Advanced Google Shopping Optimisation Tips & Techniques
Shopping campaigns are a vital tool in the PPC arsenal of almost all e-commerce businesses. Not only do shopping campaigns provide a highly effective ad format, which highlights the visual aspect of a product, they are also highly automated, making them reactive and responsive. However, because they are so automated, the user also has far fewer options when it comes to targeting, in comparison to standard search ads. This can be of benefit though, as Google will use all the tools at its disposal to display an ad to the people searching for the right query and the right product at the right time.
The problem with shopping campaigns comes when the advertiser wants to optimise and gain better control over their campaigns. As mentioned, they are automated, so Google does most of the targeting automatically. This means that the advertiser needs to find innovative strategies to ensure their ads are pushed to the relevant audience.
In this article we’re going to outline some practical tips and techniques on how to optimise Google Ads shopping campaigns to get a positive Return On Investment (ROI), reduce cost per acquisition and increase sales for your business.
Basic Targeting Options
The biggest factor in Google’s decision on which search queries should be shown a particular shopping ad are defined by the shopping feed. The shopping feed is located in the company’s Google Merchant Centre and is used to list all the products for sale. However, a company’s feed is much more than just their list of products. Google will use all the information and attributes defined in the feed to decide when and where to show the related ads. With this in mind it clearly makes sense to make sure that the shopping feed is as rich in data as possible. Not only will that allow Google to show ads in the most relevant search query results, but also to not show ads for irrelevant search queries that waste budget.
Attributes are columns in the shopping feed which define certain characteristics of the products being advertised. They also help to define what kind of product it is that is being sold. For example, Google provides a taxonomy document which is used in the “google_product_category” attribute, in this column you can select the best option for your product from thousands of product types i.e. from ‘animals > bird cages’, to ‘activewear > water sport accessories’. Google will then use this information along with the other attributes in your feed to dynamically select the targeting for a given product.
When setting up a shopping feed, it is essential to carefully consider the following three attributes. These attributes are arguably the most important factors in creating a shopping campaign and will facilitate targeting in the most effective way possible:
- Product title - Run title testing (Brand, Gender, Product Name, Size) to see if the brand at the start of end of the name is better... use feed rules to do these kinds of tests. The standard format is the one shown above, however, one size doesn’t fit all and should be tested. This format in practice could look like this. Aldo Women’s Silver Strappy Heel - Size 5. However, you could test different formats such as Silver Aldo Strappy Heel – Women’s.
- Product type - Google product category used to be the most important, but now your own definition of product type carries more weight. Include at least three product levels which are descriptive. Keyword stuffing in this is fine (but make sure it’s contextual and makes sense). It is possible to use your website information architecture and navigation to help develop product type entries, providing those are clear and well structured.
- Price – the price of your product is your main and most significant differentiating factor on a SERP (Search Engine Results Page). Price is also known to act as a quality score. The lower your price the better (in comparison to your competitors). If you have a product in a range of sizes, consider focusing ads on the smaller size as the picture is likely to be similar or even indistinguishable at the size it’s shown, but the lower price will benefit the performance of the advert.
Campaign priorities are fundamental to running any kind of Google Merchant Centre account with multiple shopping campaigns. It might be the case that you’ve split your campaigns by product type. For example, fashion retailers might split campaigns by men’s and women’s or trousers and tops etc. However, if there are products that fit into multiple campaigns it can start to cause issues, because running the same product in multiple campaigns may increase the bid amount. This happens when campaigns from the same account compete for the same search terms.
Using the Campaign Priorities function can help avoid this problem. There are three campaign priority levels, namely low, medium and high. When more than one campaign is set to bid on a search term, the campaign with the highest priority will bid. Eliminating competition and maximising budget value. For accounts with just one campaign, Google recommends setting the priority to low. However, for accounts with a more complex structure using a mix of priority levels for various search terms can be invaluable in making sure ads do not compete.
Whilst campaign priorities are most commonly used to avoid competing with yourself. They also offer an opportunity to get creative. One of the best examples of this is a technique involving negative keywords and campaign priorities to filter and prioritise search queries. For example. Imagine an account with campaigns around shoes. They may have three shoe campaigns running as below.
