBM Sterling)
IT Operations
– ServiceNow (ITSM) integration to create executive friendly incident & resolution
dashboards
– Splunking VMWare & Active Directory
Thank You
1

More Related Content

Sephora Customer Presentation

  • 1. Copyright © 2015 Splunk Inc. Splunk at Sephora
  • 2. 2 Evan Wayne Vice President Store Systems and IT Operations
  • 3. 3 Agenda Sephora Overview My Background Improving the In-Store Experience Splunk for IT Ops and Sephora.com What’s Next Key Takeaways
  • 4. 4 The Science (and Background) of Sephora Founded in 1970 in France One of the largest U.S. cosmetics retailers, largest in France – 1900 stores in 29 countries, 380 in the US – Largest store – Sephora.com Owned by LVMH Moët Hennessy Louis Vuitton North American Headquarters in San Francisco
  • 5. 5 My Background and Role Joined Sephora in April 2013 IT team supports in-store, online and corporate – Mobile POS on iPod, POS, COLOR IQ, Store Digital – Sephora.com & M.Sephora.com – Security and IT infrastructure (network, data center) I guarantee you have a great shopping experience My idea for a Splunk T-Shirt – “Splunk guarantees we’ll check you out!”
  • 7. 7 MPOS is a Strategic Initiative Implemented MPOS to: – Have more places for customers to purchase – Save floor space – Reduce lines at the registers Benefits of MPOS – Improved customer experience – More revenue per square foot
  • 8. 8 MPOS Should Not be a POS Many errors occurred during year 1 – Coding issues – Card-swipe failed – Network hoping – Battery Issues No visibility into problems – Log data was localized to each retail location – “Needle in a haystack” 18% of all transactions failed
  • 9. 9 Why Splunk Be the first to know before your business Centralized view of distributed store systems Ability to correlate issues to quickly determine root cause Give more people the ability to assess problems
  • 10. 10 Splunk for In-Store Customer Experience Mobile POS/POS/Color IQ/mPlus Uses – Monitor application health & performance – Centrally analyze, troubleshoot device and store issues  POS server and register failures  Type of error (how often down to device/store)  Charging/Battery/Screen Dimming Issues  SSID hopping (we’re in malls!)  Card-swipe failed – Track usage and version upgrades – IOS Console Logging via Splunk
  • 11. 11 MPOS Dashboard MPOS Transactions - Top Stores MPOS Error Codes – Start Up
  • 12. 12 The Good, The Bad and The Beautiful The good - fast time to value – Time to actionable data – 2 Weeks The bad –discovered a lot of issues after 2 weeks – Splunk uncovered about 40 new problems with our MPOS The beautiful – Splunk helped crush the problems – Failed transaction rate fell by 95% (<1% failure rate) – MPOS store penetration doubled – Revenues through MPOS doubled The bonus – Quickly & personally determine problem validity when a store calls
  • 13. 13 Splunk for Sephora IT Ops Sephora US (retail, ecommerce, supply chain & corporate) Network & Security – Security Information & Event Management (SIEM) tool – Corporate firewalls – proactive alerts on excessive session count and repetitive denies – Internet traffic – WAN acceleration, web filters and Advanced Persistent Threat (APT) solutions Enterprise Services – Single Sign on (PING Identity)  All corporate and retail applications migrating towards SSO strategy Enterprise Service Bus – Application logging and correlation for over 150 interfaces 1
  • 14. 14 Splunk for Sephora.com Sephora.com/ Loyalty Program/ Mobile Web/ Mobile App/ Beauty Insider – (Engineering, DevOps, Production Support) 50gigs of logs per day • Real-time dashboarding for: • Session count • Order Per Minute (OPM) • Exception reporting/Traffic Monitoring • Application health (Apache -> back end systems -> DB) • Integration check points (JCPenney, Intl Shipping, giftcards) • Proactive monitoring for site optimization & trend analysis • Fraud detection in real time (orders per second/# of failed logins) • Mobile traffic patterns cross-channel 1
  • 15. 15 Splunk Infrastructure at Sephora 3 Splunk Search Heads* 3 Splunk Index Servers 2 Forwarders
  • 16. 16 Top Takeaways Splunk extends beyond internal IT to the customer experience Splunk has value for both pre-sales and purchase use cases Don’t need a large # of users to get a huge benefit from Splunk The size of that benefit can be in the millions of dollars
  • 17. 17 AHA! Moment “Splunk has helped us move from ambiguity to transparency”
  • 18. 18 What’s Next Retail & Supply Chain – Physical registers (NCR devices) in stores  Example: Are reboots manual or forced/abnormal – New POS Solution – Scan guns – Retail Inventory (1500 devices) and Supply Chain (1200 devices) – All retail IQ solutions (skincare, fragrance, lip) and Online Reservations Sephora.com – Order Management System (OMS – IBM Sterling) IT Operations – ServiceNow (ITSM) integration to create executive friendly incident & resolution dashboards – Splunking VMWare & Active Directory

Editor's Notes

  1. If you have any of the AHA moments, could you add? Could be other team members I added this one from our discusison or other