BigQuery 中的 Google Analytics 站点搜索

Question

I am currently trying to pull 1.000.000 different search queries in the next 12 months from a webshop in Google Analytics. Since you only can show 5.000 queries per page and the loading times and the manual export in Excel files are very troublesome, I am searching for a solution via Google BigQuery.

So I activated "Google Analytics API" in the Google Cloud Platform and am stuck now. Is there a fast and easy way to get a spreadsheet with the following?

1. Every single query
2. In the last 12 months
3. From a specific data view
4. optional: export as Excel

I would very much appreciate your support that would save me hours and hours of manual exports.

Cheers

Answer 1

This feels like cheating, but if you need a simple solution that does not require programming or any setup, you can use the Google Analytics Query Explorer .

This is a technology demo that demonstrates the use of the Google Analytics Reporting API, but it works well to get data from the analytics accounts that are connected to the Google Account your are logged in with in your browser.

If you visit the sites you will see a number of dropdowns - the first three allow you to select an account/property/view (you need the view, the other options are just to filter down to view level).

Then you can select a timeframe, metrics and dimensions. You need to select a metric (since you plan to query a hit level dimension you probably want to use page views. You have to select a metric even if you do not plan to use it) and a dimension - the relevant dimension is "ga:searchKeyWord".

Then you can click "run query", and download the result as tsv (tab separated values).

The API will return 10 000 results per query, but unlike the Analytics interface, this is blazing fast, so even with manual paging by changing the start index (or by changing the timeframe) downloading 1 mio results will take only a few minutes.

If you are doing this regularly, you probably want to run an API script and do this programmatically, but for a quick result the query explorer should do.