English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Online experimentation platforms collect user feedback at low cost and largescale. Some systems even support real-time or near real-time data processing,and can update metrics and statistics continuously. Many commonly used metrics,such as clicks and page views, can be observed without much delay. However,many important signals can only be observed after several hours or days, withnoise adding up over the duration of the episode. When episodical outcomesfollow a complex sequence of user-product interactions, it is difficult tounderstand which interactions lead to the final outcome. There is no obviousattribution logic for us to associate a positive or negative outcome back tothe actions and choices we made at different times. This attribution logic iscritical to unlocking more targeted and efficient measurement at a finergranularity that could eventually lead to the full capability of reinforcementlearning. In this paper, we borrow the idea of Causal Surrogacy to model along-term outcome using leading indicators that are incrementally observed andapply it as the value function to track the progress towards the final outcomeand attribute incrementally to various user-product interaction steps. Applyingthis approach to the guest booking metric at Airbnb resulted in significantvariance reductions of 50% to 85%, while aligning well with the booking metricitself. Continuous attribution allows us to assign a utility score to eachproduct page-view, and this score can be flexibly further aggregated to avariety of units of interest, such as searches and listings. We providemultiple real-world applications of attribution to illustrate its versatility.

Continuous Attribution of Episodical Outcomes for More Efficient and Targeted Online Measurement