- Shopping – Shoes (High)
- Shopping – Shoes - Summer (Medium)
- Shopping – Shoes – Summer - Sandals (Low)
For a search query to reach “low” it will need to pass through both “high” and “medium”. Campaign priority is simply a means of telling Google where it should start so each click costs the lowest possible amount.
The High priority ads in a campaign are designed to capture only the most generic searches with low purchase intent and so are set with a low budget. Their purpose is to educate and they are aimed at users at the start of the Buyers Journey [Link: https://www.innovationvisual.com/insights/mapping-content-for-different-stages-of-the-customer-journey] .
Medium priority ads are designed with slightly higher bid amounts and are aimed at searches for terms with greater purchase intent than high priority ads. Their purpose is to give the searcher the information they are looking for during the consideration stage of the Buyers Journey.
Finally, low priority ads are set with the highest bid budget and are targeted at the most purchase intent focused search terms. They target decision makers at the end of the Buyers Journey, the ones ready to make a purchase.
In order to ensure the right ad is served at the right time to the right person it is possible to use negative keyword lists to filter search terms and ensure they are matched with the right campaign. These lists act as filters pushing high purchase intent queries to the low priority campaigns with the highest bids; ensuring we only spend more when it is most likely to result in a sale, conversion or provide the most value.
These negative keyword lists can be made up of a number of different things including “product names”, “brand names” and “style names”. In general, the more specific the term, the more likely it is that the user knows exactly what they’re looking for, meaning they are more ready to purchase than someone searching for a more generic item or phrase.
Applying negative keywords lists to high-level campaigns ensures those high value search terms filter through to the better matched low priority campaigns and so get the highest bids. The stages go as follows:
- Someone searches for “Aldo Silver Strappy Heel”.
- “High” looks at this first, however, there’s negative keywords for “Aldo”, “silver” and “strappy heel” as these all indicate a higher value of purchase intent.
- “Medium” then looks at this search term. However, there is a negative keyword for “Aldo” against this campaign. It is important to note, if “Aldo” hadn’t been present as a negative search term this campaign would have been served to the searcher.
- “Low” now looks at this search term. This is the highest search intent campaign. As the last campaign in the priority flow it will now bid for this query. This campaign is also likely to have a higher maximum CPC (Cost Per Click) allowing it to bid more to be served and so is more likely to show and show in a better position on the SERP. “Low” doesn’t have any prioritising negative keyword lists.
To summarise, structuring shopping campaigns in this way helps to focus ad spend where it can more easily convert to a sale. It can also be extremely helpful with a limited budget as it means ads will still be shown for broader terms but the majority of spend will be dedicated to those search queries with the most purchase intent and so the best chance of converting.
Google regularly releases new format ads, whether it's RSAs (Responsive Search Ads) for search, RDAs (Responsive Display Ads) for display or even the recently announced gallery ads for search. Ad formats come and go, some stay and evolve while others are removed entirely, Shopping is no different, and depending on the products being sold, it may be possible to make use of the shopping ad evolution titled “Showcase Ads”. You can read more about the implications of this format in more detail in our Google gallery search ads introductory article.
In short, showcase ads provide increased space to entice a searcher to click through to a website. Showcase ads allow users to include an additional image alongside the products as well as space for some well-thought-out copy. They also allow users to display a collection of products. This means a searcher is more likely to see their ideal product and click through.
Showcase ads are a brilliant format, however, they won’t work for everyone. In our tests we’ve seen that they rarely show on more niche product types. Instead they’ll show for those high volume and arguably competitive search queries like “dress” or “shoes”. That’s not to say you shouldn’t try this format for yourself. If you’re running a shopping campaign or are planning too, why not trial this format for a couple products to see if it works for you?
In another twist by Google, they’ve also launched an additional campaign subtype for shopping, called “Smart Shopping”. Smart Shopping is a combination of standard shopping and display remarketing campaigns. Google uses the information from a feed (as it would a normal shopping campaign) and automates the targeting. The real difference between “Shopping” and “Smart Shopping” is that Smart Shopping ads will be included in the display network to deliver a far wider reach. If the purpose of a campaign includes brand awareness as well as sales using Smart Shopping could be an ideal strategy.
