Advanced Analytics for Google Shopping – Tracking Product Ads in Google Analytics

Ninjas don’t show up on radars, but that doesn’t mean they don’t use them.  And when it comes to Google Shopping, every analytics or paid-search ninja should know exactly which products and product targets are responsible for clicks and conversions within their campaigns.

If you’re managing a paid Google Shopping campaign (self-proclaimed ninja or not), it’s useful to enhance the scope of your analytics radar so that when it comes time to evaluate campaign performance you have more than just one way to slice and dice your data.  In this article, I’m going to demonstrate a couple other ways to evaluate your Google Shopping performance that’ll hopefully help take your analysis and strategy to the next level.

Get ready to sharpen your analytics swords!

Your Product Tracking Sucks in Google Analytics

Admit it.  You probably haven’t even looked at which products are converting for your site, let alone thought about how to set it up.

Granted – Google Shopping is still in it’s infancy and they’re still transitioning over 100% of free Google Product Search traffic to the new paid ones, so it hasn’t really been worth your time… But is that really a valid excuse for not knowing what your Google Shopping ads are doing?

You’re still paying Google for that traffic, right?  (The answer is yes) And as more of the traffic shifts from free to paid you’re going to want to know as much as you can about your campaigns and how to improve them.

But before we jump into all the details, let’s really quickly address how the new Google Shopping is different from traditional Adwords.

With text ads you can run A/B tests on specific keywords where you change the ad copy or landing page to identify the intent behind specific searches and exactly how far down the purchasing cycle that person is.  The main problem is that much of those systems that we currently use to analyze text ads don’t necessarily translate to Google Shopping.  For two main reasons:

1. No keyword-specific bidding for shopping ads (only negative bids).
2. Google decides which of your products are relevant to the user’s query and shows that one.

This means that you’re probably not tracking exactly which products are getting visibility or the search terms that are triggering their ads since a lot of it seems out of your control.  You’re watching Campaign and Adgroup performance in Google Analytics but there’s not a whole lot of actionable data that you can pull from that if you’re like many ecommerce sites that sell over 10k products.

 

Setting Up Advanced Product-Level Tracking in Google Analytics (Putting On Your Ninja Uniform)

Chances are that if you currently manage Google Adwords for text ad campaigns, you’ve already synced your Adwords data with your Google Analytics account and you’ve used those reports to evaluate performance for your campaigns, keywords and adgroups.

But as I mentioned earlier, this level of segmentation for Google Shopping can stagnate campaign analysis so let’s set this up so that we can take a deeper dive into the data later on.

Tracking Product Targets in Google Analytics

You can already access macro performance reports for product targets by going to Advertising > Adwords > Keywords tab in Google Analytics.

If you have traditional text ad campaigns also running it may be difficult to find the product targets among all the other keyword-level data in this report so here’s a handy regular expressions filter that you can use to filter out only the product targets.  All you have to do is copy/paste the following code into the advanced filter, and filter on Keyword:

[\=\*]

*Remember to select “Matching RegExp” instead of the default “Contains” otherwise it won’t work.

 

This will give you essentially what you’re already able to see in the Adwords account but with actual revenue per target.  This is slightly better but you still don’t have a whole lot to work with.

But what if we were able to see exactly which products were getting clicks/conversions within each product target?  We can do this but it requires a small tweak to the data feed.

The {adwords_producttargetid} Parameter

In your Google Shopping data feed there’s an adwords_redirect column where you can set up custom tracking parameters for your Google Shopping ads.  We can actually track product target performance on the item-level by adding the {adwords_producttargetid} parameter to the end of the urls in this column.

If you already have tracking parameters appended to your adwords_redirect url (anything after the “?” in the url) then all you need to do is add &id={adwords_producttargetid} to the end.

If this is completely new to you and/or you don’t have any tracking parameters you’re going to have to add ?id={adwords_producttargetid} to the end of the url.

*The difference between the two above is the leading “&” vs the leading “?”.

 
By adding the {adwords_producttargetid} parameter, you’ll be able to see the Google-generated ID for the specific product target that triggered your ad.

You can access this report in Google Analytics by going to:

1. Content > Site Content > Landing Pages
2. Choose your Secondary Dimension for Traffic Sources > Campaign.
3. Once you do that, edit the Advanced Filter to include only.

 

Shazaam! Now you have a list of all the landing page urls (aka the Google Shopping ads that are being clicked) within a Google Shopping campaign.  Each ID parameter will tell us exactly which product target is rendering our ads.

 

 

Segmenting The Analytics Report in Excel (Unsheathing Your Ninja Sword)

Now let’s head over to Excel and work some magic.

First, download the report by going to the Export tab and selecting your desired file type for Excel. CSV is usually most compatible with Excel.

Which will give you a report that should look something like what’s below.

*It’s possible that you’re appending other valueTrack parameters that might be showing up in this report (like adType=pla) so if you are, just do a Find/Replace to eliminate them, leaving us only with the landing page url and the ?id= parameter.

