How Do I Find the Search Terms Google Matches to Our Products?

Roy, how do I find out which search terms Google is matching to which products in our Shopping Campaigns?  Ideally, I'd like to find the search queries that matched to a specific SKU in our feed.  I am playing around with the data in the dimensions tab but am not exactly seeing a view, filter or column that helps make that association clear to me. Derek

Hey Derek, there’s not a great way to get this data by SKU. The best you can do is to use Google Analytics to find search queries matching Shopping Campaigns based upon how your feed is setup.

  1. If you are using Google Analytics and have it connected to AdWords navigate to: Acquisition > AdWords > Campaigns > and click into your shopping campaign
  2. Next, click on keywords in the top navigation
  3. Then underneath keywords, use ‘secondary dimension’ and use search query.

That will show you the search query data based upon your product IDs. I often see a lot of product IDs being * though as there are ‘all product groupings’ or something else and so not all the data will associate to a specific product; but that’s usually the best query data you can get for shopping terms.

Another way I've done this is to create a Shopping segment, and then look at some of the reports. The downside is that from within Google Analytics the "keyword" is actually, for reasons I don't fully comprehend, a string representing the product group that generated the click, and not the specific product.

If you set the segment to Shopping campaign visits, and then check out the "Search Queries" report under Acquisition > AdWords, you can at least see what queries are generating conversions. If you have AdWords and Analytics connected from the Analytics side you can see clicks or CPC in addition to sessions, conversion, etc.

If, however, you had a specific product in mind, you could also refine that "Shopping visitors" segment by going to the Enhanced E-commerce settings in the segment and adding a condition for that product--I just don't see a way to do it en masse.

Hopefully, Google will make it easier to find search queries matching Shopping Campaigns in the future.

Thanks for the question Derek.  Drop me a line if you have any other questions.  I love to talk shop.

Up and to the right,

Roy

Why Do We Have Such Low Impression Share Relative?

Roy, I hope you are doing well and are surviving the holiday season.  I was wondering if you could answer a few questions for me about our low impression share and how we can do a better job of optimizing impression share relative to our competitors.  I've spent some time in the Auction Insights Report in Google AdWords and our impression share is 11% compared to some of our competitors that are in the 30-40% range.  Here are a few questions that I'd love your get your insights on:

  1. If we don’t see ourselves listed as a seller of an item, but it is in our feed with the correct GTIN is it possible we are bidding too low or does Google display ALL sellers for a given SKU?

  2. Do you know what constitutes an impression in Shopping Campaigns? Is an impression only when it shows up in Search Results or is it any time a user is exposed to our ad?

  3. What are your tips for improving our low impression share?

Matt

Hey Matt, impression share is a diagnostics tool, not something you should consider "optimizing for". You are optimizing for profit dollars. Full stop.

My understanding is that an impression is counted any time your site's name appears, and there's a link to you, once per user per session. So, if the see an Ad on the SERP, and then click to seem more vendors, and then click to see all vendors, anyone who showed up along the way gets an impression (and only one). However, if you started your search over, with a new query, those are new impressions. Don't quote me on that, though.

Where you are successfully matching impression share is a function of bid. Bid, in turn, is a function of AOV, Conversion Rate, and COS. Therefore, Impression share is tied to spend, if you consider the site's metrics as a separate matter.  You can check out my free eBook Getting Started with AdWords for a little more background on how bidding works.

If you're concerned with overall match rate, it's actually an unusually high impression share that shows up as a symptom. Imagine there's a product ABC, but you're the only one in the world that calls it an ABCD. Then, Google would be reporting your share of impressions on searches related to the ABCD--of which you would own almost all of them. A very low impression share suggests that you have room to improve, but it's not directly a matching issue.

If you have products that aren't attracting any clicks, but you're certain you have the right UPC and other product data, then the next variable in the list is bid. I don't know if there's a threshold that would prevent you from showing up at all (even in the expanded list), but very few shoppers get that far anyway. To test, you could try to isolate a specific problem product, and push its bid far beyond normal, and see what happens. If it starts getting clicks, then you have your answer (and you might be able to see it in the wild).

But bottom line is: you're not trying to grow impression share. If you aren't the cheapest on every product, and have the most overwhelming merchant reviews, that could very easily lead to nothing more exciting than a lot of unwarranted expense.

Thanks for the question Matt.  Drop me a line if you have any other questions.  I love to talk shop.

Up and to the right,

Roy

What's Your Experience with the Data Driven Attribution Model?