Our experience with Smart Shopping to date suggests that the standard format shopping campaigns (when fully optimised and properly managed) tend to perform better. On the other hand, these wider reaching campaigns can often have a useful, broad based effect at the start of the sales funnel to help draw people in. It is important to remember searchers are unlikely to convert on their first visit, so utilising re-marketing strategies, where possible, are a key to success on the Google ads platform.
Other agencies and leaders in their field have also shared similar results with us, with some campaigns performing extremely well and others failing to provide competitive results. So, this format is all about testing.
When testing this format for yourself we recommend the following:
- Don’t pause your current campaigns, instead take some or one of your best-selling products and test it in this new format.
- Leave the campaign to run for at least 3 weeks (the longer the better). Automated and smart bidding campaigns need time to optimise. It’s likely your campaign won’t see positive results in the first week. Depending on the number of conversions you regularly receive this could take longer.
- Smart Shopping doesn’t let you see placements or much of anything really (which makes it a bit of an enigma). However, this doesn’t mean you should leave it to its own devices. If you see something strange or it’s costing you huge amounts of money, adapt or pause the campaign while you figure it out.
- Focus on conversion value or conv. Value/cost. These figures will tell you what’s working. To do this successfully you will need to have enhanced e-commerce tracking setup beforehand.
It is important to note this campaign type is still in beta (at the time of writing) and will improve as time goes on. What might not work today, may be your ideal campaign type in a few months to a year.
Search Term Analysis
When running a shopping campaign, you should always aim to review your search terms on a regular basis. This ensures you’re not wasting money on irrelevant searches and on searchers with no purchase intent. As Google does all the targeting for you in shopping campaigns, this is even more important than in organic Search. Negative keywords are the only way to truly ensure your product ads are being shown to the right people for the right things at the right point in their journey.
Search term analysis can be a daunting task, especially on big accounts which receive thousands of clicks on any given day. However, there are strategies to manage it.
- Regular analysis will reduce the amount to review in one go. Keep on top of it and it will be far easier to manage.
- Focus on the search terms which appear more than once and are costing you the most money.
- Utilise phrase and exact match. In a lot of cases, broad matches can be, well, too broad. Google also defines negative keywords differently in shopping.
- The more you optimise your campaign’s negative keywords the easier it will become. At a certain point you’ll start to see fewer and fewer negatives and you will know your campaign is optimised. However, you should always keep an eye on this, as what people search for changes and how Google targets your products will change to.
- Take your search terms from a smaller window. If you haven’t reviewed your search terms in years. Just take them from the last couple months. Search terms will often be repeated and if there’s a serious issue you’ll still be able to see it.
We’ve touched on using negative keywords to segment search queries in priority-based campaigns. However, another strategy is to use negative keywords to segment by “branded” and “unbranded”. This way you can focus spend on your branded search queries (that have more purchase intent).
This is another example of how to focus spend where it’s needed and where there is most value available from a user clicking on your ad.
Scripts can be a great tool to use on shopping campaigns. Not only do they take a lot of the hard work out of tasks like search term analysis, they also help to optimise a campaign better and faster than any human can.
An ideal script for a shopping campaign works by adding a search term as a negative keyword if it doesn’t include a certain range of keywords (from a predefined list). This allows search term analyses to be automated when a campaign may be too large to review manually or have too many generic terms.
There are a whole range of scripts available, many can be altered or adjusted by a developer to fit your needs and some are ready for use out of the box. If you have the budget, scripts can also be made from scratch to complete a number of complex functions specifically tailored to your business.
Ad Schedule (bid adjustments)
Every campaign you run should have some kind of ad schedule. This is an essential and basic principle of PPC marketing. You want your ads to show when you’re most likely to get a conversion or sale. Now this point is moot if a business has an unlimited budget and an endless supply of cash. However, this is rarely the case and most people will want to optimise their account to get the best ROAS (Return on Ad Spend) possible.
