Language offers additional insights to sentiment and content. The same content can be described with psychologically close or distant language. According to the Construal-Level Theory (Trope & Liberman, 2010), psychological distance influences decision-making. Seven of the eight examined brands psychologically approach customers with their English brand language but psychologically distance themselves from customers with their German brand language on Twitter. Only one brand shows no psychological distance difference between their English and German brand language on Twitter. Implications on decision-making and brand positioning are discussed.
Language has been analysed for sentiment and
content. Approaching language from a
psychological perspective, language may also be
psychologically close or distant. For example, a
brand message may read “From dusk till dawn we
have you covered”. The message hints at enduring
protection and security. With its vagueness the
message is abstract and psychologically distant.
Another way to communicate protection and
security is as follows: “Protects your feet from
rain, mud, and ice”. The second example is much
more specific and concrete and, thus,
psychologically close. Both examples talk about
protection and security, yet in very different ways.
Therefore, the same content can be described in a
psychologically distant or close way.
Psychological distance is important because it
influences how customers process and store brand
messages in memory
The Construal-Level Theory of psychological
distance (CLT)
Psychological distance influences customer
decision-making and is thus important for sales and
marketing. When a choice is perceived as
psychologically distant, such as selecting a
restaurant for a Christmas dinner for example,
customers prefer fewer options
The psychological distance in language can be
driven by any of the four interrelated distance
dimensions
Brysbaert and colleagues have compiled a corpus
with psychological distance ratings for 40,000
words in English
A brand is a name, term, sign, symbol, design,
or combination of each with the purpose of
differentiating one company from another in the
market place
Twitter is an important channel for brand communication due to its notable role in reaching customers and managing customer relationships. Therefore, brand language on Twitter is examined. The sample selection, data collection, data cleaning procedure, psychological distance scoring, and analytical approach are described next. 3.1
The brand tweets were gathered on 13th January 2020 with a historical search using TwitteR from the R CRAN repository. Table 1 details the English and German Twitter handles from which 800 Brand name DrOetker EON Lidl Lufthansa Nivea Siemens Tchibo VW tweets per Twitter handle were scraped.
All brand tweets are stored within the R software environment. The text of the brand tweets was further processed as illustrated in table 2 with an example from Dr. Oetker from the dataset. In a first step, numbers, website links, emoticons, and special characters were removed from the tweets text. In step two, all stop words were removed from tweets with the packages tm and NLP from the CRAN repository in R. In step three, all letters were made lower case. This step only applies to English tweet text as German nouns start with a capital letter according to German grammar. In step four, the remaining linguistic content was replaced with psychological distance ratings. As table 2 illustrates, plurals or tenses were not changed, e.g., “placed” was not changed to “place”, or “waffles” to “waffle”. While such changes would allow more content words to be found in the dictionary corpus, the meaning of the original tweet text would also be changed and ambiguity added as the example of “place” illustrates. Place may refer to the noun, i.e. a location, or the verb, i.e. to position or hire something. Hence, no such changes were made to the tweet text. Step four, the psychological distance scoring, is explained in detail next. 3.2
For tweets in English a corpus was used entailing
concreteness ratings for 40,000 words
These definitions relate well to the ‘mental
travel’ notion
The rating scale was anchored with one (abstract, language-based) and five (concrete, experience-based). If participants felt that they did not know the word well enough, they could indicate this by ticking the option ‘N’ instead of giving a rating. Due to missing values and exclusion criteria, each word was rated between 25 and 30 times. As each word has been rated by at least 25 different people, the ratings contain less bias and are more objective. The language concreteness corpus is thus suitable to measure psychological distance in brand language.
For tweets in German a corpus with
psychological distance ratings for 350,000 words
was used ranging from zero abstract to ten concrete
The ratings for the English tweets were normalised to range from zero to ten. Prior to running the normalisation function, the minimum and maximum values of the original scale were temporarily added to the Twitter data to reflect the full scale range of the original scale. The normalised data were further analysed for statistical significance. 4
A repeated measure ANOVA shows that, on average, the psychological distance ratings for English brand tweets are significantly different from the average psychological distance ratings for German brand tweets (F (1,7) = 14.06, p = .007). On a scale from zero (psychologically distant) to ten (psychologically close) English brand language is psychologically closer (M = 5.00) than German *** p < .001, ** p < .010, * p < .050 brand language (M = 4.41). The same holds true for the psychological distance median ratings (F (1,7) = 5.43, p = .053, Me = 4.58, Mg = 4.38). In order to examine whether the individual brands communicate differently in English and German, the psychological distance in English and German brand language has been compared for each brand separately. Figure 1 shows that, with the exception of VW, English brand language is psychologically closer than German brand language. Especially Tchibo and Dr. Oetker use psychologically much closer brand language in their English tweets than in their German tweets.
These differences have been tested for statistical significance by comparing the mean psychological distance ratings per English brand tweet with the mean psychological distance ratings per German brand tweet for each brand with paired t-tests. The rationale for using paired t-tests is that a brand is viewed as an entity that once communicates in English and once in German. The examined entities or brands, however, remain the same and are thus not independent of each other. According to the results in table 3, all differences between the psychological distance in English and German brand language are highly significant with the exception of the brand VW. 5
Examining how eight international brands communicate on Twitter reveals that the majority do not consistently use psychological distance in their Twitter brand language. Only VW uses psychological distance consistently in their brand language and thus positions the brand effectively on Twitter. VW’s brand language shows no psychological distance difference between their English and German brand language. However, the remaining seven brands, e.g., Dr. Oetker, EON, Lidl, Lufthansa, Nivea, Siemens, and Tchibo psychologically approach customers with their English brand language but psychologically distance themselves from customers with their German brand language. Therefore, these brands appear accessible and affordable in English but inaccessible and attractive, yet unaffordable, in German on Twitter.
By using different types of brand language, these brands weaken and potentially harm their brand positioning. Today’s connected world is likely to accentuate this effect as customers can easily view Twitter handles in different languages from the same brand. The investigated brands use the same brand name, logo, symbols, and colour for their Twitter handle. Therefore, they want to be perceived as one brand or entity regardless of the communication language. Otherwise they would have created a separate brand to sell their products or services within a brand portfolio. L’Oréal or Procter & Gamble, for example, have a large brand portfolio with different separate brands, some of them selling similar products.
For a brand conglomerate, such as L’Oréal for
example, it would make sense to use
psychologically close brand language for their
affordable brands, e.g., Garnier, Maybelline, to
position the brand as accessible and affordable. For
their premium or luxury brands, e.g., Lancôme,
Yves Saint Laurent, psychologically distant brand
language would be more suitable to convey the
brand’s exclusiveness, attractiveness, and
desirability
Language offers additional insights to sentiment
and content. Using the example of eight
international brands, this research shows that brand
language differs in terms of psychological
distance. Psychological distance is important
because it guides whether customers focus on
feasibility or desirability considerations when
reading the brand message
Given the pioneering nature of this research, there are a number of limitations and areas that warrant further investigation. The two dictionary corpora employed in this research differ in terms of compilation method and size. The German corpus builds on the English one but uses machine learning to generate more psychological distance ratings. Machine learning could help to augment the English corpus or create a psychological distance dictionary corpus for another language. Another limitation concerns the choice of brands. The selected brands are internationally known but originate in Germany. Further research should compare these brands with brands that originate in Anglo-Saxon countries to examine a possible country of origin effect. Another fruitful area to explore is differences in psychological distance between the English and German language, and other languages, if more dictionary corpora become available, to examine the presence of systematic differences across languages and cultures.