Help! When Google Shopping Performance Tanks
I get it.
You're a busy in-house marketer managing all the channels, and Google Shopping had been working great for you, so you weren’t paying much attention to it as a channel source of revenue. It just kind of… worked.
Then, performance hit the pooper and the CEO is yelling at you in one ear, and Google is shrugging their shoulders in your other ear.
What can you do?
Well, there are hundreds of ways Google Shopping can break, but here are some of the first things I look to when we’re running an audit on dropped performance in Google Shopping.
There are many others, but if you rule these out, you’ve probably ruled out 90% of things that happen to an account!
1) Product Feed Dysfunction?
The first place to look, is in Google Merchant Center. I like to check out the Diagnostics Tab especially. I mean, this is an obvious one, but should always be the first place you look.
Make sure to look at the sometimes hidden Account Issues Tab as well (in the screenshot to the right of "Item issues").
- Do you have more Product Disapprovals than normal?
- Have your highest selling products been disapproved?
- Is there a drop in product count?
- Are your key selling products still in the feed?
This will allow you to immediately identify whether an obvious feed issue is the culprit. If that’s the case, you’ll have to work with your feed provider to discover the source of the problem, but at least it gives you a reason as to what’s going on.
2) Product History Reset in GMC?
This is a big’un but not everyone is aware of it.
If you see a Performance Drop for no clear reason, check to make sure your Item IDs have not been reset, or the feed container has not been reset.
Google has confirmed multiple times that there is some sort of product history tied to the item ID in the feed, so changing that ID can reset the product history. This can be good for a problematic product that keeps getting disapproved, but bad for a high performing product with years of great history to show for it.
Also, we learned in the past year from Ginny Marvin (Google’s new ad liaison) that resetting the feed container also wipes product history in that account, even with the same item IDs. So investigate whether a developer recently changed something up in GMC without you knowing it.
3) Issue with Specific Products?
It may seem obvious, but sometimes we get lost in the campaign data and forget to investigate the products themselves. Dig into what products are actually dropped. I think a really easy way to do this is in either the UI or the Reports tab at the ID level. Simply pull a report of products during the timeframe of the drop, compared to a previous timeframe. Then identify what products, if any, are responsible for the dip.
If the drop is centered on a few top selling products, dig into why that is. Visit the product detail page (PDP) for those products. Are they all in stock? Can you make it through the checkout process? Any issues on mobile with those products? What about price competitiveness, did competitors begin offering those products at a lower price? Those are all things to watch out for.
If the drop overall is spread across products, then you’ve ruled out a specific product issue (maybe).
3) Landing Pages Being Targeted Change Dramatically?
With Google’s Smart Shopping (RIP) and Performance Max campaigns, Google dynamically selects the products, but also various pages on your website. Make sure to check the Expanded Landing Pages report in Google Ads for the specific campaign where performance has dipped (or all of them), to identify whether Google is still targeting the same landing pages or products. There may be a reason Google is no longer targeting the same products, but this can at least give you some insight.
One time in a client, we noticed a massive uptick in traffic without a corresponding rise in sales. We investigated the landing pages and discovered they had selected a random category page (probably to test) in the static ads, likely in Gmail, and traffic surged with no sales. We were able to show our client that this was due to Google testing and not something we had done or could manipulate as the ones managing Shopping Ads.
4) Anything Change on the Website Recently?
Ah yes. The website. You know, it’s the crucial part of ad campaigns that you, the advertiser, have no control over but that can guarantee your failure.
As we noted previously, once you have investigated the feed and campaigns, ensure you dig into the website to identify any issues here.
One of the best ways I have found to do this, is to annoy the developers with a single question:
“Hi Tiffany, can you let me know what changes to the website were pushed live between January 17 - 22?
That’s it. That’s the email.
What you’ve done, is identified the specific date of the drop (let’s say, January 20), and asked for details around that time since drops may lag in reporting. You’re also not asking for specific details to lead them to only investigate something like the cart. You want all the deets. Did they completely change the category page layout? Did they change the PDP layout without telling anyone? Did they add seven pages to the checkout process? Gather information, rule out, and then make deductions and bring the team in for further investigation. This is especially helpful if you’re an agency and there may be multiple people involved client side.
