Automatic Content Recognition (ACR): Whys and Hows

Automatic Content Recognition technology is entering widespread use throughout the Connected TV (CTV) industry. Its basic purpose is to track the video content that users watch on their smart TV or other connected devices, and empower precise targeting based on that data. It’s effectiveness is undoubtable – advertisers are adopting their approaches thanks to the advances allowed by ACR on both linear and CTV. Marketers, meanwhile, see that ACR can be an effective solution for boosting both in-store traffic and online sales.

As a result, media buyers need to know the whys and hows of ACR to make the most of it, while understanding its limitations. In this article, we will explain everything you need to know about ACR: what is the meaning of ACR, where it applies, who supplies it, and how it can be used in the most effective way.

What is Automatic Content Recognition (ACR)?

The ACR abbreviation stands for Automatic Content Recognition. It is a technology embedded into CTV environments or into Over-the-Top (OTT) video streaming applications. It refers to the ability to recognize and identify streaming content, down to individual objects on a video, by sampling its parts and matching those with database records.

ACR-supported devices (typically smartphones and smart TVs) help marketers to analyze content consumption behavior by comparing digital ‘fingerprints’ that are embedded into content to generate viewership insights. For instance, ACR technology allows seeing whether or not a particular ad was played on a specific device in a specific time frame. This information is crucial for publishers to figure out ad rates. These digital benchmarks help in calculating ad impressions, defining ad reach, as well as managing ad frequency more properly.

Other data points (information gathered through ACR) include:

• type of platform – we can define whether an ad was played using linear, CTV, MVPD (Multichannel Video Programming Distributor), or some other type of VOD-enabled (video on demand) device;

• location data for both stationary and mobile screens;

• demographics and IP addresses;

• viewing behavior – including content preferences, ad completion rates, average watch time, channel surfing patterns, completed view rates, and other parameters.

How does Automated Content Recognition Work?

What does ACR stand for? As we already mentioned, ACR samples parts of video (or audio) content and compares that data with references stored in the publisher’s database to recognize the content being played. This is possible due to the advent of fingerprinting. There are two types of ACR fingerprinting: one to process acoustic (ACR audio) media, and one for video content (ACR Video).

At the initial stage of the fingerprinting sequence, the publisher’s media goes through special algorithms that generate library-side fingerprints. These traces are stored in a special database connected to content servers. When a user browses his/her media through ACR-enabled devices, these generate their own fingerprints and send them to a server that matches both library and device content pieces. Based on matches, marketers get their metrics and insights to deliver highly targeted ads.

Now let’s see the precise difference between the two types of ACR.

• Acoustic Fingerprinting.

   

  Acoustic fingerprinting is a digital audio signal-based technology that mainly serves to identify songs and tunes. It is widely used in radio broadcasts, streaming services, copyright licensing, and even for video files identification. As we are primarily concerned about the latter, we’ll focus on that.

Acoustic fingerprinting reads frequencies, sound spectrums, amplitudes, tempo, and some other characteristics to identify the signature of a signal. Then, signatures go through compression and hashing to completely mask off the rest of the information. The signals themselves are usually made inaudible to humans.

• Video Fingerprinting.

   

  Just like acoustic fingerprinting, its video counterpart analyzes small clips sliced off the original video and compares them to the pieces available from the premade database. At the same time, video compression, resizing, or quality settings do not alter or change the embedded fingerprints.

Similar to Audio Fingerprinting, in Video Fingerprinting, small video clips are made from the original video, and certain characteristics are extracted from it. These techniques take care to ensure that image manipulation technologies like compression or resizing do not affect these fingerprints, and the content can be recognized nonetheless.

What Benefits does ACR Bring?

Brands utilize ACR TV for multiple reasons. The most obvious are frequency optimization, unique reach abilities, and improved targeting. Frequency optimization is important for ad efficiency since marketers can’t always tell if their ads are getting optimal display density to reach the needed effect. Reach control helps to ignore households or users that were previously targeted by your ads to augment unique customer reach (or vice versa). Improved targeting refers to the ability to programmatically target specific audiences that consume specific types of media.

Disclaimer: be advised that some of the US and EU laws limit ACR usage, especially the functions related to personal identification identificators tracking. Please consult with legal advisors before applying the technology in your work.

The technology is also valuable thanks to its data collection capabilities. Behavioral information tied to IP addresses and/or device MAC (Media Access Control) addresses greatly enhance household targeting. Along with this, ACR data allows us to learn actionable specifics about audiences to better understand how to reach them with ACR marketing campaigns. Through ACR, marketers are able to determine which households were exposed to specific types of content, and therefore advertising messages, on both linear and connected TV.

For brands that are considering automatic content recognition for their digital marketing campaigns, we suggest following one of two strategies: Linear-first or CTV-first. The Linear-first approach is best suited for companies striving to improve incremental unique reach and optimize frequency in the first place. The CTV-oriented approach focuses more on maximizing the spend on a target audience and driving a full-funnel sales strategy.

