Hi all,
The Analysis team would like to formally announce the addition of a couple of metrics to the Search Metrics dashboard http://searchdata.wmflabs.org/metrics/.
First, we added the Augmented Clickthroughs (the metric formerly known as User Satisfaction): http://searchdata.wmflabs.org/metrics/#kpi_augmented_clickthroughs This is an average between the clickthrough rate and the proportion of user sessions' dwell times passing a pre-specified threshold. Going forward, we plan on obtaining qualitative user-provided feedback which will allow us to use this metric to estimate User Satisfaction.
Second, we're now using survival analysis methods to estimate and summarize page visit times: http://searchdata.wmflabs.org/metrics/#survival These metrics are the times at which we lose 10%, 25%, 50%, and 75% of our users on pages they got to from the search engine results page. For example, consistently only half of our users make it past the 75 second mark.
We'd like to extend a special thanks to Erik Bernhardson, who was crucial in the implementation of the data collection mechanisms that enabled us to construct these metrics.
Cheers, Mikhail Popov on behalf of the Analysis team
Nice work!
Dan
On 5 October 2015 at 11:33, Mikhail Popov mpopov@wikimedia.org wrote:
Hi all,
The Analysis team would like to formally announce the addition of a couple of metrics to the Search Metrics dashboard http://searchdata.wmflabs.org/metrics/.
First, we added the Augmented Clickthroughs (the metric formerly known as User Satisfaction): http://searchdata.wmflabs.org/metrics/#kpi_augmented_clickthroughs This is an average between the clickthrough rate and the proportion of user sessions' dwell times passing a pre-specified threshold. Going forward, we plan on obtaining qualitative user-provided feedback which will allow us to use this metric to estimate User Satisfaction.
Second, we're now using survival analysis methods to estimate and summarize page visit times: http://searchdata.wmflabs.org/metrics/#survival These metrics are the times at which we lose 10%, 25%, 50%, and 75% of our users on pages they got to from the search engine results page. For example, consistently only half of our users make it past the 75 second mark.
We'd like to extend a special thanks to Erik Bernhardson, who was crucial in the implementation of the data collection mechanisms that enabled us to construct these metrics.
Cheers, Mikhail Popov on behalf of the Analysis team
discovery mailing list discovery@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/discovery