Data Driven Attribution Model in Google AdWords
Data Driven Attribution Model in Google AdWords

Hey Roy, have you had any experience working with the Data Driven Attribution Model in Google AdWords? 

Do you have any recommendations on when you'd use it?

Nathan

Nathan, happy to share my perspective on the Data Driven Attribution Model.  Google defines the Data Driven Attribution Model as follows:

Data-driven attribution gives credit for conversions based on how people search for your business and decide to become your customers. It uses data from your account to determine which ads, keywords, and campaigns have the greatest impact on your business goals. You can use data-driven attribution for website and Google Analytics conversions from Search Network campaigns.

You can learn more about the Data Driven Attribution Model by visiting Google's support site.

The Data-Driven is currently a black-box, and I don't trust it.  Attribution model and performance are chicken and egg, as the model becomes a self-fulfilling prophecy.  If you picked First Click, then you'd invest more into high-funnel campaigns, and then high-funnel campaigns would perform "better", for example.  If they're claiming that they can find the combination of model variables that increases the total at the bottom of the funnel, then that'd be cool... but the Conversion Optimizer, eCPC, Ad Optimization, and all of the other "magic black box" features in AdWords generally underperform (or outright fail) for mid-size retail accounts, as far as I've found.

I find that most of the time when Google launches some automated data analysis system that the first 6 months to 1 year are not very good and the early adopters don’t do well unless they have so much data the system can learn very quickly.  It seems that it really takes 1-2 years to work the bugs out to where it generally works.

So, let's put it in the "wait and see" category, but I'm very skeptical.  Till I see evidence suggesting I'm wrong (which happens from time to time), I'm going to stick with Time Decay.

Thanks for the question Nathan.  I love to talk shop so don't hesitate to reach and drop me a line.

Up and to the right,

Roy

Why Am I Not Showing Up in Google Shopping

Common Google Shopping Campaign Questions

Dear Roy, I am taking over my family's e-commerce business and I am new to the world of paid search and have a few questions for you about Google Shopping.

  • What are the reasons we do not appear for some keywords in Google Shopping

  • If we wanted to appear for certain keywords in Google Shopping what should we do and how would that effect our costs?

  • Why are we not showing up for Mobile searches within Google Shopping?

Thanks for sharing your insights

Sincerely,

Gene

Hey Gene, no problem, I love talking shop and am happy to help.

What are the reasons we do not appear for some keywords in Google Shopping?

There are a number of variables that Google is considering when it determines who to show. First among them is the relevance of the product data to the search. While they don't tell us how they determine that, we can certainly make some safe assumptions. First, they are building an abstract catalog of all products behind the scenes, based on all of the data they collect from all of the retailers on the platform. Therefore, if you were selling Widget ABC, and your competitors were as well, and Google could tell that they were the same product (by UPC, MPN, Brand, etc.), then they're going to start to paint a picture of what they think that product is, and what words are relevant to it. The downside is that because Google doesn't actually have firsthand knowledge of ABC Widgets, it becomes a bit of a democracy of competitors, and it's difficult to tell what they're looking for.

Second, they are doing more traditional keyword matching, especially against the title, but also the description. There is some semantic analysis here, as well, so closely related terms are included, in part. Given the sample keywords in your example, this isn't as likely to be the issue, as those are basically searches exactly for your product title, but it's a variable that's core enough to what Google's doing, that I have to mention it.

Third, they are checking the relevance of the landing page. This is conjecture on my part, but we know that they're parsing landing pages as part of their ad validation process, so they do know something about where they're sending the shoppers, and it's entirely plausible that there's a Quality Score-like calculation going on behind the scenes, as well.

Those three are the difficult ones, as improving them involves a lot of content work (which is generally manual, and therefore expensive and hard to scale), and there aren't that many types of feedback that Google provides. Among what we do have to work with is Impression Share. Some of your products are currently seeing an impression share of 58% for the past three months. That seems high to me, based on the performance you're seeing in your sample queries, and it's more likely that the items in your catalog are under-matching to the products that Google thinks shoppers are looking for. A high impression share means one of two things--you are either dominating a given audience, or you are only able to access some small segment of the true audience, wherein you are a big fish in a small pond. This is what I fear is happening here--it might not even be a bidding problem at all.

Fourth, and last, is the bid. While it's true that AdWords is fundamentally an auction platform, relevance is key to ensuring that shoppers keep coming back to Google to find things, and so Google is sensitive to that above all else. Once a list of viable contenders for that ad impression are vetted and ranked, the bid is the final ingredient.

If we wanted to appear for certain keywords in Google Shopping what should we do and how would that effect our costs?

