Runs experiments on people. Lubricated by olive oil: https://teapaw.org/t.co/wA5Vwu1wlk
Melbourne, Australia
Joined January 2008
- Tweets1.1K
- Following962
- Followers549
- Likes4.8K
Rob Kingston retweeted
How are effects of online A/B tests distributed? How often are they not significant? Does achieving significance guarantee meaningful business impact?
We answer these questions in our new paper, “False Discovery in A/B Testing”, recently out in Management Science >>
Using @home_assistant to get updates when Victoria's / Melbourne's COVID cases rise or fall in your suburb.
gist.github.com/kingo55/4e92… cc @philhawthorne
Me: **intimately details a GA misconfiguration affecting Safari users on a third-party booking platform**
Support: "Your users should use Google Chrome instead." ¯\_(ツ)_/¯ #measure
Rob Kingston retweeted
Today @dapperdrop, @allmywant and @robkingston release Mojito JS Delivery v2.3.0 with a bunch of optimisations and a generous 0.29KB saving in container weight. Read the blog post: mojito.mx/blog/2020/06/15/re…
Do you perform ramping when launching split tests?
@lukasvermeer was kind enough to share a brilliant method for cleanly ramping experiments using hash functions and beating Simpson's paradox.
Here's how it works: mintmetrics.io/experiments/p…
Quite proud of the work @mintmetrics_io team has done to make the Mojito Chrome extension for previewing split tests. Stunning light/dark designs by @SzilvieK, UX by @dapperdrop, all implemented flawlessly by expert developer @allmywant - chrome.google.com/webstore/d…
Rob Kingston retweeted
“Large-scale testing is not a technical thing; it’s a cultural thing that you need to fully embrace. [...] How willing are you to be confronted every day by how wrong you are? And how much autonomy are you willing to give to the people who work for you?”
hbr.org/2020/03/productive-i…
When (and how) to use bandits for split testing - @mgershoff demystifies bandit algorithms & the stats that power them: docs.google.com/presentation… #measure
Rob Kingston retweeted
Here's how we safely deploy A/B split tests using @Bitbucket Pipelines for CI: mintmetrics.io/mojito/deploy…
Rob Kingston retweeted
Inspired by @SimoAhava excellent post on using reCaptcha with GA (simoahava.com/analytics/impr…) @paul_boocock has prototyped using it for bot detection with @snowplowdata discourse.snowplowanalytics.…. Would love feeedback on if we should turn this into a formal integration?
At our local Web Analytics Wednesday last night, I shared some of our worst and most flawed split tests with the dodgiest data. Here are the slides so you can avoid running into these problems for your own experimentation:
mintmetrics.io/wawmelbourne/… #measure #optimization
Rob Kingston retweeted
1/3 Good exercise for an experimentation class.
Look at the figure below from @LaunchDarkly and find at least two flaws. Here are three:
1. Variation A, called baseline, shouldn't have a stat-sig number.
Just launched - experimentation add-on. Now its even easier to quantify the improvements your teams are shipping. Learn more: buff.ly/2MIadj2
Howdy @stefwill - I'm bit of a WAW fan here! Missed out on the last one, but would be very keen to do a talk if you're looking for speakers?
Rob Kingston retweeted
*opens an article about Google Analytics*
Sees a sentence "The beauty of Google Analytics is that numbers don’t lie."
*closes article*
Mojito Docs are now live! If you get stuck on any of the documentation, hit up @dapperdrop or me. We want to make it easy to use.
Snowplow is doing their first Melbourne meetup! Who else is joining? meetup.com/Snowplow-Analytic… #measure