This article explains…
- the content-to-commerce approach
- why contextual advertising is so important
- how artificial intelligence goes hand in hand with videos
- the role of emotions in the success of contextual advertising
- how to achieve an above average CTR
Although the definition of Content-2-Commerce has existed for a while marketing specialists still haven’t been using it to its full potential. This would possibly help achieve many business goals. The importance of this approach is that it uses artificial intelligence, machine teaching (also deep-learning techniques), big data as well as omnipresent personalization. This is happening thanks to the increased availability of power calculation and technology.
For many portals and internet media the main goal is to sell from surface advertising, the reason being every person that spends money or sells a product is able to predict the effectiveness of this action. The sale of advertising reminds me of selling insurance policies. The insurer sells policies for the amount of time that he can predict risks. The market is quantified and the success can be measured. This factor determines any action and success in business. It is less significant, in the era of blocking ads (Poland is one of the top users of AdBlock) and communication chaos caused by information overload, that ads are less effective and receiving a better CTR is more difficult.
According to the Online Video Forecasts report published by Zenith in 2017, the longest time most viewers spend watching videos online is on stationary devices. The average is 19 minutes a day. This is only one of the statistics confirming the quick and stable growth of video content consumption, which follows growth in potential marketing lying in this medium.
In the context of so many factors lies a question:
Is there any possible solution for more valuable consuming of video content by the user? Is there any way to monetize a video other than using ads?
The answer is the effective use of content-to-commerce.
What is content-to-commerce?
Shortly speaking it’s the combination of the chosen context, for example videos and pictures with suitable products online. Content to commerce isn’t anything new, but along with the development of artificial intelligence and the Similarity Search technique, searching through recognition, it’s possible to automate processes and put more focus on precision and adjustment of content and context.
It is said that content is the king, when in reality context is the Caesar and dictates the conditions. Content placed in the right context has a greater chance to form a successful e-commerce connection. TV publishers, owners of portals with a great number of videos and pictures, bloggers, owners of online stores should be looking into this. With the productive and efficient help of publishers, these owners can increase their profit and receive valuable action.
Clients will accept and react positively to ads if they are personalized and added in the appropriate context. Clicking them will calm their impulsive shopping decisions. Emotions accompanying the recipient of a video or picture can be turned into a transaction thanks to in-video commerce. This trend can be observed not only in normal video content, but also in public content published by traditional publishers: TV and their VOD platforms.
Video consumption is associated with emotions that can quickly be turned into a transaction just with the help of a click. In order for this to happen a few conditions must be fulfilled:
- the right content– qualitative, valuable videos or pictures, that will easily give in to Similarity Search: the detection of objects in a video/picture. The technology used in Plastream during tests showed 95% precision in the face and object recognition and over 98% precision in recognition of products in supermarkets. This is the first step in transferring a customer to the online store.
- the right context– advertisement needs to be identically matched to the message. General, irrelevant content to context isn’t enough for clients. The brand needs to do something more in order for the recipient to get acquainted with the content on a purchase level. Something more is placing advertisements in the context as deep as possible- right when introducing the product, first impulsive purchase- in the correct moment of the video. Example- video of a woman walking on a beach in a white dress. The context is not only the white dress, which can be found online and to which a link can be added to a similar product in an e-commerce store. The context is also the situation: the beach, the sun, the ocean- this is the perfect context for advertising a travel agency or airlines. Context is also text that might be describing a totally different thing that a company wants nothing to do with. (ruling out) How many times have you fallen for this? We live in times where a shopping impulse to purchase a dress is induced by a four-second scene in a YouTube video. It’s worth using this in marketing!
- the right connection with e-commerce- let’s go back to the situation from the bullet above. Finding a white dress online shouldn’t be a problem but finding the exact dress or a very similar one can be difficult. Not for platforms based on Al that used advanced mechanisms to search online stores for the perfect match.
The effect of fulfilling these conditions is the ability to obtain an above average conversion, higher than average CTRs and low CPC- as well as the feeling of delivering more value to recipients who would most likely want to satisfy their shopping impulses.
How does this work in practice?
Deep learning and neural networks used in platform technology content-to-commerce (just like Plastream) allow users to find exactly what they want at the moment in which they feel it’s what they want. Platforms combine content with products allowing recipients immediate acquisition of products they see on video or pictures which fulfils their shopping needs. This mostly happens with clothing or accessories but interior equipment and tourism can also use this technology.
Pictures and videos are subjected to DETECTION– the process in which products are shown are detected and catalogued with precision, for example, pieces of clothing. (long sleeve, short sleeve) In addition, a text analysis is carried out that summons the results so they are as close to the original as possible.
A blogger writing about a dress -> thanks to the text analysis we can focus on searching for the actual dress despite the buildings, umbrellas or cars in the background.
An article about a celebrity’s shoes -> we will focus on the celebrity’s shoes only despite the celebrity’s pants, shirt and hat.
The next stage is the EXTRACTION function- here the exact features of the product are recognized which helps differentiate it from others. Some examples of these features are the pattern of the fabric, the exact colour or style.
The penultimate stage is SIMILARITY SEARCH which focuses on finding the right product in the base of millions of imported pictures analyzed by developed algorithms with the help of the unstructured engine Big Data or the relational database of an e-commerce store.
|dress||red, cotton, short, short sleeve||similar products in e-commerce stores|
|jacket||blue, glittery, long, long sleeve||similar products in e-commerce stores|
|trousers||black, with holes, jeans, long||similar products in e-commerce stores|
It’s interesting how there are many choices on locating the results on the picture. Thanks to advanced technology there are few options:
- displaying results after pausing playback
- displaying results during played material or directly on the picture
- displaying tags suggesting the product, brand or service
- displaying results in the form of a personalized widget above, under or next to the picture.
Guaranteeing the matched content located in the right context allows the noninvasive and very natural location of chosen products in online stores and lowers risks of negative feedback. Simple mechanisms allow recipients to get to know the content quickly and help them make a quick decision. Thanks to efficient content tagging the product can be found quicker, the terms of purchase can be read and a transaction can be made.
Getting to know the product to purchase it is only a click away. They can forget about screenshots, persistent searching on forums, questions and comments- all of which in the past had aimed to find a certain product after viewing a picture or emission of material. Now thanks to new technology they can buy a similar product or identical product because they see it in the video.
Content to commerce has modified rules between online stores and publishers, but the availability of new technology has helped create an actual difference for everybody:
- recipients receive content of greater value that matches the context they want to view and explore. The result? General satisfaction and better UI.
- advertisers (entities of e-commerce) are happy with the valuable movement- new recipients go to them after viewing certain videos and pictures, they are usually more determined to purchase the product than others who discovered the product through ads with no context.
- publishers can calculate their e-commerce and gain extra profit from their business without conflict of interests with other monetization options. They can also note their increased retention index thanks to the noninvasive form of advertisement on their portal. Everyone mutually benefits from their activity.