The first thing I'd do is analyze whether or not you have a bidding problem. As I mentioned, I think you may have (at least in part) a match issue, as your impression shares are high. Despite this, the "Benchmark Max CPC" tells another story. Google provides this value to give you a ballpark idea of what other advertisers are paying per click for that kind of traffic. Some of your products bids are currently around $1.00, but the Benchmark is $2.64, and your average actual CPC is only $0.62. That benchmark actually conflicts with the impression share data, in my opinion. It may indicate that there's a bid ledge somewhere around $0.60, where the competition gets far more fierce after you reach that tipping point. Here's what Google thinks the audience looks like, based on the past week's data:

Google Shopping Bid Simulator
Google Shopping Bid Simulator

So, we're in the $0.86 to $1.14 range, spending between $1,500 and $3,000 in their hypothetical date range window. If we were participating at the benchmark level, we'd expect to see costs leap to over $13,000 in the same period--between 5 and 9 times as much as you're spending now, in dollars. We can see that they think the actual CPCs will be about $1.60 at that level, though, so it's clearly non-linear.

You can read quite a bit more on my approach to this kind of data here, where I do a bunch of interpolation to try to find the optimal point on the curve for profitability.

What would the output of that $13,000 spend look like? If we assume the Conversion Rate of 0.11% holds, then we know that the 8,310 clicks will produce about 9 conversions, whereas the ~3000 clicks at $1.00 would produce about 3. Using our Average Order Value of $1,838, we see that this would be about $16,500 in revenue for the $13,000 in spend--almost certainly a losing proposition. The law of diminishing returns is in full effect, here. While I don't think I'd trust Google's forecast (based on such a short date range) to represent literal figures, the relative comparison of the two hypotheticals is likely close. That's about a 2:1 ROAS on the smaller volume, versus ~1.2:1 for the higher volume case.

Why are we not showing up for Mobile searches within Google Shopping?

This one's easy, because you're choosing not to be:

Google Shopping Mobile Device Bid Adjustment
Google Shopping Mobile Device Bid Adjustment

If I had to guess, I'd say that this has only been the case since that drop in click count around 8/18, based on the fact that there had been data for the Mobile segment in the past three months, and that drop-off caught my eye as being likely related to this. And, when I zoom into the past 7 days, and see no impressions for Mobile, that hypothesis is confirmed. This could be a strategy to separate Mobile traffic into its own campaign, but I don't see that being executed. I'm also not a fan of that strategy, outside of some very specific use cases.

I hope this helps answer some questions and if there is ever anything I can do to help with Google Shopping or AdWords drop me a line via the Contact form.

Up and to the right!

Roy

Max CPC's Relationship to Actual CPC and Impression Share

Relationship Between Max CPC Bid, Actual CPC and Impression Share

Hi Roy, I had a question in regard to Max CPC bid, Actual CPC and Impression share. One of our campaigns in PLA has been over-target ROI significantly for a few weeks. Impression share is at the 60% range and I would love to push that into the 80% or more. What’s odd is that our max CPC bids in some cases can be in the $2-$3 range and our average CPC’s end up being in the 30-60 cents range. I did check in on our auction insights and it looks like we greatly outrank our competition in impression share. This implies that CPC’s aren’t being driven up due to lack of competition. I feel a bit skeptical about this though. Any thoughts? I don’t understand why google wouldn’t want to sell me those extra impressions.

When the actual CPCs are coming in much below the Max CPC Bid, then a lack of competition is the first thing that comes to mind to me, as well.  While your impression share is strong, it's not so high that I expected to see $2.50 bids coming in at $0.50, though.  However, the way you're interacting with other sites in the search auctions is still the only thing that affects CPCs like that, relative to bids.

What this might uncover is an opportunity with the product data, such as the titles.  While I'm sure you're feeding in UPCs, it may be possible that Google isn't matching your products with all of the searches that they might be relevant to.  For example, if I passed in a bunch of titles that were too short, then I would see apparently reduced competition--along with a significantly reduced impression pool, relative to a feed that was more robust.

Negative keywords can also have an impact so make sure there aren't any overzealous negative keywords limiting you from quality impressions.

Bid adjustments also may play a role in the delta you are seeing between the max CPC and actual CPC.  Let's look at a typical mobile bid adjustment. If half of the clicks are mobile and mobile bids are half as much as desktop bids then the total average bid would be 3/4 what it would otherwise be and it wouldn't be reported at all.

Have questions?  Drop me a line via the Contact form--I love to talk shop.

Up and to the right!