Too good to be true?
Kununu, TripAdvisor, Glassdoor, TrustPilot, eBooking, Google My Business and many more are popular and frowned upon at the same time. But what distinguishes authentic and helpful reviews with added value, how can they be recognised and what contribution does artificial intelligence make?
People are confronted - consciously and unconsciously - with evaluations. They learn to evaluate what they have experienced and are also subjected to such evaluations themselves. It is therefore hardly surprising that it is now possible to publicly rate employers, training, services and products. Ratings are "big business". It depends on them whether employers find the right talent and travellers are lured to the "wrong" hotels. These subjective assessments vary in quality and usefulness, depending on the assessment criteria.
But how can the essence of subjective impressions be efficiently recognised?
How can an effective benefit be recognised?
First of all, the question arises: What drives the evaluators to express themselves in the first place? There are people,
- ... who share positive experiences out of joy in a self-reflective manner
- ... who, out of anger, self-reflectively share negative experiences
- ... who, out of greed, unreflectively share positive experiences
- ... who, out of envy, share negative experiences unreflectively
All raters have a common motive: they want to achieve something. From the reader's point of view, ratings can be:
- positively useful
- negatively informative
- positively misleading
- negatively destructive
Every evaluation is first of all interesting. But are the statements also relevant? How can the wheat be separated from the chaff? The evaluations based on large amounts of data paint a picture as already described by contemporaries in antiquity: "The whole art of language is to be understood", said Confucius.
What to look for in ratings
In general, it is important to pay attention to the following when using assessments:
- Look out for linguistic conspicuities such as long sentences, many technical terms, repeated words, few adjectives, typos and punctuation errors, emotional terms that do not make sense, expletives, oxymorons, neologisms, etc.
- Without commentary/justification of a rating, a score is hardly usable
- Excessive positive scores by countless users make one wonder. The principle of "too good, to be true" applies surprisingly often
- A conspicuously large number of negative ratings are also suspicious
- If the majority of the ratings are in the upper middle range, the score is credible
- If evaluations show a similar effect of the language, this indicates the same author
- Evaluations are probably authentic when the language used is balanced. They appear direct, entertaining, simple, friendly, critically self-reflective and follow a recognisable thread without internal contradictions.
- Pretended, manipulative statements, on the other hand, seem long-winded, complicated, aggressive, unreflective, (too) friendly, demanding, intellectual, convoluted.
Support through AI
Innovative AI speech analytics is able to recognise the effect of language. The AI diagnoses emotional and rational messages, from which strengths and weaknesses become apparent. In this way, pearls of great added value can be distinguished from unwelcome and possibly harmful developments. Deceptive glossing over, but also war rhetoric positions become visible in this way. This helps to distinguish such assessments from helpful, authentic statements.
Stephan Siegfried is a lawyer, entrepreneur and has published several books and professional articles on the subject of speech analytics.
Contact: firstname.lastname@example.org or www.sprachanalysen.ch
Would you like to learn more about the application possibilities and opportunities of VIER Emotion Analytics? Your contact Philipp Grochowski is looking forward to your enquiry!