Your Path To Data-Driven Quality–ALM Forum 2014
I was honored to present Your Path To Data-Driven Quality at ALM Forum 2014
.
You can download the deck here:
| PowerPoint | |
You can watch the video here
Track
Changing Role of Test
Title
Your Path To Data-Driven Quality
Abstract
Bing analyzes Petabytes of data per day. Facebook instruments everything. Outlook.com analyzes web performance for millions of real users. Why? Because testing is measurement, and measurement requires data.
Testers have traditionally relied on test results, but the data sources now available to the tester and the ability to process these are expanding like never before. Data-Driven Quality (DDQ) strategies such as Testing in Production (TiP) are essential tools for most testers, but the question is how do you implement these strategies to benefit your specific product? If you are seeking this answer, Seth will show you how to move forward. The data you will need can be big or small, real-time or delayed, synthetic or organic. So it is critical that you know what data you need and understand how much is required. With this presentation, you will become equipped with the tools and knowledge you need to get insights from this data about your product quality. Seth will take you through the critical thinking and framing to understand how data-driven quality can benefit your product and team. He will then enable you to create a roadmap for how you can implement your data-driven quality strategy and take your testing to the next level.
Twitter-verse
@setheliot "Exchange team using machine learning to predict future availability issues" <- like a pre-crimes division? #almforum
— Mark Tomlinson (@mtomlins) April 3, 2014
I'm totally going to steal @setheliot's "Hadoopin 60 seconds" slide from #almforum
— Curtis Stuehrenberg(@cowboytesting) April 3, 2014
@setheliot "Detection: collecting evidence about your product or service as you progressively deploy it to the world" #almforum
— Mark Tomlinson (@mtomlins) April 3, 2014
@setheliot "DDQfeeds your questioning: How is Performance? What do users do? #almforum
— Mark Tomlinson (@mtomlins) April 3, 2014
"Downtime at the Microsoft exchange group is defined as any time our service is not up." @setheliot laying own knowledge at #almforum
— Curtis Stuehrenberg(@cowboytesting) April 3, 2014
If you have enough data you will find correlations to almost anything. #almforum @setheliot
— Curtis Stuehrenberg(@cowboytesting) April 3, 2014
@setheliot Know the difference between Correlation and Causation - sunscreen does not cause drownings. #almforum
— Mark Tomlinson (@mtomlins) April 3, 2014
GQM - Goal, Question, Metric. Figure out what you want to learn BEFORE you decide on a metric. @setheliot #almforum
— Curtis Stuehrenberg(@cowboytesting) April 3, 2014
@setheliot "Testresults -> scoring engines using Bayesian analysis" <- learn more here https://t.co/JOFnCbdE3g #almforum
— Mark Tomlinson (@mtomlins) April 3, 2014
@setheliot pulling out #Bayesian analysis. I feel like yelling "youin my house now, son." #almforum
— Curtis Stuehrenberg(@cowboytesting) April 3, 2014
Data-Driven Exploratory Testing? @setheliot say "maybe" :) updated TESTHEAD: Testing the Limits at #ALMForum: https://t.co/z8cjtVI0f8
— Michael Larsen (@mkltesthead) April 3, 2014