placement of advertising Archives

Emotional ad positioning and emotional ads

by Sentimenti Team | Jun 6, 2019 | SentiBrand

Emotywne pozycjonowanie reklam, czyli targetowanie reklam wg emocji od niedawna nabrało nowego wymiaru. Okazało się, że komunikat można oprzeć nie tylko o kontekst, analizę ruchu na stronie, badania demograficzne, płeć i wiek ankietowanych, ale po dane sięgnąć też niemal w głąb ich serc. Jak to możliwe? Wystarczy rozpoznać ich emocje.

Emotional ad positioning and example: New York Times

In 2018, the New York Times conducted a study on the emotions of its readers. This was based on self-learning algorithms and combined with an analysis of feedback collected from readers about how they felt after reading the content of specific articles. The result of this research was an emotion prediction tool that indicates joy, sadness, hope, and 15 other emotions in readers, among others.

Not content to let this tool predict the emotions that NYT readers might potentially experience while reading future articles, the company went straight to selling advertising space. It was offered to owners of products with emotional character close to the emotions contained in the given articles. The possibilities turned out to be impressive: the tool made it possible to examine and create the emotional content of a given article and to better match the marketing message to it.

Marketing content so emotionally targeted and appropriately placed among other content achieved up to 80% better results than classic behavioural targeting (on average by 40%). The tool even made it possible to separate content with negative or disturbing undertones, so as not to add advertising messages that might fit the content or the reader’s profile, but are completely inconsistent with the tone of the text: New York Times.

Emotion targeting – emotional ads: perspectives

The agorithm can be applied not only to the articles contained on this site, but also to news and publications of other types. Therefore, it opened a whole new field for campaign creators. This resulted in 50 campaigns and over 30 million collected feelings, sentiments and emotions. Advertising messages were usually placed next to entertainment or corporate social responsibility content.

Interestingly, similar research was conducted in other editorial offices, including USA Today and The Daily Beast. The analysis was based on phrases (keywords) and emotions related to their meanings, and an attempt to answer the question of what mood current readers of a given text are in based on behavioral analysis of their actions on the website and frequency of returning to specific, emotionally charged content.

USA Today’s research has shown that readers don’t limit themselves to positive news, but read everything. This means you can target your message to them not only when the context is similar to the rest of the content, but also when readers are in a similar mood to the context of the content. Therefore, such a method allows you to more effectively create content for better communication of brands.

The Daily Beast, on the other hand, instead of trying to guess moods, indicates where on the site readers will spend the most time; in these popular places it tries to contextually place the marketing message. All based on positive emotions and negative emotions in advertising.

The future of the advertising market?

The described activities based on data analysis, algorithms and artificial intelligence are beginning to be the future of the advertising and public relations market. How do the market of ordering parties perceive these new solutions? It would seem that with such precise tools for targeting recipients, there is no need to worry about anything else. And yet opinions are divided.

According to some experts, basing a campaign solely on such “bought emotionality” is one-dimensional, restricts and narrows the field of activity and should therefore be associated with other methods of communication. On the other hand, it is an excellent solution for companies looking for safe solutions, making their marketing message more precise and targeting the most determined customers.

Sentimenti and emotional advertising. Identifying emotions in online advertising

Since the New York Times includes emotion ad positioning on its pages, the solution must work and be effective. Positive and negative emotions are taken into account. Is it possible to apply similar mechanisms in Polish?

Until now, this was not at all obvious. Algorithms for automatic processing of our language have been improved to such an extent that they are great for text analysis. But what about the emotions expressed in them? Until now there was no database of words, phrases or even whole texts written in Polish.

That is why the Sentimenti team created it. The database was created in the course of research we talk about on the blog and at academic conferences. It turned out that with good data it is possible to create an effective system of sentiment and emotion analysis and with it – ad positioning.

Interia Emotions – different emotions, one goal

We are now at a similar stage to where the New York Times was about a year ago. We have an application that efficiently analyzes text and the emotions it contains. We have started cooperation with Interia portal – we are creating an emotional map of its thematic services. From here it is only one step closer to taking the overtones of an article into account in ad positioning.

What is very important, emotive ad positioning does not mean additional duties for journalists. We will not tell anyone what emotions to express, because in practice each emotion creates an appropriate environment for ads. Text has a sad meaning? It is best to place an ad with ecological overtones. It expresses fear? This is a good context for pairing the article with an ad for insurance or dietary supplements.

