Advertisers will tell you that they don’t like to waste money. This is especially true heading into the holiday season.
However, that is exactly what many advertisers do by not taking advantage of Look-a-like technology, which has made great advances in recent years.
Look-a-likes can reduce waste, save time, and improve performance by defining highly effective targeting parameters before the first impression is even served.
What Are Look-a-Likes?
It is a method of profiling the behavior and actions of existing site visitors and customers (sites they visit, searches they perform, subject matter they read) to find common attributes that can be used to develop new audiences to target with your display campaigns.
The concept is simple. Serve display ads to people who look and behave like your existing customer base and they have a high tendency to convert.
Of course, exploration and optimizing is required to tune in targeting parameters, for new creative, landing pages, products, and seasons. But the time and money spent on the initial exploratory phase of a campaign can be greatly reduced by jump-starting a campaign with look-a-likes.
While advertisers often know lots about their target customers, a look-a-like analysis can tell precisely how to target prospects like those target customers with tactics such as search retargeting, domain targeting, and contextual targeting.
For instance, it is not difficult to determine the last search a prospect performs before coming to your site. However, with a sound look-a-like analysis you may find distinct patterns among your client base in the last 100 searches they performed prior to coming to your website.
How It Works
In the days leading up to the launch of a campaign, the advertiser places a pixel on their website. Once enough pixels have been fired, a company like Simpli.fi performs a statistical analysis to determine how to target users similar to those that have already converted.
Here are some examples of highly effective look-a-like types we see clients commonly use:
- Search Retargeting Look-a-likes (A.K.A. Search-a-likes). In this case, an analysis is done to determine the search patterns of existing converters while they are visiting other sites around the web. Search terms that are common to likely converters can form the core of a keyword list for search retargeting in order to target new customers. For example, see the graphic below: User A shows the behavior of a user who has already converted. Using data gathered from user A’s behavior on the web, users B and C can be targeted with display ads based on them exhibiting similar search behavior to user A.
- Site Look-a-likes. Targeting likely converters based on the sites visited by existing converters can also be highly effective. In this case, statistical analysis provides a list of sites that existing converters are more likely to visit than typical users who have not converted on the advertiser’s site. These domains are then be targeted with a campaign that serves only to those specific domains.
- Contextual Look-a-likes. In many cases, converters show unique patterns of visiting pages about particular topics, or that contain particular words. Once the contextual categories and/or keywords that are favored by existing converters are identified, campaigns that target similar pages are run and optimized.
- CRM/Offline Look-a-likes. Look-a-likes can even work with offline data without exposing any user PII (Personally Identifiable Information). To do this, offline converters are cookie matched with online browsers, and search, site, and/or contextual look-a-like analyses are performed. Armed with insights on how their offline converters behave online, advertisers can then launch campaigns targeting users with search, site, and contextual patterns similar to their existing offline converters.
In addition to reducing waste at the start of a campaign, look-a-likes can deliver a host of benefits:
- Expanded Delivery. As more data is collected during a campaign, additional search terms, sites, and contexts are often identified that enable advertisers to target more users who have a high probability of converting.
- Improve KPIs (Key Performance Indicators). Whether the advertiser’s key performance indicator is CPA, CTR, CPC, ROAS or something else, campaign optimization using look-a-likes drive improved performance.
- Deliver Insights. The insights attained by doing look-a-like analysis help marketers implement more effective search campaigns, and also create marketing programs with specific creative, landing pages, and other elements that appeal to newly discovered target audiences.
Any advertiser looking to reduce waste and improve performance during this holiday season should be using look-a-likes. In fact, any website experiencing increased holiday traffic should capitalize on this opportunity to gather as many insights as possible regarding their existing audience so that they can improve their targeting efforts in 2013 and beyond to acquire new audience members.
Think of it as creating clones of your ideal customers based on the most favorable behaviors and attributes of those customers to your product/service/brand.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.