Machine learning and automation campaigns on Google keep going from strength to strength as they continue to adjust their algorithms and improve their automation offering. Smart Shopping is one of their newest campaign types, which selects the right bid for every user and each auction to maximise sales.
If you’re not already sold on the benefits on Smart Shopping (clearly you haven’t read our Bloom and Wild case study), leveraging Google’s machine learning algorithms can massively reduce the time you spend optimising Shopping campaigns so you can do more with your time. Not to mention it has a higher ROI than compared to standard shopping campaigns.
If you’re looking to branch out into this campaign type, we’ve put together some best practices to ensure you’re delivering the best results:
Before you start, ensure that you are using high-quality image assets; make sure conversion value tracking is correctly in place and your feed is optimised to meet best practices.
This campaign will work on delivering the highest revenue it can achieve within your budget, so be sure to assign a budget that you are comfortable to spend. That being said, you need to give it enough room to test and learn, so this spend level should match your average on Shopping & Display.
Smart Shopping will take priority over standard shopping so we would recommend pausing old campaigns and not trying to run both campaign types at the same time. If you don’t want to move all your product types across, ensure the category you are testing is running on Smart Shopping only. If you have different goals by category, it would be best to split this out.
Setting a tROAS is optional on the campaign level, specify this if you have a target to hit but keep in mind this might restrict volume.
There is a 15-day learning period so do give it time before you make any changes!
For reporting, ensure to exclude the 15-day learning period and then compare the next two-three weeks data with previous for the most accurate comparison. You may also need to exclude the last three days to account for conversion delay. Despite its many triumphs, the downside of Smart Shopping is its lack of reporting that we’re used to from a standard shopping perspective – things like audience performance or search queries.
With that in mind, you’re ready to set up your first Smart Shopping campaign! The key is to be patient within the first couple of weeks as there will be a lot of fluctuations – this is completely normal as the algorithm tries to find the best audience for you. This can be quite difficult, as we’re all quite used to making changes when we see a performance dip, but taking a step back will be key in ensuring this campaign works for you.
If all of this feels overwhelming or leaves you with one or two burning questions, please don’t hesitate to get in touch. Just drop us a message here.