When it comes to driving ecommerce revenue from Google Ads, Google Shopping is one of the most popular campaign types, but often also the most overlooked.

It’s as much the Shopping feed optimization part as the Google Shopping campaign optimization part that just gets left “as is” without any optimization. Hardly anyone goes beyond the standard Google Shopping campaign structure and really think about how you can improve performance by segmenting your search terms.

In this post, we are going deep on everything related to your Google Shopping campaign structure:

  • Campaign segmentation to enable bidding based on actual search terms
  • The optimal ad group structure to enable the right products to show
  • Why breaking out product groups might be the most overlooked part of your Google Shopping optimization routine
  • (NEW: August 2019) How it stacks up against Smart Shopping Campaigns

Let’s dive straight in. The biggest challenge with Google Shopping campaigns is the lack of keyword data.

If you’re selling sofas, then there is a huge difference in conversion rate for the term couch vs black leather couch with chaise lounge. 

The first step we have to do is to regain this control.

Regaining the Lack of Keyword / Search Query Control With a Proper Campaign Structure

One of the biggest missing parts of Shopping Campaigns is the lack of control over the search terms – aka what search terms are triggering your ads.

Many search marketers have asked for the ability to add keywords in Shopping campaigns, but most of us have realized it’s not likely to happen. Google is moving towards less manual campaign management – not more.

Like I’ve seen Kirk Williams ask for so many times (I SUPPORT YOU), our best shot is to get keyword-level modifiers. Meaning that we’d be able to set a bid modifier for search terms that contain specific keywords (like with mobile). This could be pretty interesting.

Keep up the good fight, Kirk:

Kirk Williams wanting keywords as a bidding signal in Google Shopping campaigns

But it’s most likely not going to happen, so we have to think in alternatives.

The Solution: Splitting Campaigns by Search Query Pattern

One of the best ways to regain some of the control until Google (maybe) gives it to us is by splitting your campaigns by search query pattern.

This sounds more complex than it actually is.

What it means is that you find a specific pattern in the search queries that result in a higher ROAS or conversion rate than the other search queries.

The most common example in eCommerce are searches that include a brand, model or product name.

An example is running shoes:

  • Marathon running shoes → Generic search query
  • Nike marathon running shoes → Branded search query

Typically what you see is when someone includes a brand name in their search they are further in the customer journey than a generic search.

This means searches that include a brand tend to have a higher ROI than generic searches.

Query-Split Campaigns Allow You To Bid According to Search Query Pattern ROI – Not Just Product

With a standard Google Shopping campaign setup you can only set a bid for a specific product, category, or brand (amongst others).

When you split up your campaigns in Branded and Generic segments, you are able to bid differently for searches that include a brand name versus searches that don’t.

It doesn’t give you the power to bid higher for each individual search query, but it does give you some power back.

But imagine you do it like this:



Searches that includes brands will fall in the brand campaign (i.e. Nike Marathon Running Shoes).

Searches that doesn’t include brands will fall in the generic campaign (i.e. marathon running shoes)

Searches that have been excluded from both the Generic and Branded campaigns will appear in the Catch All campaign.

You Catch All campaign should be “last chance” search terms such as highly generic searches. Let’s say that you’re appearing for the search running shoes. This is highly generic and showing very specific product ads is unlikely to result in great results.

Especially as you’re bidding the same for this search term as you are for a search term like marathon running shoes or Nike running shoes.

By excluding a low-performing search term from the Generic and Brand campaigns you will be able to bid lower on that specific term via the Catch All campaign.

I will go into more detail how to create the query-split campaigns below. But first it’s important to find out if it will work for you.

“Yea, But Will This Work For Me?”

As with so many blog posts, you are likely wondering: Will this have an actual effect on my  campaigns? It sounds cool, but is it really applicable to me?

Luckily, there’s an easy way to find out.

By excluding all search terms that include a brand in your Search Terms Report, you can see exactly what the performance difference is when a brand is included in your search terms:

Finding high-performing search term patterns for a segmented Google Shopping campaign structure

In this example, any search that included the brand HP accounted for 41% of all costs and 38% of all conversions. That means it’s definitely worth dividing campaigns in brand and generic searches.

If there are significant searches for both including and excluding brand, then it makes sense to split.

This was an easy example where HP is the main brand sold. However, if you have dozens or maybe even hundreds of brands, you’ll have to get your spreadsheet skills out.

But with hundreds of brands, then you’ll definitely see evidence that you have to split your campaigns in brand and generics – guaranteed (just saving you some time here).

