🥳 You have completed all topics in the Handbook. Click here to claim your certificate!

3. Audiences and targeting

Ads need audiences. Unlike roadside ads and billboards, digital advertising can actually target ads to very specific audiences. This improves the relevance of the ads and, hopefully, generates business value.

When you are driving your car or walking down a street, the billboards and physical ad displays you see are the same for every one else on the street, too.

We’re not yet in the “Minority Report” world, where digital ads in physical locations can tailor their content by viewer.

However, digital advertising on the internet has been founded on the premise of audience segmentation and targeting.

The online world gives us unprecedented capabilities of choosing just the right pairs of eyes for any ad we create. We can choose the potential audiences with extreme granularity – ranging from interest-based cohorts all the way to creepy-ish demographic data such as gender and age.

With measurement tools, we can collect enormous amounts of data about our visitors and use that to fine-tune our audiences with real user data rather than with imaginary target profiles.

However, all these capabilities raise pressing concerns about privacy and misuse of data. Because audience segmentation relies on granularity of data, a “more is more” approach has been prevalent in the digital advertising industry. The longer you have been online, the more information these big advertising technology companies have on you, and they are very reluctant to give it up.

In technical marketing, audience generation, segmentation, and targeting are concepts you’ll frequently encounter when working on marketing campaigns. Understanding the limitations (both technological and regulatory) will help you find the appropriate balance in creating powerful ad audiences while protecting the data of end users and customers alike.

How audiences are generated

Audiences are simply segments of user data.

When you collect information about your site visitors, app users, and business customers, you are collecting data that can be used to segment the users.

Example

When someone visits your site or uses your app, you can ascertain the browser or device they were using, whether they’ve visited the site or used the app before, and you can observe if they visit certain key pages or perform meaningful actions during their session.

With this data, you could generate these user segments, for example:

  1. Users who use the Google Chrome browser, who are repeat visitors to the site, and who end up subscribing to the newsletter.
  2. Users who bounce immediately after visiting the site.
  3. Users who are first-time visitors to the site, and who view your blog content.

Depending on the measurement tools you use, you can then save these segment settings, after which you can always apply them again to look at your data through the lens of each audience.

When segmenting user data, you are looking for patterns of behavior unique to each segment. These patterns will help you determine if, for example, it might be useful to target certain ads only to a subset of the available segments to improve the ad’s chances of creating business value.

If you have access to more data sources, you can build even more comprehensive audiences, feed them to a data management platform (DMP), and improve the granularity of your audiences further.

For example, let’s take the first segment from the list above. We can add granularity to it by replacing “repeat visitors” with “buyers”, because we can compare the email address they used to subscribe to the newsletter with data in our customer database. We can then build three additional segments as subsets of the original one:

  1. Users who have made more than one purchase before (“loyal”).
  2. Users who have made a single purchase before (“customer”).
  3. Users who have not made any purchases before (“prospect”).

This is incredibly valuable information. For example, you might have a new customer benefit that should be advertised only to visitors who have not made a prior purchase. To improve the hit rate of this ad and to avoid spending ad budget on customers who cannot make use of the benefit, you can target the campaign to just the “prospect” segment above.

This is an incredibly handy feature of online advertising, as it allows us to do hyper-targeted advertising, keep our return on ad spend high, and avoid negative sentiment that comes from showing ads to users who don’t find them at all relevant.

Targeting ads to specific audiences

What types of audiences you can build and what types of targeting mechanisms are at your disposal depend on the advertising service you are working with.

Search advertising, for example, has the unique edge of being able to utilize the user’s search preferences to segment search engine users as ad audiences.

Example

You could build an ad that targets the keywords “in-person analytics course” for visitors from “London, United Kingdom“. When a visitor from London then searches for analytics education, your search ad would show up next to the search results.

Your ad would not (likely) show up for a search about “online analytics blog” or for visitors from some other country.

Similarly, on Facebook, you have an enormous amount of segmentation opportunities due to Facebook being able to target users based on how they interact with brands, what they chat about, and what types of content they focus on the most.

You could create a Facebook ad that targets “middle-aged men from Finland” who are interested in “heavy metal music”. This ad would then show up when people in that cohort use Facebook, and especially when they interact with content that is related to metal music.

While this type of advertising represents a huge chunk of what the digital advertising industry is, it’s important to remember that companies like Meta and Google are the exception and not the rule. There are only a handful of such big tech companies online that can simultaneously utilize a dataset of billions of users and provide advertising opportunities that target these users.

There is a long tail of advertising networks and publishers that are not part of these walled gardens. Their ability to utilize audience data and to serve relevant ads relies heavily on their relationship with the bigger ad networks and on technology that is slowly becoming more and more difficult to rely on.

Deep Dive

Walled gardens

When a brand has digital properties and services that people utilize on a daily basis, they have the unique opportunity of enabling in-service advertising that uses their own user segments and their own placements for the ads.

Example

Google and Bing have billions of users using their search services on a daily basis. They can utilize the data gathered within their search engines to show ads to the users within their search engines.

