Google Marketing Live 2019 – New Products and Multi-Channel Campaigns

5/21/2019

Google just announced some significant changes to its advertising ecosystem at the annual Google Marketing Live event in May.  The theme that the search and advertising giant is focusing on for 2019 is the adoption of highly automated multi-channel advertising campaigns.  Across search, shopping, apps, display, and YouTube, Google is extending the reach of individual Google Ads campaigns to serve across more of their platforms utilizing machine learning algorithms to drive strategic decision making on ad placements and budget allocation between channels/properties.  Additionally, the introduction of a few new ad formats will continue the trend of an expanded footprint for ads on Google products. We will highlight some of the key announcements and what this means for companies who advertise on Google.

New and Updated Product Announcements

Smart shopping campaigns: This campaign type was introduced in 2018 and was built for small businesses to utilize their product data to deliver ads across Google.com, Google Display Network, YouTube, and Gmail.

Shopping campaigns: Traditional shopping ads will now be eligible to serve on YouTube and Google Discover in July of this year.

Showcase Shopping ads: This highly visual variation of shopping ads will be extended to serve on Google Images, YouTube and the Google Discover feed in addition to its current placements within select Google Shopping results.

Google Shopping Actions: This relatively new feature that was released last year enables customers to complete transactions through the Google Shopping Actions cart without leaving the Google infrastructure.  This feature will be further integrated into Google Shopping results and extended to Google Images and YouTube (already served on Google search and Google Assistant).

Discovery campaigns: This new ad format will be a swipeable carousel style ad eligible to serve within the Google Discover feed, the YouTube homepage, and on Gmail.

Gallery Ads: A new ad format that will be introduced to all advertisers later this year, this highly visual format will be served at the top of Google Search results as a scrollable/swipeable carousel of between 4-8 images, up to 3 headlines, and a 70 character tagline.

Smart Bidding Update: Smart bidding is Google’s use of their machine learning expertise applied to Google Ads bidding units in real time.  Working to improve the functionality, they have added a few features, most notably seasonality adjustments which will help the machine learning algorithms to adjust to performance shifts during promotions or seasonal variations.

Automation vs Human Intelligence in Digital Marketing

There is a need for restraint and skepticism when increasing the utilization of computer-aided automation within digital marketing.  There are repetitive aspects of the management of digital advertising that can benefit from automation, but the best results are realized when human intelligence is the driver of strategy.

Think With Google recently interviewed Marie Gulin-Merle, Calvin Klein’s chief marketing officer and PVH’s chief digital officer.  As they discussed what she called “the end of digital marketing” describing the increasing importance of digital on marketing, she made a couple of key points worth sharing.

First, to underline the importance of digital marketing to retail sales, she indicated that 50% of sales are digitally influenced.  This is a number we have seen in our work with retail clients as well and is underestimated by many companies when determining how much of their marketing budgets should be allocated to digital channels.

Later the discussion touches on the use of machine learning and digital analytics (“tech”) in the field of marketing, to which she says;

“… I think tech helps on, level 1 automation and repetition of what works, level 2 identification of patterns of what works, what doesn’t work, and then you need human beings to make sense of the data.  Great marketers will be the ones who really know how to embrace tech the best. … at the end of the day, marketing is about ideas.”

You can find the full interview here.

https://www.thinkwithgoogle.com/marketing-resources/data-measurement/digital-marketing-mindset/

We think this is important in the context of some of the changes Google is making to Google Ads campaigns, specifically when they are expanding the reach of individual campaign types across Google’s platforms which is increasingly reliant on machine learning to drive the strategy and performance.

We have found the best results when creating omnichannel strategies that are optimized at the individual channel level, where we use ads that are customized/optimized for a specific platform.  One example that concerns us is in how the new smart shopping campaigns use product data to display ads on the search network but also generates fairly generic advertisements that run on the display network, YouTube, and Gmail.  We tested these, but get better results when utilizing traditional shopping campaigns to display Google Shopping ads and display campaigns to run dynamic remarketing creative that is made by a creative professional and incorporates a company’s brand image and message.