How To Structure Your Google Shopping Campaigns in Practice

The way all this looks in practice is the following:

Using negative keywords and campaign priorities for your Google Shopping campaign structure

Image inspired by Crealytics that run the Shopping tool camato

With the high priority setting you show the Generics campaign first. So if there is a search on Google that matches one of your products, the Generics campaign will show up first. This sounds counterintuitive, but stay with me.

When you add your brands as negative keywords to the Generics campaign, you are essentially telling Google that all searches except your negative keywords, should appear in the Generics campaign.

Important: Negative Keyword Lists & Product Exclusions

This also means if you add any negative keywords in the Generics campaign, but not in the Brands campaign, then these searches will appear in the brands campaign.

So, you should have two different negative keyword lists:

  • All Shopping: is used for excluding searches terms you don’t want to appear forever in any Shopping campaign
    • This should be applied to all your Shopping campaigns
  • Generics/Brands Shopping: is used for search terms you want to give a last chance in the Last Chance campaign
    • This list should only be applied in your Generic and Brand campaigns.

As you’re working with your search term reports that you can’t just add negative keywords “to campaign or ad group level”. You have to add them to the lists, otherwise you’re not actually excluding anything.

If you only add a negative keyword to Generics campaign, then the Brands and Catch All campaigns will start showing ads for those keywords.

Excluding Products

The same applies to excluding products. If there is a product group that is not doing well in the Generics campaign, then you have to exclude it from the Brands and Last Chance campaigns as well.

It’s crucial that you remember this.

If you want to exclude a product group from the Generics campaign, but want to let it remain in the Brand campaign, then there is a workaround. You simply set a bid for $0.01 for that specific product group in your Generics campaign.

If you’re running automated bidding, then I recommend creating a separate ad group that you set to manual bidding where you can put all your unwanted product groups to $0.01 bids. It’s not ideal, but it will do the trick.

How to Split Campaigns If You Don’t Have Brands

I’ve seen situations where an ecommerce store either did not carry well-known brands or where the performance difference was too small.

If this is the case for your store, it doesn’t mean you can’t use the campaign segmentation tactic to get better results from Shopping, but it does mean you need to get a bit more creative.

While using the same technique of filtering your search terms, then you can come up with other potential search patterns that make sense for you to segment your campaigns by.

Campaign segmentation is all about finding commonalities in search terms that convert better than others.

A couple of examples that I’ve used in the past:

  • Top keywords (manually chosen)
  • Sizes, genders, and other “qualifiers”
  • Specific attributes (color, material, etc.)
  • Category keywords like marathon, trail, concrete, indoor, etc.

Just remember if you have a very large store, then you might run into the 5,000 negative keyword limit on your negative keyword lists. Very few advertisers will ever experience this limit, but just be vigilant of it.

To Use Mobile-Only Campaigns or Not?

I’m a generally a fan of splitting up your important campaigns into mobile-only campaigns. I like it and I generally see better results when doing it.

Shopping campaigns are no different. I might actually be more inclined to split up Shopping campaigns because if you create Shopping campaigns with only one ad group, then it’s a necessity to split up your shopping campaigns in mobile-only campaigns.

The reason comes down to the bidding. There isn’t really anything else that can move the needle when it comes to getting better results on mobile Shopping campaigns than better bidding.

The advantages:

  • You can set very specific bids per product / product group
  • You can apply specific mobile ad schedule bid adjustments

The disadvantages

  • You get double the work

However, in the end I argue that doubling the work on your mobile Shopping campaigns isn’t really that big of a deal.

Comparing the performance gains advertisers get from Shopping with the time investment, most advertisers can afford spending more time with Shopping.

Ad Group Structure: To Split Up By Ad Group or Not?

Before we dig into this, I want to specify what difference between an ad group and product group is:

Ad groups contain product groups.

Product groups are a way of grouping products together. This can be done in any way you want. Typically, you’ll create product groups based on brand, product_type, price, profit margin or another attribute.

Finally, a product group can also be a single product.

When managing Shopping campaigns, splitting your products into multiple ad groups within your campaigns can be a sound strategy.

Two of the biggest advantages of multiple ad groups are:

  1. Stronger control over bid adjustments (RLSA, Mobile, Ad Schedule) as you can set a bid adjustment on the ad group level
  2. Better insights to what search terms specific product groups appear for

Since there is no way to see what search terms a product group is appearing for (REALLY, GOOGLE?), the only way to get more insight is by splitting your product groups in individual ad groups.