Walled gardens provide buy-side, sell-side, and auction mechanisms within their own services

This is often referred to as a “walled garden” approach. Companies like Google, Bing, Meta, and Amazon can build extremely efficient advertising services by acting simultaneously as the ad network, the DSP, the SSP, and the publisher. Advertisers target users that use the service, and their ads show up in placements within the service itself.

Ready for a quick break?

You are seeing this ad because we think it’s time you took a small break. Go get a beverage and walk around a bit to get your synapses back in order. We promise we won’t segment you further based on your choice of beverage (sparkling water, right?).

Cross-site audiences and retargeting

While contextual and in-service ads are powerful (at least for the big tech companies) in their own right, digital advertising’s most prominent and most controversial aspect is what programmatic advertising relies on the most: advertisers buying ads on publisher properties.

If you look at vacation flights on a travel aggregator, you’ll very likely run into ads with these flights prominently displayed as you browse different sites and use your apps.

How is this possible? How can a website completely unrelated to the flight booking engine still show ads about what you did on that flight aggregator?

You can thank technology known as cross-site tracking combined with the advertiser participating in retargeting (or remarketing) of their ads.

Cross-site tracking means that ad networks run little pieces of code on the sites that you visit, collecting information about which sites you visit and what you do on those sites under shared identifiers commonly stored in third-party cookies.

Then, a publisher can sell their ad placement to the ad network, and the ad network can display an ad in this placement based on the user’s cross-site browsing behavior.

To continue our example from above, the flight aggregator pays the ad network to retarget the flight the user just saw, so that the ad about that flight would show up on participating publisher sites.

In this case, the publisher has an ad placement available on a news website, and based on the user’s prior browsing behavior the ad network decides that this placement would be perfect for a reminder about the flight the user saw previously.

The publisher site and the travel aggregator don’t need to have anything to do with each other except that they both make use of the same advertising paradigm and participate in the same auctions.

Ethical concerns

The idea that web and app users are constantly being followed while they use these online services is problematic.

Your browsing history defines you. It tells ad networks intimate details about your preferences, your online behavior, and the likely user cohorts you and similar online profiles fall into.

While this data is a gold mine to advertisers and publishers alike, it’s also invasive and relies on an extremely unbalanced approach to user’s right to own their data.

You might think that it’s great to be reminded about abandoned carts and to get discounts on flights that you just viewed, but that’s just the tip of the iceberg when it comes to what cross-site data can be used for.

It is for this reason that web browsers have committed to disabling third-party cookie (and similar storage) support and why regulations like GDPR in the European Union seek to give data subjects agency over their own personal data.

Don’t miss this fact!

Segmenting and targeting users based on their previous behavior on the internet can be a very effective way of making sure the correct audience sees the correct ad. At the same time, it can be a very invasive way to engage a user, which is why technology and regulations race to prevent misuse of a user’s interaction history with any given brand.

It’s also why walled gardens have an additional unfair advantage: they don’t need cross-site audiences to profile their users. They run miniature versions of the web within their own walls, and they can create even more robust audiences using highly targeted and extremely reliable behavior data, as there are no technical limitations to how they can collect this.

As a technical marketer, it is likely that you will be lobbied by advertising technology vendors regarding different “solutions” to the lack of cross-site tracking support in modern browsers. While the decision how to proceed is of course up to the risk evaluation you’ll need to do in the companies you work with, it’s important to understand the technical premise of how these digital advertising networks operate.

This way you can make informed decisions on how to support your companies’ digital growth while still respecting the data protection rights of your customers.

Key takeaway #1: The right audience makes the ad experience relevant

With granular data collection, marketing technologies can collect a lot of information about users in order to associate them with interest- and demographics-based cohorts. The amount of detail that can be collected from an internet user in even a short period of time is staggering. This correlates directly with the effectiveness of the audiences, because the more you know about a user, the more accurate your segmentation-based decisions can be.

Key takeaway #2: Cross-site audiences are most effective and problematic

Even though audiences can be generated from the data collected from a single digital property, especially the programmatic advertising world has long relied on cross-site audience building. This means that the user’s browsing behavior – what sites they visit and what they do on these sites – is used to determine their interests. This allows for a far broader understanding of the user, but it comes at a huge privacy cost. Users don’t necessarily want to be followed across the webs. They don’t want advertising networks to know which sites they frequent.

Key takeaway #3: Collecting all that data is intrusive

Even though we can collect a wealth of information about internet users, it doesn’t mean that we should. Data protection laws and data minimization principles regulate whose data can be collected, how long it can be stored for, and for what purposes it can be utilized. Rampant data collection over the past decades has resulted in stricter privacy laws to avoid user data being harvested for audience generation without a proper legal basis to do so.

Quiz: Audiences and targeting

Ready to test what you've learned? Dive into the quiz below!

1. What technology has been historically crucial in enabling cross-site tracking?

2. Which of the following segments would you NOT be able to create with your website's data alone?

3. What are walled gardens?

Your score is

0%

What did you think about this topic?

Thanks for your feedback!

Unlock Premium Content

Simmer specializes in self-paced online courses for technical marketers. Take a look at our offering and enroll in one or more of our courses!

0/27 Topics completed

Good job!

Complete all topics to claim your certificate.

Next Chapter