Machine learning for Shopping is here!

This month, Google announced that Shopping campaigns can now be “goal optimised”.

This means that a Shopping campaign is now able to use AI to automatically optimise your ads to maximise conversion value to reach your ROAS target.

Another benefit of setting up campaigns in this way, is that it combines Display Re-marketing and Shopping in one campaign so that your ads automatically get tailored and shown across Google properties such as Google search results, the Google Display Network, YouTube, Gmail and Search partner websites.

We’ve noticed this helps to increase ad reach in a controlled manner, at speed, whilst generating a return. This can be a challenge to do across all these properties, in one go, otherwise.

To set up a campaign you will need:

1) A feed in Google merchant center that is linked to Adwords.

2) Remarketing tag installed (ideally configured for Dynamic ads).

3) Revenue/transaction tracking (typically this campaign works well when the existing campaigns have converted 50+ times in the last fortnight).

4) A Shopping campaign that is set up with a goal

The results are impressive

We’ve been rolling “goal optimised” Shopping campaigns through the month and have seen some great results.

The stats below show the results we have seen so far for one of our B2C customers when running the new “goal optimised” & older campaign alongside one another:

A) Shows the older Shopping campaign converting at £5.04 CPA & 860% ROAS

B) Shows the new Shopping campaign that is “Goal Optimised” converting at £2.26 & 1334% ROAS

You will notice that the impressions are significantly higher on the shopping campaign that is ‘goal optimised’. This is due to a higher number of impressions from ads showing up on the Google Display network as you would typically see with a remarketing or display campaign.

We recommend to test it for yourselves

We are seeing similar results across all customers who we have rolled this out for, so if you haven’t heard of, or executed this strategy already – we suggest you test it right away!