Depending on how many products you have, this can become a time-consuming task. Especially if you split your campaign in mobile too. Optmyzr has a super simple tool that allows you to do this with the click of a button (Check out my free step-by-step guide to setting this up.)

Use Multiple Ad Groups To Show The Best-Converting Ads for Specific Search Terms

When you have multiple brands that carry the same product categories you want to create individual ad groups for specific product groups/brands so they can own a keyword:

How to use ad groups in your Google Shopping campaign structure

Let’s say that you have Reebok, Asics and Nike athletic shoes.

You’ve discovered that Reebok sells the best for searches for crossfit shoes.

Asics sells best for trail running shoes

Nike sells best for lightweight running shoes

If you only had one ad group, there would be no way for you to determine what brand/product should trigger what keyword. For single-brand advertisers, this is a non-issue, but if your store carries many brands it can become a huge issue.

By creating an ad group for each brand, you can exclude the wrong search terms from the ad groups / brands that you don’t want them to show for:

How to use negative keyword on ad group level in your Google Shopping campaign structure

Product Group Setup

In regular search Ads you have keywords. The equivalent of a keyword in Shopping Ads is a product group. Basically, it’s a group of products (or a single product) that you can associate a bid with.

Below you can see the difference between what searches that a product can trigger vs a keyword:

How to improve titles for your Google Shopping feed

The difference between regular keywords and product groups is that you can’t really control what search terms you’re bidding for with a product group.

As default, Google will create a single product group that contains all your products.

It’s crucial that you split up your product groups is a way that makes it easier to bid for each individual product based on its performance.

The Number One Reason to Split Up Product Groups: Bidding

In order to set a specific bid for individual products, you need to split up your product groups. This can be done in just one ad group, or in several, which I’ll cover next. For now, let’s stick to the bidding aspect.

The main reason why you want to set specific bids for individual products, or product groups consisting of a few product, is due to the performance varying.

Let’s look at a couple of examples:

Some products have a 1,000% ROI while others have a 300% ROI. You want to set a separate bid for each of these.

Some products require a $3 bid to maximize their impression share. For others, just $1 or even less will work. The “magical” S-curve in bidding is crucial when it comes to bidding on individual products.

Some products do well on mobile. Others don’t. This is where splitting product groups in different ad groups or campaigns come in.

Some products should be excluded completely. Others shouldn’t.

If you don’t split up your product groups, there is no efficient way of discovering these differences or acting on them with your bidding.

Real Life Example of an Advertiser Not Using Product Groups

I’ve run across many cases where the product groups weren’t adequately broken up.

Often it’s because the overall results from Shopping are pretty good.

In the example below you can get a better idea of what I’m referencing.

The overall performance is 1,189% ROAS, but when we start breaking down the individual product groups you can start seeing that there are specific product groups that aren’t performing as well as the others:

pasted image 0 17

Granted, they’re still performing pretty good, but look what happens when you split up some of the product groups that are performing below average some more:

pasted image 0 12

And once more:

pasted image 0 16

You’re starting to get the idea, right?

By breaking up the product groups again and again, you can start seeing that certain product groups, or individual products, don’t perform as well as the others. When you don’t break up your product groups you lose the ability to see this.

And especially when your overall Shopping campaign is performing well you will end up ignoring the lower-performing parts of your campaign.

Ideas for splitting product groups

Product groups are split up in layered groups. This means that you split up product groups within a product group:

pasted image 0 18

You can essentially split out product groups based on whatever you want.

My favorite way to split them is typically like this:

  • Brand (usually as an individual ad group)
  • Product Type level 1
  • Product Type level 2
  • Product Type level 3
  • Item ID

It looks like this in practice:

How to break up product groups in Google Shopping campaigns

Some other interesting ways to split product groups can be:

  • Pricing (via custom label)
  • Margin (via custom label)
  • Brand
  • Product Type
  • Other commonalities

This is especially beneficial if your products vary in price within the same product type and brand. However, in most circumstances, then when you split a product group via two levels of product types (i.e. Running Shoes → Marathon Running Shoes) then the pricing for all the marathon running shoes are somewhat similar.

Yes, they might differ $50-75, but it’s not several hundred dollars, which can really tip the scale when you’re trying to find the optimal bid.

Don’t Go With Just One Shopping Campaign

Remember back in the day when the Google Ads experts would tell you not to put all your keywords in one ad group? Well, this is the equivalent to that.

