Audience Targeting

Sections in Chapter 3:

Audience Profiling

How to configure targeting parameters before the launch of a campaign to develop a relevant, high-quality audience pool.

Learn more

Campaign Optimization

How to configure campaign settings and controllers in real time to reach the ideal audience.

Learn more

Campaign Expansion

How to use deep targeting to create look-alike models and extend your audience pool.

Learn more

Audience Profiling

The History of Video Audience Targeting

Historically, marketers who wanted to reach targeted audiences with video advertisements had to make educated guesses about their target demographics. Even the best advertisers relied on only several imprecise factors (age, gender, income, location and interests) to target their audiences. Once marketers launched campaigns, they would invest huge amounts of time and money manually tracking campaign performance for months before they would be able to make guesses about what was working for their business and what was not.

Programmatic Audience Screening

Modern programmatic technology enables a remarkably precise targeting process that contrasts starkly with past targeting capabilities. Programmatic technology has the power to screen hundreds of millions of consumers across hundreds of attributes to define targeted audiences for specific business goals. Many programmatic platforms use machine learning and artificial intelligence to constantly refine and optimize advertising campaigns based on more and more data. Machine learning technology constantly search for new customers from profiles of current best customers. Programmatic marketing delivers the precise personalization of one-to-one marketing to millions of consumers in just seconds.

Important Targeting Parameters

How to Choose Appropriate Parameters

Different programmatic video advertising platforms offer different targeting options. Given a specific target audience profile, these platforms gather data from ad servers, bid requests, publishers, consumers and clients to create a data map of where to target that particular audience profile. The following targeting parameters should be used to create a detailed audience profile prior to the launch of a campaign to maximize success:

Keywords

Construct a list of all keywords relevant to your target audience. Keyword targeting should synthesize all relevant language surrounding a particular market into keyword format and weigh these keywords for relevance. Keywords target relevant users over a defined period of time. Keywords can be used to generate an estimated audience size and an analysis of historical audience consumer performance. A focused list of keywords guides advertising campaigns towards relevant, in-market consumers.

Categories

Identify the commercial categories with which members of your audience identify. Category targeting uses categorical information extracted from ad networks to determine the relevance of a particular publisher to a specific consumer. As an example, a channel advertising basketball shoes could define category of ‘basketball fans.’ Categorical data follows the nesting structure of existing data so that deeper nesting categories are prioritized for relevance and efficiency when multiple categories overlap.

Geography

Develop a rough idea of where your target audience is geographically located. Some target audiences are not geographically specific, while others can be highly localized. Geo-targeting is a targeting process that aggregates all available data about geographic location to target consumers in specific areas. Geo-targeting uses information about coordinates, zipcode, state and IP addresses to deliver advertisements only to specific geographic locations.

Demographics

Develop a basic demographic outline of your target audience. While not all campaigns will target specific demographic groups, some may target a specific age, gender or marital status. Depending on the targeting technology used, demographic information can be optimized manually or by machine learning. When launching a campaign, make sure to confirm how the targeting algorithm works to avoid instituting overly restrictive demographic limitations.

Devices

Identify which devices your target audience will be most active on. If advertisers determine that their audience members will likely be active on many different devices (or are not sure which devices they will be active on), they may decide to run campaigns across all devices. If advertisers are confident that members of their target audience will be disproportionately active on certain device types (e.g. young adults on mobile devices or in social media platforms), then it may be advantageous to target the specific device type.

Operating Systems

Identify which operating systems your target audience will be most active in. If advertisers determine that their audience members will likely be using several operating systems, they may decide to run campaigns across all operating systems and formats. If advertisers are confident that members of their target audience will be disproportionately active in certain operating systems, then it may be advantageous to target these specific operating systems and develop content for the formats available.

Look Alike Segments

If you use other marketing channels, identify your highest performing consumer segments. Look-alike targeting uses these existing segments (usually the top 20% of purchasers or consumers with active shopping carts) and uses data from these profiles to generate a new pool of consumers with similar attributes. Before performing look-alike analyses, specifically define what ideal audience members look like and determine whether overall campaigns will benefit from a new look-alike audience.

Relevant Audiences

Analyze your competitors to identify similar relevant audience pools. Relevant audience targeting uses data from competitors, content management systems (CMSs), forums, referrals and social groups to target audiences with specific interests. The more information that is known about competitors, the more precise the relevant audience pool will be.

Campaign Optimization

The Difference Between Controllers/Settings and Targeting Parameters

Historically, marketers who wanted to reach targeted audiences with video advertisements had to make educated guesses about their target demographics. Even the best advertisers relied on only several imprecise factors (age, gender, income, location and interests) to target their audiences. Once marketers launched campaigns, they would invest huge amounts of time and money manually tracking campaign performance for months before they would be able to make guesses about what was working for their business and what was not.