Depending on your goals you will want to define your ad schedule based on either conversion value, revenue per click, conversion rate or ROI. Which metric you decide will depend on a number of factors. Traditionally conversion rate is the best indicator and will work for most accounts. However, if you sell a more diverse range of products with large changes in price, profit margins, a more ROI or RPC (Revenue Per Click) based ad schedule may work best.
How to work out your ad schedule bid adjustments:
Conversion Rate Based:
Bid adjustment ratio = [(Total CVR / Device CVR) – 1] * 100
Revenue per click based:
Bid adjustment ratio = [(Total RPC / Device RPC) – 1] * 100
Return on investment based:
Bid adjustment ratio = [(Total ROI / Device ROI) – 1] * 100
Lastly, when you’re building your ad schedule it is important to remember to take bid adjustments with a pinch of salt. There are a huge number of factors to take in when working these out. Not only should you look at the hour of the day, you should also look at the day itself, the time of year (seasonality) and much more. Additionally, you should try to avoid potentially skewed data, if some days and times only have a small amount of data, but have 1 conversion or sale, they might look amazing times to optimise for. However, this could be argued as an anomalous result that should be ignored (until there is more data to support a different tactic).
Bids and Automated Bidding Strategies
We touched on how to optimise bids in the previous section with ad schedules. However, possibly the most important aspect of optimising any campaign not only for Shopping Campaigns but Search Campaigns too, are the bids. The bids are what ensure ads show in the right SERPs and also, when used correctly, provide value for the business.
If bids are too high or too low, they won’t deliver good results. Too high and they won’t give positive ROAS, too low and they may not show at all (and definitely not to the right people). Bids can also be limited by overall budget. In competitive markets budget limitation is the norm and in some Search Campaigns for example, it is common to see smaller clients, with smaller budgets use up their entire daily budget on one click. Overall it is often necessary to invest in PPC before seeing a return on the investment.
Shopping Campaigns are the perfect place to utilise automated bidding strategies in Google ads. As there is less data available on the relevant bids to use for a given product, automated strategies help eliminate the guesswork. These strategies also use Google’s increasingly advanced machine learning algorithms and other clever things to optimise the ads automatically. The data Google uses to do this is hidden from advertisers. It includes data on searchers interests, their recent searches, where they are in the Buyer's Journey for any given type of product, as well as other relatively vague data that Google collects and audits from its many sources. Put all this together and you can have a campaign which is entirely focused on providing the best results for your given objective.
What Options are Available?
Maximise Clicks - Maximise Clicks is an automated Bid Strategy used to generate as many clicks as possible within your specified budget.
Target ROAS - Target ROAS (Return On Ad Spend) lets you bid based on a target ROAS figure. This Google Ads Smart Bidding strategy helps you get more conversion value or revenue at the target ROAS you set.
For most people we would recommend using “Target ROAS” when using this strategy so Google can aim to get the best return on ad spend automatically. The trick is to ensure the set target is managed effectively. If a campaign is performing well, try increasing it in stages, this way Google will try to deliver even more value. It is important to watch these campaigns carefully though, a significant drop in impressions might mean the campaign is not being entered into enough auctions because Google doesn’t think it can't get enough value against the established ROAS, so won’t show at all. In these cases, the “Target ROAS” should be lowered incrementally until performance improves.
Periodically it is important to switch to enhanced CPC just to ensure Google can refresh its data, if campaigns start to stagnate, this may be the cause. Switching allows Google access to more up to date information as leaving “Target ROAS” on permanently can risk Google over-optimising and damaging returns.
As a final piece of advice, when launching a fresh campaign, it should be started on Enhanced CPC or manual CPC. This is because there won’t be conversion data for Google to optimise for at the beginning. Once the campaign has accrued a significant number of conversions it can safely be switched to an automated strategy.
Google Shopping Experts
Hopefully you’ve found this guide useful and you’ve gained a better understanding of some of the more advanced and technical aspects of Google Shopping campaigns. However, if you would like more support in the strategy, creation and optimisation of your shopping campaigns, feel free to get in touch with our friendly team of PPC experts.
Our team of experienced team of Google Ads and e-commerce shopping experts have worked with numerous clients across a range of sectors to deliver exceptional results and ROAS, why not get in touch to see how we could start making Google shopping ads work for your business?