Advanced Google Shopping: data insights from behind the scenes

- Luke Metcalfe

At Crealytics, we love experimenting with Google Shopping to see what works and what doesn’t. There are all kinds of campaign optimization tricks to be discovered when you spend time testing. What’s more, as the hunger grows for more in-depth Google Shopping best practices, it’s crucial to stay on top of changes and developments in the Google’s algorithm and the sector itself. Luckily, we’re pretty nerdy when it comes to this kind of thing!

Last month at SMX West, I treated an audience to a presentation on Advanced Google Shopping, in the hope of addressing some of the biggest questions on the lips of PPC marketers today:

  • How should current changes in the Google Shopping economy affect my approach?
  • What impact does price have on performance and ranking?
  • How should I test my Google Shopping product images?
  • How can I ensure my product titles are performing?

I’ve compiled some of the most useful bits so that you can apply them to your Google Shopping strategies.

The Google Shopping economy

2016 was a great year for Google Shopping. In fact, there was such a flood of growth that it surpassed Text Ads in terms of spend. In the first half of the year, Shopping owned a 56% share of all paid search spend in both the UK and US and 57% in Germany.

But something a bit weird happened. As you can see from the graphs below, although Google Shopping clicks kept on increasing, its CPC (cost-per-click) actually decreased.



Why was this?

Well for one, and as predicted by Google in its Alphabet Earnings Report, mobile click share for Shopping has increased profoundly over the past few years. In fact, since last year, the majority of clicks have been generated by mobile rather than desktop.

As we all know, mobile traffic is increasing across the board, however, Google has been intentionally driving more people towards its Shopping results. The screenshots below demonstrate the escalating squeeze that Shopping is putting on organic results; this has resulted in a growth surge of 78% for paid search vs organic’s 16%.

Google has also been increasing impressions for generic terms on its Shopping results, which has, of course, led to an overhead in the queries available to advertisers.

The way that internet users make their journey to purchase is also changing. Navigation is now primarily taking place on Google, as opposed to advertisers’ websites. The diagram below shows how, even though a user might navigate to an advertiser’s site, they will keep returning to Google to continue their research via a main hub. Google is now collectively the first, intermediary and final stop in their journey to purchase. What this could mean for the future is that fewer people browse for products broadly across the web, doing it instead via Google Shopping alone. This opens up the floor for more competition, increasing people’s ability to see another retailers’ products and buy from them.

For every three or more clicks on Google Shopping ads, there’s a 46% conversion, compared to the 29% conversion gained when those clicks take place on a Text Ad. This tells us that Google Shopping ads lead to more purchases by enabling people to undertake the research they need to – at the very point preceding a purchase – and directing them to the checkout.

Due to the fact Google Shopping is more likely to lead to a sale, it’s a more profitable medium. This is made all the more impressive given that in a side-by-side comparison based on clicks, Shopping is actually more expensive than Text Ads. We can see below, that despite higher CPCs, Shopping delivers 62% more Revenue Per Click (RPC) than the Text Ad equivalent.

Google Shopping is also a fantastic performer when it comes to mobile conversion. Marketers have typically found that, despite growth in mobile traffic and clicks, mobile users are hard to convert. The fact that Shopping conversions on mobile are just 3% less than Text Ad conversions on desktop is a testament to Shopping’s potential.

Google Shopping economy: key takeaways

  • Although CPCs for Shopping are higher than Text Ads, RPC and CR are both higher than the Text Ad equivalent, making Shopping more efficient overall.
  • The strong, ongoing growth of ad inventory supply is leading to CPC deflation.
  • Shopping continues to perform well, giving marketers extra competing power.


Naturally, price has an impact on everything when it comes to eCommerce, and with Google Shopping it’s no different. Indeed, impressions and clicks in Shopping can be very sensitive to pricing and price changes.

How does price affect performance in Google Shopping?

Taking a look at the graph below, which shows the knock-on effect of a price increase for Aasics sneakers, we see to what extent clicks are negatively affected by such an increase. It’s significant, with a 5% price increase incurring a 60% reduction in clicks.

The decrease in clicks was so severe, it demanded further investigation into the connection between price and performance. So, after dividing a retailer’s products into two groups:

  1. Products they sold at a higher price than the competition
  2. Products they sold at a lower price than the competition

we then compared their performance.

What we found was that even though both groups were roughly the same size, the cheaper products received 134% more impressions than the more expensive ones. What we’re seeing here (and on the graphs below) is Google’s algorithm in full operation; it recognizes that cheaper products perform better – they receive more clicks, creating more revenue for Google – and consequently, gives those products more impressions than the expensive ones.

Related: A study into Google Shopping’s price bias (and 4 ways to overcome it)

Furthermore, the performance of those cheaper products, compared again to their similar but pricier counterparts is better. In fact, on Google Shopping there’s a 280% increase in conversions for cheaper products, as well as a higher CR resulting in a 30% lower CPO.

These are significantly important insight for retailers: in Google Shopping, cheaper products are king.


When we talk about ranking in Google, there are multiple factors that can affect your position. Price is one of them, but all retailers need to consider:

  • Retailer rating – new retailers will be rated down, so user ratings are absolutely essential.
  • Bid amount – an important one, but one that can be blown out of the water by cheaper product prices.
  • Performance history – more history leads to more impressions, simple as that.

All of the above are considerations for any search campaign manager, but the impact that product price has on position is one that certainly can’t be ignored.

