10 Ways to Save Money
on Google Shopping

Digital marketing agencies live and die by the digital audit, and our own agency is certainly no exception. Nailing the breakdown of a potential client’s current strategy and where the opportunities lie hidden is essential to structuring any successful pitch. Recently, we’ve noticed a worrying trend among many of the Ecommerce companies that we’ve assessed:

They all waste a lot of money on Google Shopping.

We aren’t even suggesting the usual marginal overspend on say brand awareness campaigns that don’t push conversions. We’re talking excessive investment into products that haven’t converted in months, or sometimes ones that have never converted at all. And these products all rub elbows with the highest converters, with not a single bit of variance in the strategy.

Not to worry though, we’ve pulled together a list of 10 best practices for
avoiding waste on your Google Shopping campaigns.

1. Optimise your feed

The basics- We’ll assume most people reading this will already have a decent handle on at least the basics of Google Shopping, but for those who don’t, the way the system works is in tiers. Your Google Ads interface will pull product data from the Google Merchant Center (GMC) which in turn pulls data from a CSV file of your own creation called the feed. The feed is essentially an ever-changing sheet with constant columns denoting everything from Product ID to GTIN and even stock. So long as all three parts remain in harmony then your shopping campaign should have minimal issues.

GTIN Numbers - GTIN stands for Global Trade Item Number and it is the standard by which Google operates. Missing or incorrect GTINs can absolutely torpedo your GMC and indeed is typically the most common of product errors we encounter. Using you own internal product ID numbers will not suffice. Luckily, the GMC will alert you when this is the case, and then it is simply a matter of fixing your feed to allow your available products to be shown. This is one of the most common errors we see and also one of the most egregious, so be sure to review your GMC to make sure you are not doing the same.

Include product variants - Poorly optimised product titles are a great way to ensure that your products never see the light of day on Google. Typically, the matter of poor titles comes down to length and clarity. The more specific your title, the better and these should always be some variation on Brand + Product Type + Colour + Material, with allowances for whatever categories that would differentiate your products from one another on an individual level. This way you can better match your own products to customer intent, driving conversions across the board. We also suggest attempting to place your most relevant keyword that you hope to appear for as close to the front of the title as possible, even on the very beginning if possible.

Utilise Product Types and Categories – Of course, you can go beyond just title optimisation and examine your Product Category. This can be a frustrating exercise at times, as these are pre-defined categories from Google that you’ll be assigning your products to. There is a list that can be downloaded from Google here. Some are vague, for example – “microphones” is a single category under:

Electronics > Audio > Audio Components > Microphones.

This is sort of broad when you consider that it also can go the level of “Floral Tape” vs “Decorative Tape” by going to:

Arts & Entertainment > Hobbies & Creative Arts > Arts & Crafts > Art & Crafting Materials > Crafting Adhesives & Magnets > and THEN choosing either Floral Tape or Decorative Tape, depending on your product.

So you can see where the headache comes in. The best way to tackle this is to make sure your Product Types are all filled out properly, before setting automated bulk rules in Excel to match Categories to Type. While the field is optional, and updating the feed can be time-consuming, we suspect that Google will continue to move away from Product Title keywords as a method of Quality Score Ranking, and move into the other product fields. Essentially, the more you can tell Google about your product, the greater the chance it has of being shown.

And if you don’t want to tackle the feed optimisation solo, there are a variety of paid online tools like Feedonomics or DataFeedWatch that can help streamline and customise your feed. These tools provide an easily manageable way to essentially pull data from a source like Oracle or Shopify and then use that data to update your Google or Bing Shopping feeds (or even Facebook or Amazon). This takes a lot of steps out of manual feed management, freeing up time to optimise the account in other ways.

2. Separate products based on margin

To start this off think of a simple exercise. Let’s say you sell novelty holiday jumpers for dogs. You sell them all for similar prices, but the material and labour costs increase from your Chihuahua line, to your Labrador line, to your Great Dane line. Your ROAS is about a 4 to 1 across the board, everything is good, right?

Wrong! Your Chihuahua line has a profit margin of 95%, while your Great Dane sweaters, so labour intensive, only have a margin of about 35%. The overall ROAS is good, but the product focus is all wrong.

What you need to do in these situations is to separate these products based on their product margins to better understand what your bid strategy should be. Obviously, your bids should be much higher for those products that fall into your high-margin group while your low-margin products should have reduced bids. These are categories that you can apply to your products in the feed itself, and then you can make automated rules on bidding, or at least make manual bidding a little easier on yourself (more on automation later).

3. Utilise single product ad groups (SPAGs)

Or product groups in this case. A big mistake we see in terms of account organisation is that a campaign will have an product group with anywhere from 20 to 100+ products in it. Edits are being made as time goes on at the product group level, and rules and bids are set across products that have nothing in common in terms of margin, price, CTR, and conversion rate.

