Artificial intelligence is changing the face of in-video commerce

Every day we receive several thousand advertising messages: whether on the Internet, television, radio or on the street. We are becoming resistant to ads and, above all, tired of their presence, and marketing clutter does not allow to get acquainted with the content that is really desired. In 2017, as many as 46% of ads in Poland were blocked with ad blocking plugins. This is the highest result in the world and this indicator is constantly growing – a year earlier 36% of Polish internet users declared the use of adblock.

Overcoming this problem and reaching the target audience is becoming more and more challenging for marketers, publishers and advertisers. To achieve their goals, they must look for new solutions, often originating in modern technology, and modify previously implemented methods. One of them is in-video commerce, which thanks to the use of artificial intelligence and machine learning is already changing the face of modern advertising.

The hopes of video marketing

A number of statistics attest to the popularity of video marketing: by 2019, online video traffic will account for 80% of all online consumer traffic worldwide (SmallBizTrends), and Facebook alone generates an average of 8 billion video views every day (Social Media Today). We’re watching video not just on computers anymore, but also on phones: YouTube alone is seeing 100% year-over-year growth in mobile video consumption (HubSpot). The video format makes the content presented easy and quick to digest, and allows a greater emotional charge to be conveyed. No wonder that video attracts not only viewers but also publishers and advertisers.

For e-commerce businesses, this means more advertising opportunities and the chance for potentially more effective campaigns. For publishers, video is high-quality content that is readily consumed by web users. Video, as a medium, can combine the interests of publishers (seeking to monetize content) and online stores (seeking to increase conversions).

The answer to these plans is in-video commerce collaboration. This advertising content is embedded in the video, matched to the content and context in which it is displayed.

Why context matters?

In the age of the so-called. Banner blindness – the effectiveness of advertising is increased by the context in which it is presented to the recipient. Ill-considered creative and inappropriate time and place of its publication irritate users – they see inappropriate ads, not matching their potential purchase impulse and, as a result, they take actions to close the ad or the whole website.

Contextual advertising is matched as precisely as possible to the content it accompanies. Makes a natural, even obvious impression and offers a personalized shopping experience. The chances that a user will further interact with such prepared advertising content therefore increase, putting both the publisher and advertiser in a comfortable situation.

How in-video commerce works?

In-video commerce is about presenting relevant products to the user while the video is playing. Products are matched to the content, i.e. identical or very similar to those seen on screen. This personalization effect, combined with the emotions that accompany video viewing, triggers a purchase impulse that, thanks to the precision of in-video commerce, can be satisfied immediately. After clicking on the ad, the user is sent to the store, where he/she can complete the transaction in a few seconds. The time of purchase decision plays a huge role here, but the selection of content and appropriate recommendations in e-commerce stores are crucial. High-quality content combined with relevant advertising can positively impact user retention and conversion on both transaction sites.

It’s a win-win-win situation: the recipient receives personalized, relevant content, the publisher serves up the right user experience, and the e-commerce store has a chance to precisely reach the user.

How it looks like in practice?

Imagine a program about mountain climbing. Sharp peaks contrast with billowing clouds, and the protagonists of the episode overcome obstacle after obstacle to reach their destination. Such an image carries an emotional charge, associated with wanderlust and the desire to travel in the viewer. There is a chance that after getting familiarized with such content, the user will be inspired to go on another trip, think about buying new trekking boots or a backpack perfect for further expeditions. This is an ideal place for advertisers to present products relevant to the content and respond to the potential needs and desires of the recipient.

Content is analyzed: objects and the context in which they are presented are recognized. Relevant image elements are described with tags e.g. trekking, mountain, backpack (with the exact size, color or type), which are then linked to products in e-commerce stores. Everything happens in seconds, enough to trigger and satisfy purchase impulses.

Artificial intelligence as an engine for in-video commerce 2.0

In-video commerce is not a new term, but it is only the use of artificial intelligence and machine learning that can automate this process and make it as accurate as possible without incurring expenses, human resources and time. Analysis of content and its appropriate description with tags, as well as linking it to product feeds of e-commerce stores, allows for a new dimension of monetization, and tagging content itself opens up the possibility of creating new content quickly. This process, performed manually, would be impossible or very time consuming.

Plastream automates these processes, changing the face of in-video commerce. By using artificial neural networks (deep learning) to automatically analyze image content, it is possible to match content to product images in e-commerce stores and automatically reference. This is an important support for publishers in contextualizing ads to visual content. Plastream goes a step further, however, by responding in real time to changes in its product range if a product is currently unavailable. It searches for the most similar replacement in terms of many features such as color, cut or material.

The system is able to recognize not only objects, but also the characteristics assigned to them (color, fabric, length, pattern), faces and the context of the material. Such tagging allows for quick and automated creation of new content. If we need pictures of a celebrity in long dresses, with the right tags, it can be searched in a short time and content can be prepared.

Many marketers think that in-video commerce is the future of advertising. However, thanks to advanced and constantly developed technology, this format is available today.