In June 2016, Twitter started a randomized experiment inside its real product: one group of users saw posts in chronological order while another group used a home feed ranked by a personalization algorithm. The control group covered nearly two million daily active accounts.
In June 2016, Twitter started a randomized experiment inside its real product: one group of users saw posts in chronological order while another group used a home feed ranked by a personalization algorithm. The control group covered nearly two million daily active accounts.
In June 2016, Twitter started a randomized experiment inside its real product: one group of users saw posts in chronological order while another group used a home feed ranked by a personalization algorithm. The control group covered nearly two million daily active accounts.
In June 2016, Twitter started a randomized experiment inside its real product: one group of users saw posts in chronological order while another group used a home feed ranked by a personalization algorithm. The control group covered nearly two million daily active accounts.
In June 2016, Twitter started a randomized experiment inside its real product: one group of users saw posts in chronological order while another group used a home feed ranked by a personalization algorithm. The control group covered nearly two million daily active accounts.