If you’ve already got your product groups set up to contain multiple products, you can see why splitting them all out into new product groups could be time-consuming, especially if your feed is quite large. There are a few tools to help you do this and indeed Excel is always a reasonable option, but there are also tools like PPC Samurai or Optmyzr that will help cut down on your time investment.

4. Move your marketing spend to the most valuable search terms

We’re all familiar with the Pareto Principle (i.e. 20 percent of X generates 80% of revenue/costs/etc.). Much of the same applies to your product groups on Google Shopping. 80% of the budget will usually go to 20% of product groups. This means that if you sort the product groups by cost, the first few groups will get most of the budget that you have allocated for the campaign. Therefore, it’s also important that you don’t just let Google pick this 20% for you. You know your products best, and focusing on the high margin products, or those with typically low CPAs will yield better results than an automated strategy.

Of course, it’s also then important to appropriately deal with search terms that have poor performance. Luckily, examining your search terms in Google Ads will shed some light on the matter. You can specifically filter these terms to include poor performing terms. Obviously, these filters can be adjusted to your specific needs, but we would recommend at least starting by finding terms that haven’t had at least 50 clicks over the last 30-90 days (depending on the typical buying volumes and purchase process lengths for your products). Once you have a list of these poor performers, you can rearrange them into their own campaign, and set a much lower bid on them. This will help you keep your spend low on products with poor conversion rates, without removing them from contention entirely.

5. Use negative keywords correctly

This rule dovetails nicely with the one just above it. If you’re seeing a lot of low CTR and CVR in your campaigns, then it is likely due to traffic quality, i.e. you’re attracting customers that don’t have purchase intent. One of the best ways to help mitigate this effect is by efficiently leveraging your negative keyword list. For example, if your ads are appearing when a user searches “[product name] reviews”, you will want to add “reviews” to your list of negative keywords, especially if these searches are not converting. Consider using Filters and your desired KPIs to find specific groups of potential negative keywords which you can then add to your ad groups or at the campaign level.

There are lots of further tips on negative keywords that would merit an article in their own right. Obviously, these are lists that in many cases will change a good amount from company to company and even a single account may just change over time. There’s a certain amount of trial and error involved here, but in the meantime, here are a few generic keywords that will help get a list kickstarted:

  • About
  • Definition
  • Example
  • Sample
  • What are
  • Guide
  • News
  • Review
  • Statistics
  • Tutorials
  • Bargain
  • Cheap
  • Free
  • Price
  • Quote
  • Create
  • Craft
  • Handmade
  • Import
  • Export
  • Export
  • Rental
  • Gift

6. Include pricing in your strategy

This is a short, but simple one. Oftentimes, it is easy for a company to get too focused solely on digital promotional strategies, while ignoring the other elements of the marketing mix. Google Shopping is a tool that lends itself particularly well to pricing comparisons. When a consumer makes a purchase through this channel, it is often due to either brand recognition or price-point comparison. It is crucial to examine your competitors pricing strategy and make sure that your product is positioned correctly.

7. Geo-targeting and Ad Scheduling

Here are two concepts we see commonly applied to more conventional Search campaigns, but rarely in Shopping campaigns. Best part is, these are both really easy to implement. Try to shoot for around 90 days of data here before you start making decisions.

Geo-targeting

Google Shopping is typically country specific, and furthermore only available in certain countries. That being said, if you’re in the UK, then chances are you’ve already set this up to cover the country and haven’t thought about it since. You are not going to be searching for new regions to expand to in this step, but rather setting bid modifiers on existing locations based on past conversion data. Again, we want to take a look at areas with low cost and high conversions, and start with bid adjustments of anywhere between 15% to 50%. Really put your budgetary focus on regions where performance is already good to see if you can create even greater returns on your ad spend.

Ad Scheduling

Similar to the above, you’ll want to take a look at your campaign performance, but this time with a focus on the time of day. When are your visitors shopping the most, and at what time are conversions the highest? Set similar bid adjustments to the above and see if you observe returns from the increased bids.

It goes without saying that whatever positive bid adjustments you make in this step can also be applied to other times and regions as a sort of “negative rule” too. If there’s a time of day or region that has minimal conversions, create some bid adjustments to lower the spend in that region. See what can be adjusted to ensure you are not pumping money into a city or country that doesn’t convert, or that you are not losing money to top of funnel searches at 3 a.m.


Remember that adjustments can also be made based on device, and this could be possible using similar strategies to the two above. Does most of your business come from mobile shopping? Crank the bids up there, while dropping them on desktop. And so on. These rules are really easy to get up and running, but make sure you continually check the data to make sure you are not missing any opportunities for conversions, or excluding potentially profitable times, regions, or devices. Just because they’re easy to set doesn’t mean you have free-reign to run the campaign on auto-pilot. “Fix and forget” solutions don’t exist in PPC — be vigilant.

