Get the Most Out of Your Website Custom Audiences

Pablo Martinez Sep 12 2018 1 PM | 6 min read

Facebook’s Website Custom Audiences (WCA) are extremely flexible, but choosing the right audience composition for the right goal can be tricky. As we offer several advanced options for structuring audiences that can help you reach better, highly relevant people with ease, we decided to list tips and tricks on how you can get the most out of your Website Custom Audiences.

Website Custom Audiences are a powerful tool for all Facebook advertisers. They offer flexible rules for targeting people who, for example, visit particular parts of your website or trigger certain pixel events. Most advertisers are already familiar with standard strategies such as excluding people who recently made a purchase, targeting shopping cart abandoners, using lookalike audiences of people who have made purchases, and using value-based lookalikes created from a customer list. However, a lot more can be achieved by leveraging Website Custom Audiences’ advanced filters.

We previously wrote about when you should and shouldn’t split targeting on Facebook. One of the key reasons for considering splitting is if audiences have different conversion values – i.e. that you should be more aggressive with the more valuable audience subsegments. But how do you identify these valuable users – or conversely, identify people you don’t want to target? Especially when it comes to prospecting it can be hard to figure out the people with the most potential. Lookalikes are a typical solution for this, but the quality of the seed audience (the people you use to build the list from) can play a huge role on how they perform. This is where more advanced setups can be extremely helpful.


Extended Possibilities with Advanced Filtering

Advanced filtering makes the already extremely flexible Website Custom Audiences even more customizable. There are a considerable amount of options but here are some that can be particularly useful.

  • Aggregation

If you create a regular WCA of all people who triggered a Purchase event with a value of more than $100, all individual pixel events of the person are considered separately. For example: if Ben buys four pairs of shoes worth $60 each, but on four separate occasions, he would be ignored by this standard rule even though he bought $240 worth of items!

With aggregation rules, you don’t need to rely on individual events, but instead, you combine the data from all the times the user has reported the rule in the given timeframe. Using these aggregate sums, you can capture users who might have lower individual purchase amounts, but are very valuable to your business overall. Aggregation also supports using averages (e.g. users whose average purchase value across purchases is over $50), counts (e.g. users who have triggered more than 10 content views), minimums, maximums, time spent on website, and another extremely useful rule: ranges.

  • Ranges

With, you can also use ranges with aggregation, so you can select a certain percentage of users triggering a particular pixel event (or set of events). This supports counts, sum of value and averages. With range rules, you can for example target the top 10% of users based on the amount of purchases, the highest total value of purchases, or highest average purchase value. This opens up fantastic possibilities for creating high-quality audiences for retargeting and seed audiences for lookalike creation.

  • Custom Parameters

Custom parameters support any key-value -pair you can think of. This means you can look for any parameter and its associated value fired by the Facebook pixel. For example, if you send some custom user profiling in your pixel, e.g. user-bucket: ‘repeat-user’, you can create WCAs based on these values. Another typical case could be to categorize different sections of your website with custom values if you are advertising very different offerings (e.g. so visitors to your open jobs section get targeted with job ads instead of products).


Use Cases

As you see from just a few examples, the possibilities are wast. So how do you leverage these options in the best possible way? I highly recommend experimenting with the options and testing to see what kind of audiences work best for your business specifically. To get you started, here are a few practical use cases.

  • Create Lookalikes from Your Most Valuable Customers

If you are sending conversion value in your purchase event (or whichever event you use as your end conversion), range filtering can be used to easily extract the most valuable customers. Create a Website Custom Audience from people who have the highest total sum of purchase value using the 90-100% range (i.e. the top 10%). Once you have this audience, create lookalikes from it and you should reach potential customers that are most similar to your most valuable customers. This strategy can also work great for promoting mobile apps if you are tracking purchases in your mobile app; simply create the audiences mentioned but based on app events instead of pixel fires.

Note that this audience can be useful in itself as well, e.g. for promoting a loyalty program or re-engaging with customers who have been valuable historically but haven’t purchased anything in a while.

  • Reach New, Highly Engaged Users

For companies, such as media outlets, finding new users that are highly engaged after they land on the website can be difficult. However, this can be made easier using ranges. First, create a custom audience of people who triggered the most content views on your site in the past 30 days. You should aim for somewhere around the 90-100% or 80-100% range, depending on the overall monthly traffic volumes of your site. Then, create lookalikes based on this audience and you should start reaching more highly engaged people.

  • Exclude Users with Very Short Time on Page

Some of the traffic to your website will always be low quality and this can often be recognized from very short time spent on your site. If you are experiencing issues with low quality traffic, try creating an audience of people who have spent less than 5-10 seconds on the website and excluding them from your campaigns.

  • Reactivate Old Customers

Repeat customers can provide immense value to a business so reactivating people who bought from you in the past can be very effective. Managing audiences for this kind of targeting can be done quite easily. Create an audience of people who purchased from you in the past 90-180 days, but also apply an exclusion for people who bought from you in the last 15-30 days or so. I recommend looking at your analytics to determine the rate at which people make purchases to determine the optimal exclusion period. This ensures that you are not annoying people with ads too soon after they last bought from your, but ensure you hit people after they are looking for something else and are open to buying from you again. When targeting these customers, remember to demonstrate value using e.g. special offers.


How Do I Use These Options?

If you are already using, you can find all of these options simply by creating new app or website custom audiences. If you are interested in trying our solutions, send us a demo request to learn more.

   Retail Re-UP: Adjusting to the Next Normal of Social Advertising CHECK IT NOW
Pablo Martinez

Read Next