The next step of the Interia Emotions project will be to investigate how exactly emotions in text react with ads. Therefore, when we check this, emotive and effective ad positioning will become a fact. Such a tool will certainly prove useful. Ads positioned based on the content of articles (rather than tracking the activity of Internet users) are less irritating for them.

Emotions in words: using Sentimenti in the advertising and marketing sector

by Sentimenti Team | Jun 5, 2019 | Conferences, SentiBrand

Emotions and marketing? Communication with clients and contractors, searching for new customers and markets for goods or services, public relations policy, social media activities, advertising and generally building the company’s image – these issues require business to collect data and make informed decisions based on them leading to the development of the organization. Today, thanks to Sentimenti’s emotion analysis tools all these activities can be conducted more effectively.

Emotions in brand communication

The capabilities of the tools for analysing emotions in text developed at Sentimenti are still being tested and extended by our research team. We decided to examine advertising content with them. This is a demanding area. The advertising message has to be short, catch the attention, be memorable. Above all, it has to be suggestive so as to induce a specific behavior of potential customers. To do this, it is best to hit their emotions. From the research conducted in this matter it is known that the marketing message with an emotional foundation is more effective, and from practice – that emotions, even the negative ones, sometimes can provide a brand with exposure in the media and social media for a long time.

We analyzed the content of several advertisements in terms of their emotional overtones and the valence of 8 basic emotions from the Plutchik model. For the study we used advertising content related to AA Prestige (cosmetics line), Plus (telecommunications), Citi Handlowy (external factoring), Renault (Kangoo Express model) and Shell (Shell Helix oil).

The analysis confirmed the results of the research – advertising content characterized by positive emotions (joy, trust, expectation) is more likely to be remembered by recipients and to inspire trust in a given brand. But that’s not all – the in-depth analysis of emotions in advertising content has also shown that even a single word (also a function word!) can be associated with several emotions at the same time.

Consequently, it is not always possible to avoid negative sentiment in advertising content. You can always check this content with Sentimenti. In this way you can better match the content and form of your marketing message to your target group. Sentimenti will show you emotional connotations where the popular two-dimensional sentiment analysis does not show any. This is a big advantage when you have more cross-sectional data to analyze. Interestingly, however, some advertising campaigns (especially in political marketing) are based on negative or at least controversial emotions. Why?

The recipient processes them faster. They are more memorable and evoke a desire for action: most anger and the repulsion leading to it, fear and…. surprise. Research shows that the recipients will remember above all this negative emotional environment, not the brand and the advertised product. Unless, of course, we skillfully reverse the sign of emotions. But to do that you need to know what words evoke what emotions. We at Sentimenti have such knowledge, we also have algorithms to evaluate emotionality in content. We can share them with you.

Emotional targeting now works

Emotions transfer easily. And not just between people, but also, as research shows, between words. Based on this insight, the NYT, USA Today, The Daily Beast and other U.S. editors have begun emotionally positioning advertising content around articles with a given message. Algorithms built in collaboration with the editors analyze readers’ opinions of articles and try to answer the question of what their reactions will be in similar content so as to match them later with the best possible ad content placed in the vicinity.

This solution allows you to create and place content more contextually, which translates into greater brand marketing effectiveness. At Sentimenti we understand the needs of both publishers and advertisers in this field. That’s why we started working with Interia Group in the pioneering InteriaEmotions project on the Polish market. Together, we work on matching the emotional context of advertising content with the content of publications. All this so that companies and agencies can benefit from the advantages of this type of positioning. We presented the results of our work at the IAB Conference in Warsaw.

Why should you start working with Sentimenti? Algorithms, machine learning and other modern technologies are increasingly active in the public space. Business is increasingly willing to use these methods. A human being cannot analyse a huge amount of data quickly. He needs to use artificial intelligence to do it better and more efficiently. This is especially true for emotions, which are perceived individually and therefore difficult to objectify.

At Sentimenti we understand this direction. Therefore, while maintaining scientific standards, we develop algorithms to analyse emotions in content of various types. Our tools have proven effective at analyzing emotions in ad content, comparing them, showing correlations, trend direction, etc. This is a massive range of data. It usually takes months to acquire and analyse. It requires an objective indicator to measure emotions. After all, several people, independently of each other, may evaluate a text in a similar or different way. The goal of effective emotion analysis is to understand what the standard is for such evaluation.

If you are interested in improving the emotionality (and thus the effectiveness and increasing the reach) of your advertising messages, if you want to write more engaging posts in social media, if you need an analysis of your current activities or an indication of the direction of their development – we invite you to contact Sentimenti!