Usually, customer feedback forms include an open question based on like/dislike: “What did / didn’t you like?”.

R3MSCORE emotional analysis method consists in gathering 3 spontaneous words: “Following your experience with the company/brand, what are the 3 words that come to mind?”

In both cases, we collect spontaneous language from the customer. However, the R3MSCORE question about the 3 spontaneous words provides a way superior added value, in terms of data resources, adaptation to any public and power of analysis.

Obtaining valuable and various answers

This question about the 3 spontaneous words makes it possible for us to approach Daniel Kahneman’s “System 1” – the field of emotions and automatic reflexes (instincts or learning).

Open questions such as like/dislike encourages more the respondent to think and justify their appreciation, which gives them a more rational stance (“System 2”).

In addition, the open question like/dislike encourages the respondent to adopt an assessing stance based on appreciation, meanwhile the R3MSCORE question isn’t about inductive reasoning. The latter gives more space to the expression of larger registers, enabling the updating of elements that aren’t rationally bound to pleasure or belong to another register:

• iconic signs,
• category anchoring,
• territory anchoring,
• “craving” anchoring,
• metaphors…

In a study led by the Paul Bocuse Institute research center, we noticed that with R3MSCORE, respondents were expressing themselves in a more various register. In this specific case, they expressed elements related to visual and aesthetic dimensions of meals that were outside of like/dislike range, yet very impactful emotionally.

In fact, traditional open questions weren’t making a distinction between some of the tested meals while their R3MSCORE equivalents presented significant differences.

Adapting to any public

The “3 spontaneous words” principle makes it easier for some publics to express themselves when they struggle to build full sentences in order to verbalize what they felt. Like so, we were able to prove that this method was particularly adapted to children, who can sometimes use swear words. This is more difficult with a like/dislike question.

Weighting discourse with its emotional intensity

With a like/dislike open question, verbatims are simply analyzed regarding their occurrences. In the meantime, the main goal of the R3MSCORE method is to weight spontaneous language based on its “emotional intensity” – how each word reveals the respondent’s emotional activation.

This way, the R3MSCORE algorithm enables us to make a distinction between quotes with the same number of occurrences but being of different emotional intensities.

For example, on a 100-respondent basis, a word expressed by a small number of respondents (3 or 4) but presenting a strong emotional intensity will appear as a weak signal. It remains meaningful in a R3MSCORE restitution, whereas it would have been ignored by an analysis method solely based on occurrences.

Talking from experience, product experts − noses for the perfume industry, culinary chefs, sensorial experts for the food industry, customer experience directors – when confronted with R3MSCORE results, have always approved the relevance of highlighted weak signals.

Dealing with result analysis in depth and helping the decision-making process

The R3MSCORE algorithm generates analysis possibilities more valuable than those of like/dislike open questions. Why? It uses the calculation of an emotional activation score for every individual that can be cross-referenced with other variables specific to the profile.

This R3M score becomes a major KPI which completes usual indicators (NPS, overall satisfaction…) and often appears to be more discriminant than these indicators.

A matrix using both R3MSCORE and rational measuring can also be used to define a MAGIX BOX. Some customers even chose to adopt a decision-making process completely based on the R3M score.

All of these arguments explain why some of our customers gave up on like/dislike open questions and turned to the “3 spontaneous words” question analyzed by R3MSCORE.