Aggregating Search Impression Share (and other metrics)

[Prototype.co]

We are trying to retrive and aggregate data using the Google Ads API (v5). The goal is the following:

*Retrieve and/or calculate aggregate Search Impression Share (and Top Impression %, Abs Top Impression %), segmented by campaign type (Search, Shopping, and a Total Aggregate).*

Note that this is not an outlandish thing to do, basically we are looking for the data available on the Google Ads online interface, right at the bottom of the campaigns table, in the dropdown of "Total: Account": "Total: Search", "Total: Shopping", "Total: Display" and so on.

Now, the customer resource is often used to query aggregate data at the account level. Clicks, Impressions, CTR and the like. But the cutomer resource does NOT have advertising_channel_type available as a resource field. It would have been too easy.

Ok, the campaign resource comes to the rescue. After all, this resource does have advertising_channel_type, and it is fairly easy to sum clicks, impressions, costs. It is also possible to work out CTR and CPC by calculating Clicks/Impressions and Cost/Clicks on the summed data. But what about search impression share?

- Averaging search impression share across campaigns is an obvious no-no.

- Using impressions-weighted averages seems to be the right thing to do, and it does seem to work for search campaigns, but falls short for shopping campaigns and account totals.

- There is another method outlined here: http://reprisemedia.com.au/blog/why-you-are-measuring-impression-share-wrong/ Basically divide impressions by search impression share to get an "available impressions" column, and then do SUM(Impressions)/(SUM(Avail.Impressions). This also seems to work fair enough for search campaigns, but falls apart totally when Shopping comes to play.

So, to sum things up...

Do you have any idea how we can get the following data (everything in red, but particularly the ones in black) using the Google Ads API:

In preferably as few steps and as few (and as small) queries as possible?

Thanks for your help!

[Prototype.co]

Sorry, inline images didn't work. Here's the illustration as an attached image.

Thanks!

[Zweitze]

You should try adding segmentation on ad network type, and check if the numbers make sense. For me it always did, but I left out results for shopping campaigns...

Aggregating Impression shares was discussed quite often on this forum, use the search function to get my insights.

For more information on my suggestion to segment on ad network types, see this topic.

[Google Ads API Forum Advisor Prod]

Hi,

You can use the campaign.advertising_channel_type field to filter the total for Display, Search and Shopping type campaigns. If you are looking for a more Shopping campaign dedicated statistics report, aggregated at several product dimension levels, you can use the Shopping Performance View instead. As for search impression share aggregation, you can refer to the thread which was pointed out earlier in this discussion.

If you wish to learn more regarding search impression share and other metrics' aggregation, let me know so I can notify the Google Ads Product support team as they are the more appropriate team to provide guidance on how metrics are calculated and aggregated.

Best regards,

Peter Laurence Napa Oliquino Google Ads API Team

ref:_00D1U1174p._5004Q275sJF:ref