Measuring impact when you cannot hold out a control group
How many engineers will you commit to internal tool buildouts?
Don't let these violations invalidate your experiment results
A holistic comparison of statistical methods for online experimentation
Ensure core metric definitions are in sync across experimentation in Eppo, BI, and other data platforms, all managed via GitHub
Accurately measure the cumulative impact of your experimentation program
Holdouts help measure the cumulative, long-term impact of an experimentation program. But getting value out of holdouts requires scale and maturity that not all organizations are ready for.
When does it make sense to bandit, and when to experiment?
Some of my personal takeaways and highlights from a slightly bigger, seemingly more diverse CODE conference.
You can’t just say “we should take a scientific approach.” That won’t work.
Optimizing search ranking algorithms can drive millions of dollars in revenue. Here's how to A/B test them.
Are mutually exclusive experiments necessary, or dangerous? Here's how to run them in Eppo.
It’s time for experimentation tools to integrate directly with the CMS instead of trying to imitate them
Shailvi Wakhlu explains why machine learning and AI products require experimentation to quantify success.
Eppo makes CUPED widely available, allowing teams to run experiments up to 65% faster than before.
How we designed Eppo Reports to facilitate a shared experimentation journey across an org
Create visually compelling, fully contextualized PDF reports built to communicate experiment results org-wide.
Why experiments are necessary to evaluate LLMs - and how you can easily A/B test between various models with Eppo.
Metrics are the vehicle that drives change in data-driven organizations.
Bayesian and frequentist approaches are fundamentally different, so why do they sometimes yield the same results?