5) Dynamic Remarketing ID Mismatch?
One of the things to keep in mind, is that Smart Shopping (RIP, again) and Performance Max rely heavily on Dynamic Remarketing. This means the feed needs to match to your PDPs so the correct products (that a person actually viewed on your website) can be shown appropriately in the remarketing ads. As a reminder, they’re dynamic in nature, because the products that show up in the ads will change dynamically based on which user is viewing the ads.
When it comes to dynamic remarketing in Google, the id you note on your page must match the ID in your product feed, and sometimes this can change or break. Sometimes a developer might change something on the page (thus the open-ended question above, remember?) and drop the ID parameter, and thus, your products no longer match up specifically to that user so Google just shows them products they think they *may* be interested in. Big difference between those and it’s not a stretch to assume that will impact performance. Just one more thing to monitor.
6) Product Targets in Campaigns Active?
The tricky thing about managing Google Shopping, is that you choose to segment your campaigns around things like Custom Labels, Product Type, product brand, etc… but that any time those change in the feed they will get dropped from the campaign targets.
In other words, let’s say your campaigns are targeting products in the Product Type “coffee makers” that you have set up. Then, someone internally changes something they didn’t realize would change Product Types, and now the Product Type in the Shopping Feed is “coffee maker”. Your campaign or ad group targeting “coffee makers” as the Product Type will no longer run since those products are no longer assigned to that targeted Product Type!
One of the easiest ways to do this, is by adding the column % of Active Products in your Product Groups Tab to determine if there are product groups with 0% active products.
Another way to troubleshoot this, is to look at product groups in the UI, and then compare the timeframe from the performance drop to the previous timeframe, THEN sort by Impressions (prev period). You can immediately see if there has been a hard drop-off, and then dig into that product group to determine if there are still products active.
This, BTW, also helps if the products are still active but there is another issue. It’s a great look at just figuring out where a drop traffic is occurring. Sort by conv value/cost changes in this same view to monitor efficiency changes. Conversions to monitor sales volume changes, etc).
One final note on this, in our Standard Shopping campaigns, we typically always have an “Everything Else” ad group (we like to name them "Misc" or "Everything Else") where we exclude whatever segments we have targeted purposefully in the other ad groups in this specific campaign, BUT that allows us to keep this as a “clean up” ad group in case there is an issue with feed matching to our previous segments. We can easily see this, when this (lower bid) everything else ad group begins to get impressions. This requires investigation, because it means a new product has been added that isn’t being categorized anywhere else in your Standard campaigns, OR something has changed in product segmentation.
7) Campaign bidding or targeting changes?
Finally, I would be remiss in not bringing some traditional Google Ads troubleshooting to bear in identifying performance issues. Check change history for a few days around the time of the drop.
Anything specific, and major happen that should be reverted? Did ROAS targets get lowered and the campaigns put into lower value auctions?
Check search terms (in Standard campaigns, for as much as you can see), any new terms that need to be excluded?
What about other channels? Did the client turn off upper funnel campaigns in Facebook or YouTube or Discovery around this time (or anytime before the event) that would begin to lower traffic to bottom funnel campaigns like Shopping?
8) Auction Insights Movement?
Finally, what is your competition doing?
It’s helpful (though, don’t run here too quickly) to keep an eye on strong movements in auction insights, especially tracked with your cost per click on average. One of the ways to do this, outside of the UI, is to utilize this cool script from Brainlabs. It’s an oldey, but a goodey!
This allows you to track competitor changes over time, and layer in your CPC averages as well so you can investigate any potential correlations.
So there you have it, those are some of the big things we look at first when we see that Shopping traffic and/or performance has dipped. I hope it was helpful to you, and please let us know if there are other things you like to dig into as well by connecting with me on LinkedIn or Twitter.
May the Auctions Be Ever in Your Favor!