How Exactly does ACR Data Help Marketing?

Reaching unique audiences is one of the main advantages that automatic content recognition brings to the table. With the advent of CTV, more and more people are opting out of cable television, representing a good chunk of viewers. This cord-cutter audience is normally unavailable for TV advertisers. However, an ACR-based advertising strategy unlocks these otherwise inaccessible demographics to reach full marketing potential. ACR tech is also beneficial due to its ease of use. It can be implemented at any stage of buying, be it planning, deployment, or attribution.

Since digital and TV marketing campaigns speak the same language, linear TV media buyers can use the metrics they’re used to (such as incremental reach and GRPs) to conveniently switch over to this new environment. Additionally, marketers can evaluate ACR’s input to know exactly how much reach and brand awareness digital adds to existing advertising efforts. For the buyers that utilize both linear and connected types of campaign, ACR builds up solidity by making sure that each investment was cost-effective while reaching new audiences and managing ad frequency.

For instance, big CTV device producers like Roku and Samsung TV deliver ads to audiences that were never exposed to linear ads, boosting their advertisers’ reach. This increase in reach can easily compensate for the viewer outflow and viewership decrease seen on linear TV. Platforms like Samba TV are utilizing ACR data to retarget mobile audiences based on IP to show them the same ads on their smart TVs and smartphones. This approach is called dynamic ad insertion. Dynamic ads are the tools that can make linear TV ads addressable and accurately measurable.

Adopting planning tools that take into account ACR data helps marketers strengthen their incremental reach. Thus, by using linear TV data and core audiences, marketers can reach incremental households that are using a smart TV or other connected devices. This audience is otherwise missed when only focused on linear. ACR-powered planning will help avoid marketing stalemate by providing insights to refine digital tactics.

ACR data can also measure campaigns both on linear and on CTV. Through looking at who saw an ad, marketers can perform closed-loop attribution. These campaign measurements can also display which households were encouraged to take an offline action, such as a physical store purchase. To make that happen, the IP entries are compared with first-party or third-party conversion data.

What are the Downsides of ACR?

Naturally, ACR brings tangible benefits to those marketers who utilize it properly and carries real advantages compared to classic data sources. However, this technological approach is not without its flaws. Since smart TVs are currently not as widespread as, say, smartphones, automated content recognition analytics cover only a comparably small share of total US households.

Another limitation comes from the lack of significant personal data to enable precise enough targeting for niche advertisers. It follows that ACR data attribution is more suited for mass products and big brands, rather than startups and market newbies.

Since ACR-related analytics builds on pre-stored databases, this creates an inconvenience for marketing researchers. Say, if we want to dig a few years back to see how a particular campaign performed, we might encounter a dead end. Since data storage isn’t free, the costs associated with that storage rise significantly when it comes to huge video databases. That means that the data needed by the researcher might have been deleted, lowering the accuracy of the analytics.

Is ACR Technology Compatible with DSPs?

The short answer is yes. ACR tech providers do work with DSPs (demand-side platforms), but only the platforms that utilize DMPs (Data Management Platform). Put another way, ACR data is available to demand-side platforms through third-party extensions. Automatic content recognition technology can be implemented in a DSP as broad audience segments, or by cooperating directly with vendors that create custom segments.

These custom DSP segments have multiple benefits for publishers. First, since they are pre-built, publishers can ignore the need to create databases themselves, a task that can take considerable time and resources. Second, they allow more precise and quick access to information gathered through direct IP matching. And last but not the least, custom ACR segments enable strategies like lookalike modeling, used to uncover high-value user clusters, based on information gathered from highly converting households.

When DPSs partner with automatic content recognition technology providers, advertisers receive a unique opportunity to intelligently target the most valuable audiences without investing anything extra. Good cases for this might be existing linear TV campaigns that utilize a data-driven game plan. If we take a typical family as an example, we realize that targeting individuals in that household might be problematic since the living room TV is likely a device that is used collectively. But advertisers can use the whole family as a targeting unit, displaying the same ad multiple times in an attempt to reach each household member without causing ad fatigue.

Takeaways

Automatic content recognition can add significant value to marketing efforts because the technology can reach unique connected TV cord-never and cord-cutter audiences while using real-time optimization and reporting tools. Digital media buys can be enhanced via ‘pairing’ the ACR data with other first-party and third-party sources such as offline insights, purchasing behavior, demographic research, and the results from previous linear/non-linear campaigns. All of these factors can greatly improve digital buys in general.

ACR stands for improved data, transparency, targeting, and addressability. Obviously, advertisers and other marketers stand to reap huge benefits from these values. From the consumer point of view, ad relevance, content suggestions, and online marketing experiences will be significantly upgraded as well (if utilized right). Content providers, broadcasters, publishers – everybody involved in the process can get their fair share of the ACR pie. And if legislative precautions or technological limitations will not worsen, the technology will take root and thrive in the CTV industry.