 

Now we can figure out exactly which products are getting clicks/conversions by:

1. Remove all instances of id= in column A.
Find/Replace (CTRL+H) – Find “id=” and Replace with “”
2. Select column A and Text to Columns in Excel.
3. Choose the “delimited” option.
4. Choose “?” as your delimiter
5. Hit “Finish”.

 

Now you’re going to have a report that gives you clicks BY PRODUCT and is also filterable based on PRODUCT TARGET.

These IDs correspond to specific targets in our Adwords campaign.

Unfortunately, right now there’s not a way for us to exactly find the corresponding IDs in the Adwords login but we should be able to figure it out by:

1. Comparing aggregate clicks/product target in Adwords to the aggregate visits/id parameter in Google Analytics (or Excel).  Depending on the nature of your targets this can be either easy or rather tedious.
2. Compare the urls that are getting visits with a specific product target id to their associated values in the feed.  Product targets are based on values in the feed so there should be some insight here if you’re familiar with navigating your data feed.

 

If we go back to this specific report in Excel, there are only two possible IDs in our new product target column and it’s easy to see which one is All Products. (Hint: it’s the one with the most clicks) So, in this case our product_target IDs are

All Products = 21619610058
adwords_labels=Top Sellers = 39787593978

We can make a note of what these IDs are for future reference because they will not change.  Save it in Evernote or whatever you like to use, just as long as it’s in a safe place so you don’t have to go through this again and again for the same account.

Using {adwords_producttargetid} To Improve Google Shopping Performance

Macro-level performance for adgroups and campaigns can be extremely valuable, especially when you’re doing a quick audit on the health of a campaign, but it’s more useful to see actual product-level data since that’s where you’re going to see the biggest gains in ROI.

Take the following birds-eye view of a Google Shopping campaign that’s the default in Google Analytics.  The Ad Group for “Top Sellers” is converting at above 5% with a per visiti value of $4.19!  The only problem here is how little traffic we’re getting.. so the obvious solution is to increase our bid for the entire adgroup right?

 

Before we do anything rash, let’s head back to our product target report in Excel and and take a look at exactly what’s going on at the product level.

The Top Sellers adgroup seems to be converting well so far, so we filter on that ID and find that only one item has been converting in this group. 3 clicks, 2 conversions for $146.60 in revenue.

WHOA. So, I thought that this adgroup was already doing well with the limited # of clicks we had, but now I just learned that it’s actually all due to one product!  Equipped with that knowledge, maybe I should raise the bid for that one product rather than doing it for the entire Top Sellers adgroup.

This also shows you that /ganocafe-classic.html has made up for about half of the total clicks for this adgroup without even converting once yet.  If I raised the bid for this entire adgroup it’s likely that that one poor performing item would continue to get the majority of traffic and our once stellar
performance would go waning into the night.

Now that’s some sexy analyzing!

 

Leveraging Search Queries For More Product-Level Analysis

Another approach that can be coupled with the above  analysis is to look at which search queries are triggering your ads to make sure the right products are being included in the right targets, labels, groupings, etc.

You can see the queries that are triggering specific product targets by going to the Matched Search Queries tab and selecting Keyword as your Secondary Dimension.  Once again, use the [\=\*] regex filter to isolate the Google Shopping targets.

For example, this merchant sells a few “gano cafe” items but his titles/urls can differ based on whether or not that term is written as one word or separately as two words. As a result, these search terms are triggering two different sets of ads that perform differently from one another.

 

Assuming that the url contains either “gano” or “ganocafe” we can go back to the target ID report we made in Excel earlier and figure out which items are being served in each adgroup. Filtering on Landing Page, we see the top ads that have “gano” in the url and it’s fairly obvious which ad is being rendered for each of those two searches above.

 

This makes me wonder why I’ve got the /ganocafe-classic.html listed in the Top Sellers adgroup when it’s actually getting less clicks AND converting less than the one in the All Products adgroup. Even though we know that one is a site-wide best-seller, that may not be the case when the traffic is coming from the paid Google Shopping real estate.

This level of analysis helps of identify search terms that are better qualified as well as product ads that convert the best.

Caveat: This analysis was actually performed on a very small data set since this particular campaign hasn’t been active for very long.  As our campaigns grow and we collect more data there should be clearer call to actions from the data and we can make better informed decisions.

But just for the sake of discussion, here are a few actions that we might consider for our example data set:

1. Changing the title for our Top Seller ad to “gano cafe” rather than the single merged term “ganocafe”. This small detail is obviously triggering different ads for their corresponding searches and the search volumes will differ between the slight variations.
2. Removing /ganocafe-classic.html from the Top Sellers adgroup. If conversions don’t improve for this item (in this target), then we can assume that this ad isn’t going to send us very qualified traffic and we no longer want to include it in our highest bid adgroup. In our case, this will require some adjustments to the adwords_labels value.
3. Increasing the bid for /gano-cafe-3-in-1.html. We don’t have enough data on this item just yet, but if it continues to convert at a high rate and profitable ROI then we may want to increase our bid here.

 

Can you think of any other ways to bolster up your strategy using the analytic methods above?  What about different ways of tracking or segmenting data that you think is useful?

Let us know in the comments!  If there are any good ones I’ll take the best one and write a new article about it ;)