Brands tracking how much they are trusted by consumers are placing their trust in the wrong metric, according to founder and CEO of leading strategic research agency, Fifth Dimension, Lyndall Spooner.
“Brand trust is a hot topic right now; however, companies that previously promoted brand trust to be a strong brand health indicator are now questioning the value of measuring brand trust based on the science,” Spooner said.
“Trust between a consumer and a corporation is different to trust between two people or the trust people have in the government or their doctor. Trust is an emotional brain state that allows us to be vulnerable to another person, believing they will not take advantage of us or cause us harm.
“When we trust people, we can share with them our secrets. When we trust people in government, we hand them power hoping they will act in our best interests. Our relationship with companies is very different.”
Fifth Dimension conducted three independent experiments over 18 months – each one diving deeper into understanding the limitations of measuring brand trust – which produced three key insights:
- Brand trust does not align to consumer behaviour, it does not predict customer value or prove customers are emotionally connected to your brand,
- Your company’s likelihood of being listed as one of Australia’s most trusted brands can be accurately predicted by an algorithm; and
- In a technically disrupted world trust has shifted from companies to people and technology as increasingly we buy from brands we don’t even know.
Fifth Dimension’s first experiment was never designed to focus on brand trust; instead, the agency was modelling the accuracy of customer experience (CX) metrics to predict consumer behaviour.
The research was conducted with over 5,000 consumers with CX measured across multiple brands and industries, analysing the statistical strength of CX ratings, which uncovered that brand trust was the weakest brand metric.
“The fundamental flaw in asking consumers if they trust a brand is that the brand trust question has two meanings, based on the respondents’ mindset. If you ask a consumer if they trust a brand, 60% interpret the question as asking if the brand is capable to do what it promises, while 40% interpret the question as asking about the character of the brand, if the brand is ethical and honest in how it operates,” Spooner explained.
“The way in which people interpret the brand trust question has a major bearing on the results of whether or not they trust a brand. People with a character mindset are 88% more likely to say they distrust multiple brands across multiple industries, while consumers with a capability mindset are 30% more likely to trust a brand, and this is when both groups have the same level of positive customer experiences with a brand.”
The research found that two-thirds (66%) of consumers said they had no reason for giving the brand trust ratings they gave because it was considered the hardest question to answer of all the CX metrics tested.
“The problem with asking ‘to what extent do you trust brand A?’ is a lack of context. If you were to ask consumers if they trust a bank, they will probably say no, but if you ask them if they trust a bank to look after their money they will probably say yes,” Spooner said.
“In our experiment respondents told us that a brand trust question with no context is not only difficult to answer it is also a question they are likely to fabricate a response to, so the data is made up.
“Imagine finding that respondents have just told you, and our modelling has proven, that their ratings of brand trust have nothing to do with their experiences with those brands. But what if I also told you the level of trust consumers say they have in your brand is predictable with a clever algorithm, and that algorithm does not need to know what consumers think of your brand or what your brand stands for.”
In this second experiment, Fifth Dimension replicated the methodologies of several well-known brand trust studies and looked to see if they could reverse engineer the results; could they predict the brands most likely to be nominated as trusted, the level of trust in a brand on a ratings scale, and the ranking of trust across brands?
“We found it was relatively easy to predict levels of unprompted and prompted brand trust, as well as accurately predict the rankings of brands within and across categories because brand trust is not about emotional connection or vulnerability. Your likelihood to be nominated as a trusted brand or given a higher brand trust score on a rating scale can be predicted from measures of your brand’s presence in the market. In other words, salience,” Spooner said.
“Your brand’s presence in the market is driven by the industry you are in. Brands that frequently engage with consumers, such as supermarkets and mobile phone providers, have a much higher market presence than brands that hardly ever engage with consumers, such as funeral directors.
“Brands that are seen to advertise a lot also have a much stronger market presence, as do brands that are well established and have a large customer base. As consumers, we see these brands and hear about these brands during our day-to-day lives and are therefore more likely to be able to recall the names of these brands if asked to list brands we know.
“That is why we can use these brand presence factors to predict brand trust. When a respondent is asked to list the companies they trust in a survey, they are not telling you the brands they trust the most, they are telling you the brands that most come to mind, and the brands that most come to mind are the ones with a stronger brand presence.
“In our experiment we asked respondents to nominate the brands they trust and then later we asked them to rate the extent to which they trusted brands we listed. If you compare what is called the unprompted levels of trust and brand rankings with the prompted levels of brand trust and rankings, you see very different results.
“In both cases, unprompted and prompted brand trust can be predicted with an equation using brand presence metrics. But the equation differs by what elements of brand presence have more impact between predicting unpromoted and prompted brand trust. The brands listed as trusted by respondents unprompted are not the most trusted brands when brands are all rated on the same ratings scale. To publish a list of the most trusted brands without measuring the level of trust in all brands reported is misleading.
“Just because I cannot recall a brand at the time you ask me a question does not mean I do not trust the brand – it means the brand is not as salient and therefore statistically is less likely to be recalled. It is not statistically less likely to be trusted.”
In a third experiment looking into actual brand choices, Fifth Dimension found 42% of brand choices made across a range of products in banking, telecommunications and insurance were with brands that were never originally in a consumer’s consideration set. Instead, these brands were introduced to the consumer during their path to purchase.
“Imagine you have spent all of your marketing efforts trying to build your brand trust scores only to lose the customer to a brand they have never heard of before. How did an unknown brand beat my trusted brand?” Spooner continued.
“You can spend all the money you want trying to build your brand trust scores and getting respondents to answer a question on who they trust, but you are only collecting useless information that has no relationship to your customer spend or potential future sales.”