Splitting up your Shopping campaigns based on the search query pattern is the single most effective method to improve your results with Google Shopping.

If you’re confused and don’t know what your next step should be, I’ve listed two essential next steps for you:

  • If you haven’t split up product groups, do it.
  • If you have split up product groups, try the Generics, Brands, Catch All campaign setup

Good luck. If you have any questions to your specific case, then let me know in the comments.

Note: I can’t want to take full credit for coming up with this strategy. The first time I heard about splitting up campaigns and using priorities to segment searches was from CPC Strategy. Since then I’ve also followed Crealytics for their insights on Shopping. This post is, therefore, more about my learnings and how I’ve used that strategy alongside everything else that goes into successful Shopping campaign structures.

Most writers want to take credit for every single thing they write themselves as if everything they write is an original idea. That’s rarely the case, so I feel it’s important to give credit where credit is due.


  • Randy McCallister says:

    Hello, we’ve attempted your branded campaign scenario and have run into major issues.

    Namely, Google is just choosing not to serve any ads from our account for searches containing brand terms at all.

    Rather than skipping the main “generic” campaign due to constraints and having the traffic filter to the medium priority “brand” campaign, we are just not getting that traffic at all.

    Google help has advised that this is unavoidable and your strategy simply will not work. Please advise.

  • Andrew Lolk says:

    Hey Randy,

    Love the Google advice 🙂

    It works. I think you just might have to double-check the way you’ve set the priorities.

    Remember that your generic campaign needs to have the high priority and the brand terms as negative keywords.

    The Brand campaigns needs the medium priority and no negative keywords..

    Also, remember that you shouldn’t change what products are active in the campaigns. All the active products should be the same in both campaigns.

    The setup is alive and well in almost all our accounts.

    I’m going to shoot you an email too 🙂

  • Brian says:

    I’m in a similar spot as Randy. When I search for some of the brands my store offers, I am not seeing any of my ads at all.

    It seems as if all I’ve done by following this campaign structure is cannibalize my branded campaign for the generic one.

  • Andrew Lolk says:

    Hi Brian,

    What have you done exactly?

    Feel free to write me directly.

    I’ve set up four of these campaigns since releasing this blog post, and they work every time.

    There is no cannibalization as the search terms appearing under the branded campaign is not the same as the ones appearing under the generic campaign.

    There can be a number of issues with your brands not coming up.

    The two most common ones are:

    1) You made a mistake when setting up your campaign priorities

    2) You’re not bidding enough

    3) You’ve misunderstood the premise of the way to set up these campaigns and have excluded products rather than working with negative keywords.

    With a store like usesi, then this approach will be highly effective when you get it right.


  • Martin says:

    Hello Andrew,

    Thanks for a very insightful blog post!

    I’m currently thinking about testing this structure out for one of my accounts. But I have a few concerns which I was hoping you could help me shed some light on.

    1. Is it correct that all 3 campaigns should have the exact same setup in regards to ad groups? Set up being 1 ad group per brand in your e-commerce store.

    2. In a specific ad group. Do I exclude the “Everything else in “All products”” product group, while having a second product group being the brand itself?

    3. When applying the strategy in regards to campaign priority and bidding. What will take presidency when it comes eligibility to serve an ad for a query? For example; Query being “running shoes”, if campaign 1 (with prio high) has a lower bid then campaign 2 (prio mid) will this query ever be eligible to run in campaign 2 due to the fact that it has a higher bid?

    4. When applying the negative kw list for brand. Is it preferable to use phrase match?


  • Andrew Lolk says:

    Hi Martin,

    Thanks for your comment!

    See my responses below:

    1) The lowest priority campaign doesn’t necessarily need to be as well-structured as the rest. It can just be one ad group for all your products. If you start getting a lot of traffic in this campaign, then you might want to split it out (or better understand the search terms in this campaign and whether they should really appear here).

    2) Everything else but the brand you are targeting should be excluded in that specific ad group.

    3) Campaign Priority will take presidency.

    4) I use broad match. The difference between broad match and phrase match for negatives is only in the word order, so I don’t bother with using phrase match.

    Let me know if that answered your questions, and if any follow ups happen please do let me know 🙂

  • Martin says:


    I might be confused but there seems to be a mistake in the chart and campaign structure. So we have 3 campaigns (All Keywords, Brand, Top keywords). Clear enough.

