VIER Emotion Analytics is able to extract psychological results from written and spoken language. The AI has learned this with the help of millions of data sets and a multi-stage modelling process. The AI-based software makes statements about psychological effects of speech (e.g. "authoritative", "visionary") or psychologically relevant outcomes (e.g. "linguistic coercion", "relativising speech") of speech.
The analysis gives managers, employees and even entire departments the opportunity to make their own language impact objectively measurable, to understand it better and to change it. In this way we can communicate in a more targeted way, prevent misunderstandings and convey exactly what we want to get across.
Our AI is designed to help improve communication and understanding. The word "emotion" comes from the Latin word "emovere" and means "to move out, to set in motion, to put in an excited state". This paraphrase reflects the central characteristic of emotions: emotions touch us and they move us in a certain direction. For us, this means: We want to make measurable and conscious what moves people and what they transport through language. It is about successful communication through the right effect of language - we do not use it to analyse the character or nature of a person.
Our AI needs language that can be analysed. This can be done in any conceivable form, as both written text and spoken language. Interviews, chats, emails, articles, transcripts, press releases, website texts, even annual reports can be analysed with VIER Emotion Analytics.
We have also established an objective and convenient standard process for very individual analysis: The participants provide a speech sample - this is done via an automated telephone interview lasting about 20 minutes, during which various questions are asked. That is all. The software then uses purely formal language criteria to determine how the language works. Things like intonation or speech tempo have no influence on this, by the way.
In fact, they are not evaluated at all. The software merely arranges the modes of action around an average value. It does not determine which characteristic of a result is "good" or "bad". This always depends on the position or activity from which communication takes place and with whom and why. For example, a high value in the area of empathy can be an advantage in customer service, but not so much in controlling.
That depends on the trade fair level under consideration. If only emotions are analysed, mood can make a difference, as expected. In the more complex, multi-level analyses, which map different levels of results from language structure to psychological effect, drawing on over 110 million parameters in the language, the current mood has an extremely small effect.
Artificial intelligence is necessarily objective, anonymous and does not evaluate - it's like automatic spell-checking. When software determines one's language effect, there is no warning finger. The result is completely unaffected by sympathy, attractiveness or other subjective criteria. This facilitates the acceptance of the results and enables the derivation of appropriate, meaningful and individual training measures for each person.
The results can be used wherever verbal communication plays a role, e.g. to optimise one's own communication or that of entire departments and companies and thus address the recipients in a targeted manner. By using the software, managers and employees can better understand people's motivation. This makes it possible to communicate more effectively in recruiting as well as in leadership development, employee development, sales, marketing and customer service.
Managers can optimise their communication
Employees communicate and interact more effectively
Sales and marketing can formulate e-mails/advertising letters in a targeted way and thus increase the closing rate.
Employee development can be more individualised
A company's communication can be standardised
The communication culture within the company can be analysed and improved
Analyst evaluations of financial market communications can be better managed
Touchpoints with customers can be optimised This increases relevant KPIs (NPS, AHT, etc.)
and much more
Of course we offer personal feedback with an individual, situational and requirement-related assessment! The current situation of the participants is discussed and the results are related to it. For example, a manager who has to communicate negative changes to the team has completely different requirements than a person who is in a technical discussion. By integrating the current requirements, a higher personal relevance of the result is created.
Yes, our solution is based on a language analysis patented in Europe and the US and the world's largest study linking AI, language and psychology: it includes more than 38 million ratings, over 25,000 people, an AI model with more than 110 million parameters and the analysis of over four billion words.
Yes and no - VIER Emotion Analytics only learns continuously in the studies that are set up for the AI to learn. Feedback from live situations, on the other hand, cannot and should not be processed for various reasons. For example, data processed by the software is not stored at all and thus cannot be assigned to any specific process. Furthermore, the quality of the incoming data is crucial. We pay attention to objectivity and representative distribution when generating data. If the system were to simply learn on the basis of the incoming data, undesirable algorithms could emerge. Checking for systematic biases in the results (e.g. by age or gender) is a central part of the quality criteria. In this way, we ensure that no unnoticed biases ("biases") are picked up even as the model learns more. Moreover, the system needs not only language for improvement, but also estimation, so that it knows what is to be learned. This data can only be meaningfully collected through a study design. Learning in live operation is therefore neither possible nor desirable.
No, currently the results are available in German and English. Further languages are being worked on.
Yes, it works, as long as the language level allows fluent communication and free expression. This corresponds to the B2 classification.
Yes, because VIER Emotion Analytics does not store the analysed client data at all. Thus, they cannot be assigned to any specific process. Furthermore, VIER generally attaches great importance to the optimal protection of (personal) data. To this end, extensive organisational and technical measures within the meaning of Art. 32 DSGVO are implemented, which pursue the goals of confidentiality, integrity and availability of the data. If a VIER solution is used that stores data, all solutions and products do so exclusively on servers in Germany - i.e. in the German cloud.
Do you have questions about VIER or would you like personal advice? Contact us now via our contact form. You can reach us at any time via the contact form, by e-mail to firstname.lastname@example.org or by telephone on +49 511 95 73 95 1111. We look forward to hearing from you and will get back to you as soon as possible!
Please reload this page in your internet browser if the form is not displayed.
VIER whistleblowing system
Do you have something unpleasant to tell us? With the VIER whistleblower system, our partners and clients also have the opportunity to anonymously inform us of grievances and violations. All tips will be checked and, if necessary, followed up!
To the whistleblower system: https://transparency.vier.ai/