Important Settings and Controllers

How to Choose Appropriate Settings and Controllers

Different programmatic video advertising platforms may offer different available sources of inventory. Given available advertising inventory and the predetermined target audience profile, these platforms will optimize an advertising campaign in real-time to fit its goals. The following specifications should be outlined prior to the launch of a campaign to maximize campaign effectiveness:

Ad Positions

Determines where you want your advertisement to be served on a publisher’s display page. Ad inventory positions are classified as either above the fold or below the fold (a reference to the fold in a printed newspaper). In the digital world, ads above the fold are immediately visible while ads below the fold are only visiblewhen a viewer scrolls. Ads located above the fold receive higher view rates and are more expensive. Video players are usually above the fold and video players located below the fold are more susceptible to bots and server errors.

Ad Sizes

Determine how large you want your ads to appear prior to campaign launch. The ad size controller controls what ad sizes will be served. For display ads, there are more than 20 different commonly available ad size templates. For video ads, the default video ad size is 300×250 pixels. Different ad serving templates serve videos in different-sized publisher video players that range from small (300×250) to medium (default Youtube player size) to interstitial (full screen).

Daily Limit

Set a maximum daily budget. The daily limit controller sets the daily amount of budget spent. Even for budgets that are manually set on a month-to-month basis, this controller follows an automated performance analysis model within the bid system that adjusts for seasonality and shifting user behavior to maximize the return on ad spend.

Day Parting

Decide if there are certain times of the day that you want to focus on. The day parting controller allows for precise control of how the daily budget is allocate over the course of the day. This controller allows for manual control of budgetary spend by the hour and provides automated minute-to-minute optimization based on the performance and pacing model in the bidder system.

Frequency Cap

Determine the maximum number of times you would want a consumer to see your ad in a single day. The frequency cap controller determines the maximum number of times a particular consumer can see the same ad in a single day. This controller acts as a guideline and not an absolute cap and it can be offset by as much as 50 percent on an automated basis depending on the performance and pacing model in the bidder system.

Retargeting Window

Decide how many days to wait before retargeting a consumer that has already seen your ad. The retargeting window controller determines the length of the qualification window for retargeting a consumer. A 10 day retargeting controller means that consumers who have visited a website in the last 10 days are eligible to be retargeted. The more narrow the window is, the smaller the the consumer audience pool will be. This controller effectively determines the size of the audience and therefore has a far greater impact on performance than other controllers.

New User Window

Decide how long a consumer should be inactive before you qualify them as a new customer. The new user window controller determines the length of the qualification window for designating a consumer as a new user. A 60 day new user controller means that consumers who have not visited a website for more than 60 days will be considered new users when they arrive at a client’s website. The more narrow the window is, the larger the consumer audience pool will be. This controller effectively determines the size of the audience and therefore has a far greater impact on performance than other controllers.

Block and Whitelist Specific Domains

Decide if you want to prevent your ads from being served on certain pages and if you definitely want your ads to be served on other pages. If you are confident that members of your target audience regularly visit a certain page, you can whitelist that page to ensure that you serve ads on that page. The block/whitelist controller allows you to manually block or manually target certain publishers’ ad inventory.

Controller Weighting

A controller weighting system is a digitized priority ranking system that optimizes overall campaign performance and pacing. Without a weighted system, advertisers cannot attribute campaign successes or failures to any individual controller and must attribute value equally.

Geography

Decide if you only want your ads to be served in a certain geographic area. The geo controller aggregates all available data about geographic location and allows clients to serve ads to consumers in specific areas. This controller uses information about coordinates, zipcode, state and IP address.

Campaign Expansion with Deep Targeting

The History of Video Audience Targeting

Many digital advertising channels will begin to drive conversions no earlier than two weeks after campaign launch due to the natural lag time caused by the massive amount of data required to target large audiences. One way for advertisers to boost early campaign performance is to conduct a deep targeting analysis even before the peak conversion rate has started to ensure that the campaign will perform well. A deep targeting analysis is an adjustment of targeting parameters based on campaign performance and market conditions.

How Deep Targeting Works

A deep targeting analysis is a comprehensive reevaluation of optimal targeting parameters that is based upon shifting marketing conditions and early campaign performance. Deep targeting analysis creates an analytical model based upon converted consumer profiles to create a specific set of targeting parameters for top performing consumers.

Look Alike Models

A look-alike model is compilation of the attributes of high-converting consumers generated during a deep targeting analysis that is used to find other consumers across the internet who are highly likely to convert. Look-alike models are typically used to launch high-intent campaigns that serve ads with increased frequency to these high-priority consumers. High-intent campaigns can have varying parameters that determine the degree of similarity that a look-alike target must have to previous high-converting consumers, ranging from broad matching to highly specific matching.

Share This

Share This

Share this post with your friends!