8. Examine your Lin-Rodnitzky ratio

This ratio may sound like it’s complex, but it’s actually quite a simple calculation you can use to look at account efficiency. It can be calculated using the following formula:

In order for this equation to produce a meaningful output, you should be examining at least 2-3 months of data. Once the calculation is complete, you should be getting a number between 1.0 and (hopefully) below 3.0. But what does it mean once calculated?

1.0 – 1.5

The account takes no risks. This probably means that the only search terms getting any traffic are brand terms or the heavily targeted search terms. You’re missing out on incremental conversions, most of which are likely to be profitable. Look at your negative keyword list and see what is dragging you down.

1.5 – 2.0

The account is doing great! There are consistently quality search terms that always bring in sales as well as more experimental queries that are tested to better identify opportunities for growth.

2.0 – 2.5

The account takes too many risks. There are too many search terms getting clicks that do not lead to conversions. This inconsistency is likely due to a certain amount of account negligence. A shopping campaign will only give you search terms and negative keywords, so in this particular case, it would be necessary to broaden your negative keyword list by adding irrelevant terms that you’ve found in your search terms report.

2.5 +

The account is being run poorly. Money is being spent with minimal returns, and even small changes will likely save a good deal of money.

When the ratio is too high:

spend less on what doesn’t work. This includes:
  • Adding poor performing search terms as negatives.
  • Lowering bids for products with a high CPA or low ROAS.
  • Increasing bids for products that deserve it based on their value per click and your internal targets.

This high ratio is much more common than a low one, as account negligence is more common than death by micromanagement. It’s worth noting here that if queries are relevant but still not converting, then the problem may be larger than your account management, and could be something more like an issue with your landing pages, ads, competition, prices, etc.

When the ratio is too low:

take more risks. You can accomplish this by:
  • Reviewing your negative keyword list. There are terms there that are too constrictive on your account, and your products are not showing for as many relevant search terms as they should.

This is much rarer than a high score but still possible, especially if you’re a bit too strict when attempting to lower your ratio. Chances are you went too crazy when applying your negative keywords. While this may seem like a solid approach, be aware that you are likely missing out on peripheral search terms that can still provide conversions.

9. Calculate your Vallaeys Shopping Efficiency score

Here’s another handy calculation with a more specific focus on shopping campaigns specifically. This one comes to us from Optmyzr, which focuses on Google Ads scripts. It can be downloaded direct from their site and imported to your Google Ads account. This score calculates the percentage of all ad impressions that come from those pesky undefined product groups, i.e. those that are labelled “everything else in X.” This is where our earlier concept of grouping out individual products into their own ad groups comes into play.

If your score is close to 100%, it means that a large portion of the feed has been split into defined product groups that can each be managed separately, and thus your control over the account is appropriate.

If your score is low, then it means a lot of your impressions are coming from the “everything else in X” category and that your campaign should be restructured to drive traffic through better-defined ad groups with greater control.

10. Automate When possible

Given the bulk of this list, you might think that the parting advice here seems a little against the grain. A lot of the actions listed thus far suggest taking control back from Google to clean up your account. However, once the account has been pruned slightly, it’s okay to put some rules in place to make the account easier to manage moving forward, so you don’t find yourself asleep at the wheel again. There are a few different ways to do this

Zero-Conversion Alerts

These can be implemented with a few “if…then” statements in your account that will tip you off if a product has had no conversions over a certain threshold of cost. This threshold can be variable depending on your business but typically we would recommend if a product has cost you more than €50 without a conversion then you should go ahead and pause it. This rule will help you keep the dead weight out of your account moving forward.

Wasteful Spend in Shopping

If you’re looking at Google Shopping Ads, one of the first observations that you’ll make is that the rules function is absent. Since rules are unavailable to us, you will have to get a little more creative. First click on “dimensions” in your shopping campaign. This is where you’ll set up your “zero-conversion” alert and your “unacceptable ROI” alerts, using the same steps as above, but without automated pausing. Because this isn’t a traditional “rule” you have to seetle for automated alerts and manual pausing. Next, click the download button. We would recommend a smaller timeframe for this one, weekly or even daily alerts would be preferred. Ideally this will always be an empty report, but if it’s not then you’ll get a great insight into what isn’t working in your shopping campaigns.

Unacceptable-CPA Alerts

Here’s another “if…then” automated rule that will also help reduce the more common types of wasteful spend as your account continually evolves. Obviously, you’re going to need to establish what exactly the level of CPA is that you deem unacceptable, and this will be based on product margins more than anything.

Once you have your minimum CPA in mind, set this rule to run once every 30 days. If you find a product that is setting off these alerts, then lower the bid by 10%. Continue to do this and consider manual analysis if a product is continually setting off alerts. Sometimes these products will actually be profitable if you look at a timeframe longer than 30 days, but we do know at a minimum that the product is being overbid on at the moment, hence the slight reduction in bidding. This will ultimately help you realise a better ROAS across the account.

Contact Us