In the inventory, you will find the primary search intent for each topic and an indication if that intent is fractured (ie. there are multiple intents associated with that topic).
On the details page (click on the inventory row to see it) you’ll find a breakdown of the search intent for the topic.
Primary intent is the main intent serviced by the majority of organic results on the first page of the SERP. It shows a percentage and the type of intent.
Fracture shows the remaining intents (if any) along with their percentage.
Currently, we support the following intents:
- Info – Know: The intent is to acquire some information assumed to be present on one or more web pages.
- Info – know simple: The query has a very simple answer usually presented in a knowledge card. Example: The president of the United States.
- Transactional: The intent is to perform some web-mediated activity. Like sign-up, download, buy, ….. etc
- Info – comparison: The intent is to acquire some information to compare entities.
- Local: Dominated by a combination of local packs, geographic markers of local businesses.
- Website Query: An informational query about a certain website that has the intent of navigating that website.
Our deep learning user intent classification machine learning model is based on a dataset of thousands of queries. It takes an active learning approach, continually collecting new data and improving its abilities.