Case Study: Specifically, for one client we adopted who previously utilized smart shopping campaigns, we were able to double revenue while reducing ad spend by over fifty percent by shifting away from smart shopping campaigns to our targeted methods.

Why would our targeted method outperform Google’s automated system?

Negative Keywords

• A key aspect to driving profitable returns through Google Shopping ads are the use of negative keywords to exclude unprofitable, irrelevant/off-target, and MAP-restricted search terms from triggering ads.  Smart shopping campaigns do not currently allow you to place negative keywords on the campaigns which poses some issues for many of the clients we serve that sell in a specific niche market or have MAP restrictions.

Lack of Customization for Non-Shopping Ads

• The brands we serve have put a significant amount of work into building their brand image and conveying that to their customer base.  The lack of customization that is available for the ads these campaigns trigger on the display network and other non-shopping ad formats poses a problem for any brand that wants to convey a specific message and user experience through display remarketing ads.

Lack of Transparency (Search Term Optimization)

• The amount of data reported by these campaigns poses a major issue for retail brands utilizing Google Shopping ads as part of their digital strategy.  Google Shopping campaigns are an excellent platform for acquiring new customers. A search terms report from shopping campaigns is a treasure trove for discovering what a brand’s customers are searching for when converting on their site.  This data can be used to further optimize both shopping and non-shopping strategy and to optimize SEO on ecommerce sites. Unfortunately, this data is NOT available through smart shopping campaigns, a real disadvantage for digital marketers and retail companies.

Lack of contextual data

• The most fundamental need for Artificial Intelligence and bidding automation technology to drive success is big data.  Most shopping queries happen just a few times, that’s not enough data for machines to learn from. So the vast majority of an account receives an uninformed bid.  In the absence of data/proof, human judgement is THE key to making good decisions.

Control and user experience  

• The smart shopping campaigns are largely a black box as search term data is concealed.   Additionally, remarketing ads are mixed in with shopping ads, and there’s no way for the advertiser to know how many of their sales are actually due to remarketing vs. shopping ads investment.  So if smart shopping campaigns are actually shopping + remarketing, how do advertisers compare their performance against what they did before? Once remarketing is integrated into shopping campaigns, advertisers have literally zero control over remarketing creative, messaging, audience list control, or serving frequency.  There are countless ways that could work against advertiser best interest. A smart shopping campaign that might be thought of as an upper funnel customer acquisition channel could actually be driving 90% of the sales from the very bottom of the funnel via remarketing ads. We’ll never know because the data is now concealed – for some reason.

What does this mean for retail companies?

Ultimately, the day-to-day processes around driving qualified traffic to e-commerce sites with Google Ads will not change too drastically as many of the announcements were updates to existing ad formats.  The addition of Discovery campaigns and gallery ads is exciting because it grows the advertising opportunity. But it will be essential to test these new formats and determine how to leverage them to drive incremental sales and brand awareness, and not just additional ad spend.  It will also be important to test the impact of Shopping Actions as the functionality becomes more tightly integrated into Google Shopping results.

Recommendations for Ecommerce Advertisers

We recommend taking a cautious approach when extending the reach of shopping campaigns to other Google properties.  To date, we find better overall results when optimizing ads to each platform individually to maximize the advantages of each one.

Advertisers should be wary of boastful stories of smart shopping automation being superior in all cases, and ask for proof, especially if you are a major advertiser spending >$10k/month on shopping campaigns.  Ultimately the technology was initially built for smaller budget advertisers with limited bandwidth for managing their advertising campaigns. Before pausing your targeted shopping campaigns to test smart shopping campaigns, consider what visibility and progress might be lost when you pause your campaigns.  Often re-enabling them after a test does not yield the same result that was available before pausing the ads.

At the end of the day, computer-aided automation and machine learning algorithms have advanced significantly in recent years.  However, a retail company investing a significant amount of money into Google Ads campaigns is best served by focusing on creating digital advertising strategies that closely align with their company’s unique attributes and goals rather than relying on computers to drive the strategy for them.  In our experience, the difference is still very substantial.