liveBook · Manning

Click-through rate

As explained in Machine Learning System Design

The proportion of users who click a displayed search result or other presented link.

Click-through rate (CTR) measures how many users click a displayed result or link. In search and recommendation systems, it provides a faster behavioral signal than purchases because clicks occur more often. section 5.2.2 section 5.3.2

Use CTR as part of a metric chain, not as the final business objective:

  • Treat CTR as an online metric measured while the system is live. section 5.3.2
  • Use it as a proxy for conversion rate when clicks are positively related to purchases. section 5.2.2
  • Exploit its greater volume to obtain more training examples and faster, more sensitive A/B tests. section 5.2.2
  • Check that optimizing CTR remains connected to purchases and broader outcomes such as revenue. section 5.2.2

Search results at PhotoStock

A search engine expects higher-ranked images to receive more clicks, so CTR evaluates how often users click its result links. Conversion rate separately evaluates which clicks lead to purchases; ranking metrics such as mean reciprocal rank and normalized discounted cumulative gain can provide offline proxies. section 5.3.2

Table 5.1 Examples of offline and online metrics
Offline metrics
Online metrics
Recall at given specificity for spam message classification
Number of user complaints about spam messages
Quantiles of 1.5, 25, 50, 75, 95, and 99
Value of expired items, total sales
Mean reciprocal rank, normalized discounted cumulative gain
Click-through rate on search engine result page

Compare CTR with offline ranking measures and other online business outcomes. section 5.2.2

How Machine Learning System Design covers it

How deep it goes
Explains definition, online measurement, proxy reasoning, metric hierarchy, and experimentation benefits without detailing CTR calculation.
Examples
Uses search-result clicks, image purchases, and comparisons with mean reciprocal rank and normalized discounted cumulative gain.
What you can do after
Evaluate live search behavior and apply CTR as a faster proxy when purchase labels are sparse.
How it connects
Connects click-through rate to conversion rate, A/B testing, ranking metrics, and revenue.

Also explained in

Compare how 4 books explain click-through rate →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Machine Learning System Design. The section links point to the parts of the book it draws on.