The Future of Targeted Search Advertising

Marketing Transformation

Today, search marketers generate user insight through what the user is looking for, via keywords. But keywords don’t give context into who a user is and why he or she is searching. Historically, search advertisers have relied on site extensions and ad copy strategies to help prequalify or segment users, a strategy that still results in a one-size-fits-all search experience. Now, the ability to funnel SEM programs through smaller, more targeted segments is within any marketer’s reach.  

In this piece from Razorfish, we highlight: 

  • How Remarketing Lists Accelerate the Buy Cycle
  • Why Customer Match Offers a Game Change
  • Adjusting the Experience With Demographics for Search Ads (DFSA)
  • What’s Next?

How Remarketing Lists Accelerate the Buy Cycle

A target-focused approach to digital advertising is nothing new. Behavioral remarketing and first-/third-party, audience-based display tactics are staples of many well-optimized media plans. Yet, while many features and extensions have been released throughout the years, the core functionality of an SEM program has remained largely the same since the early 2000s: advertisers bid on “keywords” in an auction-based system, competing for a user’s click.

Then in 2012, Google started testing Remarketing Lists for Search Ads (RLSAs) to recognize how far the buyer had already journeyed. This cookie-based approach allowed advertisers to adjust ad copy, bids and keywords based on users’ previous site interactions. For the first time, marketers could accelerate the buy cycle by aggressively targeting (or excluding) those audience segments with a history of past website interactions. Common use cases included:

  1. Bidding more aggressively for users who had already begun a checkout or form process but had not yet converted (e.g., cart abandoners)
  2. Isolating bidding to previous site visitors or converters (thereby increasing efficiency for high-cost or competitive terms)
  3. Upselling or using sequential messaging for users that had already converted

While RLSAs were a welcome and effective addition to the SEM landscape, few media and social marketers have evolved their targeted capabilities beyond simple behavior-based strategies. While Google focused on previous site behavior, Facebook was rolling out Custom Audiences, a tactic that allowed advertisers to upload and target users based on CRM data (e.g., phone numbers, email addresses, etc.)

How Customer Match Offers a Game Change

The idea of connecting media with CRM data allowed for even deeper personalized marketing. It was only a matter of time before Google finally announced its version of Customer Match in 2015.

Customer Match lets advertisers upload customer email addresses into lists that can be segmented based on any number of attributes. While functionally similar to RLSAs — bids, ad copy, keywords, etc., could all be adjusted – advertisers could now expand their personalized campaigns to users who had never been to the website, reset cookies, etc. At our recommendation, Razorfish clients were quick to take advantage of this tactic, albeit with increased scrutiny and sign-off to carefully address any privacy concerns. Common use cases for Customer Match include:

  1. Bidding aggressively for high-value customer segments, such as frequent converters' or users' high average order sizes
  2. Excluding existing customers for new-customer acquisition programs
  3. Bringing offline sales online (bidding aggressively for users who have converted through other offline channels, such as in store or via phone)


Adjusting the Experience with Demographics for Search Ads (DFSA)

In 2015, Google also began testing the integration of audience demographic data and search. With Demographics for Search Ads (DFSA), advertisers could view behavior and adjust the experience based on gender, age and/or parental status. This tactic is often layered on top of, or alongside, RLSA and Customer Match strategies, providing marketers valuable user and customer insights, such as understanding the demographic mix of a keyword or ad group. It also can greatly improve the efficiency of even a well-oiled campaign. If the conversion rate for a given demographic segment is low or high, advertisers can adjust bids accordingly to meet or exceed performance goals and targets.


What’s Next?

For Razorfish clients, finding a balance of scale and impact has been key to targeted search strategies. It can be tempting to create hundreds of target audiences based on unique behaviors or attributes (e.g., visitors to a specific product page), however, we find that creating too many targets can result in having insufficient data. Alternatively, Razorfish recommends starting with broad/high-volume groups before narrowing to finer segments as additional data is acquired.

Razorfish constantly tests new features to make our clients’ campaigns more effective. Search engines continue to release additional audience options, including:

  • Cross-Device RLSA Targeting
  • Similar Audiences (“lookalike advertising” for search)
  • Integration with Google Analytics (e.g., segmenting audiences based on page views or time on site)
  • Microsoft’s Universal Event Tracking (UET) remarketing (similar to Google RLSAs)
Targeted search evolution timeline


Today, personalization is a must-have, win-win capability for buyers and sellers. Advertisers want to deliver the right message at the right time, and users prefer ads that are personalized to their interests and shopping habits. 

With Customer Match, RLSAs and Demographics for Search Ads, advertisers can now enable this personalization within search marketing campaigns. The search industry is slowly moving to a “who, not what” approach, in which the value of a keyword + bid alone is no longer enough.

Now, the combination of knowing “who” a user is (Customer Match, RLSA, Demographics) with other personalization layers, such as day, time, device and location, has resulted in extremely effective and personalized search marketing programs.


This piece was authored by Matt Wilkinson, with contributions from Kristin Ambrookian, Jeanne Fu, Amos Ductan, Courtney Demko, Dane Manning and Adam Gale.


Research Sources

More Insights


49 Cities.
5 Continents.
1 Vision.