What is a look-a-like audience?
It’s a way to reach new people who are just like your best existing customers.
Look-a-like audiences – if done right – can dramatically expand your prospects, growing leads and new customers like never before.
Let’s look at 7 of these look-a-like audiences:

  • Direct mail
  • Facebook
  • Google
  • AdRoll Prospecting
  • Bing
  • Pre-Roll
  • Private Companies


Let’s look at how each of these differ … and how you can supercharge your look-a-like quality and response by a little known strategy we developed for our clients.
#1 Transactional Data Is the Best Way to Build a Look-A-Like Audience
These are purchases that can create powerful audiences which are clones of your own current customers.
It’s so advanced, that today I can clone a 90-95% look-a-like audience – NSA style.
It uses the metadata of thousands of transactional data on everyone to find your right prospect.
This is the best look-a-like tactic, because it’s based on purchases, not searches, likes or viewing habits.
It’s extensive and powerful data on every individual’s purchase — stocks, nutritional supplements, clothing, wine – any purchase.
Direct mail names (and in other cases, email) are generated from the transactional data.
It focuses on metadata based on transactions, so it’s prospects who buy the same kind of product as your customers. It’s an actual purchaser who looks like your purchaser.
For example, a big data company like Oracle houses thousands of data tags on prospects.
For my clients, I work with their behavioral scientists and engineers to match them to a precise product or service profile, then create the look-a-like audience.
It enables us to get a much higher response, doubling it in some cases, over traditional efforts.
It gets you that much closer to identifying your “perfect prospect.”
For direct mail, this modeling – if done right – is revolutionary and will supercharge your response … and dramatically increase your universe, in some cases, tripling or quadrupling profitable prospect names.
And, if you have a customer list of 5,000 or more names, the “closing” is miraculous.

 
#2 Facebook Look-A-like Audiences
Another great way to build a lookalike audience is by tapping into Facebook’s algorithms.
Facebook keeps perfecting its ability to duplicate clones of your customers. And while the results aren’t as good as transactional data, the results come pretty close.
Using these algorithms, we can identify the common qualities of your prospects and find similar, or look-a-like, audiences.
Facebook locates new prospects based on similar Facebook profiles and online behavior. What do they click on? We match them up.
This includes analyzing data such as page likes, demographics, interests, website visits and more.
Our very best results with Facebook are when we take a direct mail look-a-like audience model built on transactional data of a customer file, and add it to finding our audience in Facebook, allowing them to do a much better job in searching for your ideal client.
And, of course, like with transactional data, if you have a customer list of 4,000 or more, your look-a-like becomes so close.
You can choose the size of a lookalike audience during the creation process.
Smaller audiences more closely match your source audience.
Larger audiences increase your potential reach, but reduce the level of similarity between lookalike and source audiences.
A source audience of between 4,000 and 50,000 works best.
And if your source audience is made up of your best customers rather than all your customers, that could lead to improved results.

 
#3 Google Similar Audiences
Google uses data collected in the Google Display Network to prospect. Data includes demographics, searches, video views, website visits, application downloads, and more.
Google uses a type of artificial intelligence to analyze trillions of searches and activity across millions of websites to help figure out when people are close to buying.
While a long way from the quality of the transactional data or even Facebook, it’s a vital part of any marketing plan.
Again, using transactional data and/or your customer files dramatically improves results.
#4 AdRoll Prospecting Look-A-like
AdRoll finds audiences using the IntentMap, the largest proprietary data co-op that advertisers can access by contributing their site data.
About 5,000 advertisers of all sizes have opted into IntentMap, pooling more than 1.2 billion digital profiles from across the web and mobile sources.
Such diversity allows AdRoll Prospecting to perform for all verticals and find you new customers.
Want to make this work better? Use transactional data and/or customer data.
#5 Bing Look-A-Like
Bing Ads now has a look-a-like feature.
In-market is a system that lets marketers target consumers who appear to be on the verge of making purchases. Grabbing someone’s attention when they’re ready to purchase is huge.
Bing’s in-market audience focuses on 14 different audiences: 4 are dedicated to finance, three on travel, two on cars. Others include clothing, hobbies, leisure, toys and games.
And more in-market audience targets are on the way.
#6 Pre-Roll Look-A-Like
I love integrating video into a campaign. Having created more than 300 TV commercials and videos, pre-rolls are a big breakthrough.
A pre-roll commercial, for example, is when you run an ad before seeing a video on YouTube.
But here is the good news. Take your transactional data and/or customer list and clone a YouTube audience. It’s powerful.
#7 Private Companies Look-A-Like
Many other private companies are using different ways to create a look-alike.
For example, I’m successfully using a database actually built on credit data.
And another I’m testing now integrates transactional data and web searches using transactional data based on a co-op of websites.
Should you use just one company?
No!
I use all 7 and my clients love the results.
For each company and each of the systems, I use a variety of slightly differing models.
The results:

  • Better ROI.
  • Better enlargement of the universe.
  • Better profits.

For these reasons, I use them all. I have one campaign now this divided 40/20/10.
One of the keys to success beyond proper testing is knowing how to create the models.
I’ve learned you should never rely on the modeling company. They are statisticians.
They are not marketers who understand the peculiarities of each target market.
That’s why I have my trained staff to modify and adjust each model based on solid direct response psychographic market principles and knowledge.
If you’re underperforming or not using any of the above, contact Caleb Huey to discuss a test program for your campaign.
Write him at caleb@cdmginc.com or call me at 310-212-5727.

Share with Your Friends