How computer vision will continue to transform advertisement in 2020

Gone are the times when advertisers aligned with the strategy of being over-pushy and annoying. Now, marketers understand the value of personalized and relevant ads and do their best to bring informative content to the users.

Artificial intelligence and computer vision in particular now play a big role in transforming the traditional form of advertisement into a brand-new approach. We decided to have a closer look at the possible applications of computer vision in the marketing industry in 2020 but first let’s get it clear with computer vision definition.

What on Earth is computer vision and why do we need to care?

Computer vision is the subset of Artificial Intelligence that aims to train and teach computers (machines) to identify and interpret the images and videos. However, the machines are not only able to identify or classify the objects that they see but they are also capable of reacting to them.

Computer vision technology uses incredibly massive datasets to learn and the whole process of object recognition and classification is much faster and more accurate than recognition by a person. This is the reason this technology is now becoming so popular and the number of its applications grows in a steady manner - computer vision is capable to perform many tasks in a much more efficient manner than people do.

An example of a computer vision application is medicine. Now, machines can identify cancerous tumors on a medical image which allows doctors to timely take preventative measures. And this is just one example of what computer vision is capable of. Now that we are clear about this technology, we can move on to its application in marketing.

Integration of relevant ads

Modern advertisements are far more personalized and relevant than they were in the past but still, it’s a long way to go. Computer vision aims to close this gap by offering an option to place the relevant ads right over the content the user is interacting with at the moment.

Here is how it works. When a user browsers a page, he stumbles across an article about dogs. He starts reading and at a certain moment, he will be served an ad of a premium puppy chow. And that will not be a coincidence because the computer will know exactly when and how to serve this ad!

While it sounds a bit wild, it actually is possible. Some companies like GumGum already work in this direction and here is how they explain the process behind the ad placement:

• The computer vision technology scans the page (meaning, all images, videos and content) and analyzes the context to understand what exactly the page is about.
• The technology selects the most relevant ad and integrates it with the page’s content (the ad placement will be the area that gets most user’s attention).
• The user becomes more engaged and interested in such ads.

The key thing here to remember is that such advertisement is now possible due to the fact that machines start recognizing images on the same level people do: not just identifying the object but understanding it. For example, a machine now can detect a “German Shepherd breed” and not just a “dog”. With such advancement of technology, in-image ad placement seems to soon become a standard.

Collection of insights about the audience

The key to delivering personalized and relevant ads is a thorough understanding of the audience’s likes, preferences, and habits. Right now, marketers utilize the power of analytics to better understand the users but computer vision can take the collection of insights about the audience on a whole new level.

What industry experts aim to do is to use computer vision to analyze user-generated content (UGC) and, based on it, better understand their target audience and its primary interests. There are more than 1.5 billion photos that users upload to Facebook, Snapchat and WhatsApp on a daily basis - now imagine how much available insights are there for the marketers.

Smart ads based on the user’s emotions

The most important thing that any marketer needs to know is whether the ad delivered the expected results in terms of customer engagement. While online advertising can be measured and tracked in terms of analytics, offline advertising cannot be actually measured. With this in mind, some companies started to present their solution to the issue and used computer vision technology for that.

The biggest example of this brand-new offline advertising is probably the AI poster by M&C Saatchi, London. The company came up with an idea of a self-learning and self-generating poster that uses computer vision to recognize the people’s emotions and adjust the content correspondingly. The poster consists of several elements (layout, font, color, image, etc.) and it automatically generates a new version of an ad based on how people interact with the current one. The elements that cause the most interaction will be used in the next poster versions and the ones that were ignored will be eliminated.

In this way, the poster constantly renews itself and eventually comes up with the most appealing and engaging version of a single ad. And this is only one example. Companies slowly start using computer vision for analyzing the users’ reaction to physical ads and this will allow to significantly improve physical advertising.

Summing up

In general, computer vision will allow marketers to better understand our online habits and interests and serve us a more personalized content that will be less annoying and more relevant. Of course, there are certain risks related to the use of computer vision in marketing (i.e. the collection of faceprints which may be a legal issue) but overall, this technology seems to bring many benefits to the world of both physical and digital advertising.