    All Keywords = High Prioriy, low bid (+ brand and top keywods negatives).
    Brand = Mid Priority, Mid/High Bid (+ top keywords negatives).
    Top keywords = Low Priority, High bid (no negatives – executed last).

    This way, reversing the Google default priority scheme we can target the bottom funnel. I think the way it is described in the article, you do not achieve the result of segmenting and you are paying the same CPC for your low ROI keywords as for your high ROI keywords. Placing Generic with High piority achieves nothing in my opinion.


  • Andy says:

    Hi Andrew Lolk,

    My shared budget spending is far from exhausted, but the search terms still skipping. I’ve rechecked the settings, priority, negative keywords lists. Everything is fine. It’s even more weird that for same search term, sometime it skips, sometime doesn’t. Would you please help me troubleshooting this. Thanks very much!

  • Andrew Lolk says:

    Hi Andy,

    Double check that you are looking at the newest time frames. Granted that the negative keyword is in place in Generic, then there is no way for it to appear back in the generic. Sometimes a term will skip from high converting to generic, but it’s rare and usually not that impactful.

    If the bid in fallback or high converting is substantially higher than the bid in generic, then the search term can skip. It’s mostly seen in fallback campaigns though.

    Also, if you run the fallback campaign as a CSS (if in Europe), then the search terms also skip a lot.

  • Andrew Lolk says:

    Hi Martin,

    I think you are slightly confused (without offending anyone).

    There are two different options.

    The proposed option in this post want all keywords to appear in generic unless we negative it into High Converting or Fallback.

    “I think the way it is described in the article, you do not achieve the result of segmenting and you are paying the same CPC for your low ROI keywords as for your high ROI keywords”

    Segmentation works using this approach. I’m not sure where the confusion pops in.

    Your example would also work. It’s just a different way to skin the same cat. In our approach, we don’t differentiate between a branded term and a top keyword. We just have “top keywords” no matter if they are a brand, model, generic term, product name, etc. It doesn’t matter to us.

    “Placing Generic with High piority achieves nothing in my opinion”

    Compared to the way you show it, then it’s just a slight change.

    Here is how it works with the strategy in the article:

    • Generic = Any keywords that perform average or have not been processed yet
    • High converting campaign = Top performing keywords
    • Fallback = Low performing keywords

    So I’m not sure why that wouldn’t work.

  • Besh says:

    Questions: What bidding strategy would you recommend for this setup. And should the strategy be the same for all 3 campaigns (top funnel, middle funnel and bottom funnel). I’m finding that my TOP campaign has really high CPCs for some queries when it shouldn’t be (With tROAS).

  • Andrew Lolk says:

    The setup still allows you to run your favorite bidding strategy. Whether it’s manual, smart bidding (like tROAS) or something third.

    If you lean towards smart bidding I would:

    1) Run them separate (but still testing moving to portfolio)

    2) Have Fallback be manual bids

    3) Potentially set different targets for generic vs high converting.


    But, one thing to keep in mind is that seeing that Smart Bidding is already bidding at the Search term level, then the three tier approach might be overkill. It might be worth just having a two tier approach. So you have a “Regular” campaign and then a Fallback campaign with low CPCs.

    That would simplify the setup, and we love simplifying things here 🙂

  • Arturs says:

    Hey Andrew!
    This is a great read, thanks for that!

    We are currently working on a small project that only has 1 product (with a few variations), but from search term report we see that it appears in various search terms that are relevant/semi-relevant and irrelevant.
    We are thinking to adjust your structure so we can more or less control the bids for most relevant keywords (around 10-15) separately. Basically, we want to do it by excluding KWs on an ad group and campaign levels.

    We are running shopping ads in multiple countries/languages.

    What is your take on this and Will this work from Google point of view?

  • Arturs says:

    Hey Andrew!

    Really cool article, thanks!
    Wanted to ask question how well will this work if we have only 2 products with very similar search terms?
    Generally we appear for search terms that include one of the following words – laptop, monitor, stand, riser (+ adjective – wooden, high, small, ikea etc).
    We were thinking to split Campaigns by high/low performing search terms and find it difficult to add/structure negative KWs.

  • Andrew Lolk says:

    Hi Arturs,

    Yes, it sounds like a perfect use-case for this kind of structure 🙂

  • Andrew Lolk says:

    Hi Arturs,

    Let me shoot you an email, so I can see how you set it up, and get a better understanding of your challenges.

    We’re just about to update this article, so it’d be perfect timing to see where you go wrong 🙂

Leave a Reply

Your email address will not be published.