Data Stack Considerations: Build vs. Buy at Tout
Download as PPTX, PDF1 like642 views
The document discusses data stack considerations for Tout, focusing on whether to build or buy an analytics solution. It highlights the transition from a complex, unreliable home-grown system to a more flexible, best-of-breed approach using Treasure Data and Looker, resulting in a 50% reduction in operational costs. The new system aims to deliver better insights and meet the reporting needs of both customers and internal stakeholders.
Data Stack Considerations: Build vs. Buy at Tout
- 1. Data Stack Considerations: Build vs. Buy at Tout July 14, 2016
- 2. Housekeeping • We will do Q&A at the end. • You should see a box on the right side of your screen. • There is a button marked “Q&A” on the bottom menu. • We will be recording this • We will send you the recording tomorrow. • We will also send you the slides tomorrow. RecordingQ&A
- 3. Jad DeFanti VP, Technical Operations John Hammink Developer Evangelist Ming Liu Sr. Software Architect Erin Franz Alliances Analyst 2 Meet Our Presenters David Banker Director, Product Management
- 4. 4 AGENDA 1. 2. 3. 4. Quick intro to Treasure Data Quick intro to Looker Build vs. Buy at Tout Q & A
- 5. Treasure Data makes ….
- 6. WHO WE ARE
- 7. WHO WE ARE
- 8. BEFORE • Data loss • Inflexible • Silos • Complex Integrations • High Costs • High Time-To-Market ETL Team EDW Team BITeam End User System
- 10. OPTION 2 AFTER
- 11. Make it easy for everyone to find, explore and understand the data that drives your business. 2
- 12. A DATA PLATFORM THAT DOES BOTH Find, explore and understand all the data Explore Everything Find, explore and understand all the data Create Standards Define your data and business metrics Any SQL Database Analyze all of your data where it is stored Build a Data Culture Anyone can ask and answer questions How is pipeline for Q4? Will we meet our revenue targets? Which campaigns convert best? Which rep is converting best? Which customer is at risk? Can we speed up our operations?
- 13. THE TECHNICAL PILLARS THAT MAKE IT POSSIBLE 100% In Database Leverage all your data Avoid summarizing or moving it Modern Web Architecture Access from anywhere Share and collaborate Extend to anyone LookML Intelligent Modeling Layer Describe the data Create reusable and shareable business logic
- 14. Leverage a Looker Block or App or quickly implement a custom deployment for a complete analytics solution Looker makes TD data accessible and interactive for anyone from Finance to Marketing to Customer Service Leverage the TD platform in real time with Looker to get actionable insights to anyone Full flexibility over defining data at query time with LookML w/o modeling ahead of time. Best of Breed Fast Time to Value Analytic Flexibility Data for Everyone Simplified Analytics, Powerful Data Discovery.
- 15. 1 5 How Treasure Data & Looker saved us from our Analytics stack and vendor 1 5
- 20. 2 0 Analytics & Tout Platform usage growth • Tout utilized both a turnkey analytics solution and an internal home-grown service • Analytics events and costs skyrocketed in Q4 2015
- 21. 2 1 Legacy systems issues • Single source vendor • Home grown solution not reliable and difficult to support
- 22. 2 2 What problems were we trying to solve with analytics? 2 2 repetitive, long term unpredictive, short term customers' predefined dashboards customers' ad- hoc reports requests internal routine reports internal ad-hoc reports external, direct impact internal, indirect impact foundation foundation
- 23. 2 3 What were our demands and costs? 2 3 customer demand internal demand direct & indirect Engineering / OPEX cost vendor cost our options
- 24. 2 4 What to build? buy one-stop solution (blackbox) build custom solution (whitebox) rent and integrate specialized components (greybox) less flexible less time / effort more expensive more flexible more time / effort less expensive Blackbox, Whitebox or something in between?
- 25. 2 5 What were the attributes of analytics reports required? Charts & Report s Aggregated Data Raw Data smaller, faster, comprehensible bigger, slower, incomprehensible flexible, detailed rigid, brief
- 26. 2 6 How we fixed the problem • Completely re-think data analytics Best of breed over single vendor and home-grown solutions • Outsource as much as possible Scale-Out over Scale-Up
- 27. 2 7 What does our solution look like? Data Collection & Storage Reporting Database Data Analytics/Visualization Primary Owner Treasure Data MySQL / Redshift Looker Infrastructure Vendor & Customer Customer Vendor
- 28. 2 8 How does it work? Amazon S3 Amazon Redshift Looker Treasure Data coordinator events scaleable receiver Sinatra Puma Rack TD logger Fluentd Tout’s logic TD output Tout’s config Tout’s logic
- 29. 2 9 Keeping up with the business Customers will adopt our services if we provide better • Daily Reports • Monthly Reports • Better insights into Tout platform usage data
- 31. 3 1 Deliver Customized Insights Quickly
- 32. 3 2 Room to Evolve Where do we go from here? • Customized and generic platform usage reports • Predictive revenue analytics • Analytics now open to all stakeholders • Upsell Additional Analytics Reporting
- 33. 3 3 Summary • Tout now meets its customers’ and internal stakeholders’ reporting needs • Tout analytics OPEX has been reduced by 50% • Tout Engineering & Operations can now focus on applications & infrastructure most relevant to our value to customers
- 34. Q&A 12
- 35. THANK YOU FOR JOINING Recording and slides will be posted. We will email you the links tomorrow. See you next time! Next from Looker Webinars: Winning with Data on July 28 See how Treasure Data and Looker work on your data. Visit treasuredata.com and looker.com/free-trial or email discover@looker.com.
- 36. THANK YOU!