*Data from price comparison provider, based on >3000 queries

As you can see above, in 65% of all cases from our sample, the product with the most competitive price was ranked in first position – no matter what the retailer.

What do people actually buy when they click on a Google Shopping ad?

So, someone’s clicked on your Google Shopping ad. That means they’ll buy the product, right? No, not necessarily.

We analyzed what people purchased after clicking on a Shopping ad, and found that of the 64% that went on to buy from the advertiser’s brand:

  • 34% purchased the initial advertised product.
  • 16% purchased a different product in the same clothing category.
  • 14% purchased a different product in an entirely different clothing category.

Of the 36% who went on to purchase from a brand other than the advertiser’s:

  • 15% bought from the same category.
  • 21% bought from a different category.

What this means for retailers is that the goal of Google Shopping should be to get someone onto your site. Your best bet to beat this competition is to focus your attention (i.e. your Google Shopping ad spend) on your most competitively-priced products. Getting people out of the Google marketplace and to your website is the most important thing. Even if you don’t make as much as you would from a product at a higher price, that cheaper product will show up higher in Shopping results, consequently increasing the chance of someone clicking through to your site and browsing there.

Pricing: key takeaways

  • Don’t end up overbidding on expensive products – consider changing product prices instead.
  • Discounting only a few select products can be a powerful strategy when driving traffic to your site.
  • One day, price and bid management will be merged.

Image testing

Image testing looks to be the next game-changer for click through & conversion rates, and many marketers are tuning into its potential. The issue, however, is that it’s still a bit of a black box, with people scrambling over how to crack the secret code.

Image testing challenges

It makes sense that image testing should be a key part of any retailer’s Google Shopping strategy, but there are several challenges to overcome.

  1. Image categorization – if you don’t categorize and organize image types before the test, you won’t receive accurate results.
  2. SERPs vs Shopping tab – in the Shopping tab, products get clustered into just one offer list; your image may not be on top.
  3. Time lag – it can take up to 72 hours to index a new image when the accompanying URL is changed, and up to 6 weeks if the URL remains the same (server-side image swap).


The data we used to build insights on image testing came from two retailers, with 1800 images changed in total. We tested the effects of using images that featured a modeled product and a product in isolation. Up to 40 million impressions were taken into account, and we overlooked any products that were not consistently available.

Tip: when testing, change images only in your test campaign, whilst running a control stream as your baseline. Also, remember to compare KPIs throughout the test to eliminate seasonality.

After running tests with the two retailers, we found that only one had a clear, positive result.

So what does it all mean? When the image type regularly varies in Shopping, i.e. when a range of image types are displayed by competitors, there’s no significant benefit to having an image that features a model. If, however, your approach is totally standout in the list, you might be onto a winner, as we found when we got a 27% increase in CTR via these means.

Image testing: key takeaways

  • Image testing can have an effect on your performance in Google Shopping.
  • There are ways to stand out, even when many others offer the same product.
  • Images are important in allowing Google to match queries.
  • There are areas where testing can have a more significant impact, making the return on effort much larger.

Feed title optimization

We’ve done a lot of testing in feed optimization over the years, and though Google claims that everything in the feed matters, we have to disagree. According to our tests, it’s only titles that appear to make any real difference to campaign success (at least for now).

In the graphs below, you can see how changes to product titles yield the best results.


When delving into the effects of feed title optimization, we analyzed four accounts, from three clients, each in its own vertical. In total, we changed 16,800 titles and generated 5.2 million impressions.

So, what did we find?

Using a strategy that focuses on underrepresented queries, e.g:

  1. ‘low share of voice’ Shopping queries that aren’t included in titles and could get more impressions according to market size, or
  2. text ad queries – terms that work well in text ads but don’t receive Shopping traffic proved to have the strongest positive impact. As you can see from the graph below, the traffic coming from text ads was indicative of the potential for impression volume growth via Shopping.

Low-or-mid-level traffic queries with market potential are the best choice when it comes to feed title optimization.

We found that the impact a simple but strategic title change had on impressions and share of available traffic volume was significant.

Including existing queries can also achieve further success.

When we included existing queries in addition to changing the title, impressions, and share of available volume, were still affected positively.

Feed title optimization: key takeaways

It’s old news that a well-organized feed is beneficial, but your optimization efforts shouldn’t stop there. The title you give your product is one of the most competitive actions you can take in addition to managing the structure of the feed itself.

  • Feed title optimization works – delivering up to 30% more impressions.
  • Focus on product-related queries and low share of voice.
  • Don’t overwrite important terms, because this will negatively affect your gains.

What now?

Everybody knows how quickly things change with Google, but regular testing on key components such as image, feed optimization (especially titles), and pricing will help you consistently stay ahead of the game. As always, hasty decisions never pay off, so prepare for any negative impacts that events such as price changes will incur, and don’t override existing gains by implementing unsuitable strategies over the top.

If it sounds like a lot of work, we’ve got you covered. Camato is the search marketer’s Google Shopping sidekick, developed to take the bulk of manual effort out of your workload so that you can focus on the most important things. From campaign structuring, to bid management, Camato optimizes your Google Shopping activity, making it more profitable without adding layers of extra complication. Want to find out more?

Get in touch!


Here’s the original slide deck


Luke is a Content Marketer at Crealytics

Leave a Reply

Your email address will not be published. Required fields are marked *