What is Machine Learning (ML)?
“Machine learning is the scientific study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions, relying on patterns and inference instead.”
ML is a broad field with a number of approaches, but the two we’ll speak to in this article are supervised and unsupervised. Supervised learning algorithms build a mathematical model of a set of data that contains both the inputs and the desired outputs. In other words, we tell a machine what we want to happen and the machine gradually learns to generate that outcome. A common example of this is feeding a huge number of images of, say, broccoli to an algorithm, then sending random images and testing how often it identifies broccoli correctly. Unsupervised learning algorithms take a set of data that contains only inputs, and find structure in the data, like grouping or clustering of data points. This is useful for finding emergent behaviors that would be impossible to predict ahead of time.
To date, ML has been used to optimize ad delivery without the need for human oversight. It would be remarkably inefficient (and costly) for Facebook, Google, or any other digital advertising platform to serve ads purely based on who bid the most, particularly when advertisers aren’t getting the results they want and users hate the ads they see. ML allows those companies to rapidly optimize campaigns in the context of inventory demand, bidding, user response, and literally hundreds of other variables, which means more relevant ads for users and turnkey campaigns for advertisers.
What this doesn’t do is tell us why. Why did video A result in 10x sales relative to video B? What about the themes, humor, duration, content, even color, worked? What didn’t? And just as importantly, what worked for each individual viewer, and how do we improve our media moving forward?
If you are familiar with the programmatic ecosystem, you’ve probably encountered the terms like “machine learning” already. The development of this technology is expected to change the world of digital advertising dramatically. But how will it look like and what is the role of machine learning in online advertising?
Machine learning (ML) is a set of algorithms used to draw conclusions and predict results based on the input data. Some of these algorithms are even capable of improving their own performance once they get more data to learn from. In broader terms, machine learning is a subfield of artificial intelligence (AI) applied in many industries from particle physics to digital advertising.
What is machine learning in advertising?
In the industry of digital advertising, machine learning is usually used to let brands understand customers better and optimize ad campaigns accordingly. Same as for other industries, the main task of online advertising machine learning is to analyze data points and detect correlations that are not so obvious to the human brain. But even for advertising the range of opportunities is so broad that machine learning can be applied to a huge variety of problems from improving ad performance to the development of computer vision.
But there is already enough theory around the web, so it’s time to explore practical questions – what does machine learning have to do with your ad spend?
Video Content That Drive Results
Supervised ML tools provide the ability to analyze every frame of a video for content and themes, such as specific actors, objects, action, humor, locations, and emotions. When combined with performance and retention data, we can start to understand not only what content resonates, but what about that content resonates. This allows us to optimize assets and preemptively create videos that drive results.
Large tech companies that process millions of hours of video already analyze those videos frame by frame to determine the content and what’s resonating with users, but this isn’t a technology that’s in wide adoption for content creators. The reasons for this are numerous:
- It’s a relatively new technology
- Most off the shelf tech isn’t aimed at advertisers
- A robust solution requires many ML libraries and tools used in concert
- Creative is typically delivered to media teams as a one-way street, with optimization limited to serving different variants to different audiences
- Most advertisers don’t even realize it’s possible
TONIK+ is solving these problems by building and aggregating a number of ML tools to preemptively process video into individual scenes and label all of those scenes with relevant labels. Once this is done, we can then overlay performance data and provide brands with their first intra-video intelligence report, including:
- What scenes resonate best across all users
- Scene resonance by the audience, including gender, age, and location
- Performance by interest, such as fans of different film genres or sports
- Efficacy across platforms, devices, and video durations
While this information is incredibly valuable in its own right, we’ve also built tools that allow us to automatically generate new, audience and platform-specific creative based on these insights. Videos are cut to appropriate durations and aspect ratios and re-published, typically resulting in average view lengths 20-50% greater than the original videos and clients getting nearly double the seconds of video viewed per advertising dollar spent.
Personalized Video Content
The applications of this technology are boundless, and will ultimately result in video being personalized across TV, OTT, and Social on a per-user basis. Unsupervised ML will be essential in optimizing across these platforms, creating video variants and identifying audiences at a staggering scale. It will also help to preemptively optimize content, aligning content to audiences before we run a single impression. Media plans will be shaped in real-time, with emergent behaviors leading to real-time changes in the content we distribute, our budget allocations, and the ad products we leverage.
Benefits of machine learning advertising
The number of benefits provided by machine learning is steadily rising as technology develops. However, even now there are already numerous practices applied throughout different stages of creation and optimizing an advertising campaign that involve the usage of machine learning at least to some extent. Let’s explore some of these:
1. Machine learning can boost ad performance
Keeping track of big data and analytics can be quite tiresome when done manually. Machine learning has great potential to make this process much faster and more precise. There is always a huge number of factors within even one campaign that can be adjusted to improve ad performance. The machine learning algorithm can give you insights regarding what these exact factors are and how you should change them in order to build a winning campaign based on data you input, as well as your goals, budget and other variables of your choice. Furthermore, over time the system gets smarter, especially when the amount of data you gather increases and boosts the accuracy of an algorithm.
2. Machine learning helps to improve ad creatives
Different ad creatives can show different results and this has nothing to do with the overall performance of the campaign. Details such as font, formats, colors, sizing, wording – all can have a severe impact on the success of an ad creative. Machine learning algorithms can analyze creatives from your previous campaigns in order to determine what will work best in the future. With machine learning, it is now possible to predict how your target audience will react to different types of messages depending on their habits, personality types, as well as many other at times unobvious factors.
3. Machine learning & targeted advertising help to make ads personalized
Many companies already struggle with enormous volumes of user data they gather. They hire data science professionals to help them draw correlations and somehow organize all these data sets, but it still takes too much time to deal with it manually. For targeted advertising, machine learning helps to face these challenges and build campaigns where not only creatives will be designed with granular precision, but every single detail about the encounter with the user can be analyzed and improved.
Take, for example, contextual targeting. This type of targeting gets especially important after the GDPR implementation in Europe and the subsequent discussions around user privacy (which then echoed in CCPA privacy regulation). To use it efficiently, a webpage’s content has to be analyzed and understood, so that one could run ads that align with the context and match the user’s interest. With machine learning and its more complicated subset – deep learning, it is already possible to analyze nuances like authors opinions and emotional attitude to the subject. Based on this information, it is then possible to predict what kind of message would work best in this environment and for certain types of visitors.
4. Machine learning can predict your campaigns
Imagine that you could know the outcome of your campaign before it is actually finished. Well, with machine learning this is no longer a fantasy – once you start a campaign, an algorithm starts analyzing it in real-time, comparing it to your previous campaigns and adding customer behavior on the top of it. After you have a whole story it is easier to predict what a big picture might look like. This, of course, doesn’t mean that a machine could make all the decisions humans do, but at least it could save some of your time and budget.
Another important feature of algorithm-based predictions is that kind of “outside-of-the-box” thinking many humans wouldn’t dare to apply. Not limited by cultural or social assumptions that bias our strategies, machine learning can uncover lots of new potential channels we might have missed. And some of those revelations can play a defining role in your campaigns’ budget optimization.
A wise advertiser would always keep track of the latest technological developments and the best ways of their applications. The usage of machine learning in online advertising already showed impressive improvements in targeting, ad performance and campaign optimization. One can easily observe this technology in action using a demand-side platform that already employs such algorithms for campaign optimization.