Mazaru have been helping clients understand their interactions with customers more effectively since 1994 – enabling them to tread that fine line between meeting expectations and managing resources, while ensuring the interaction is perceived as a success by all parties.
Their experience as consumer insight professionals has always been underpinned by the latest behavioural science and understanding of how people interact and communicate, but their new product, Mazaru Insights, infuses that qualitative understanding with the latest AI-enhanced data analysis — to generate reliable and actionable insight at speed.
As Mazaru CEO, Fran Fish, explained, the way people want to relate to brands emotionally depends on the service they’re providing. “In some cases we want to feel close to the people who provide services to us, and in other cases we just don’t care — we just want our electricity to work. We don’t need to feel up close and personal with our local water supplier every day.
“But when something goes wrong, or when we pick up the phone, we want to speak to a person”
“Or someone might need more support, or get frustrated and call because they don’t want to read a printed welcome pack, or don’t trust what’s written on a website.”
Analysing the total conversation
By using AI to analyse what their customers are saying, Mazaru help their clients to understand the effectiveness of each element of the conversation — digital, written and ‘chat’ content — from different perspectives, in totality and at speed. These perspectives include like sentiment, readability, effort, jargon, spelling/punctuation/grammar, confidence, and rapport.
Scoring these elements enables them to really zoom in on areas for improvement which might not be obvious, such as the confidence element. As Fish explains, a brand might be using lots of hedge words which create doubt in consumers minds, like ‘“we should be able to fix that” or “we usually deliver in 10 days”. By isolating this factor in the conversation this can be examined and tweaked — helping brands to improve the score by consciously eliminating these wobbly words which sow seeds of doubt, and use more active and positive statements of intent, through retraining and re-writing contact centre scripts.
The machine learning element helps to surface these small differences which feed into big trends, and identify the areas where improvements can
have the greatest impact, while still reflecting the brand’s intended tone of voice and communication style.
Ranking the competition
In addition to improving on their own performance internally, customers can benchmark against others in the same industries as a point of comparison, whether the bar is high or low. So it may be a case of, our sentiment score is low, but, compared with other utility providers we’re actually doing pretty well.
These comparisons and the machine learning which generates them will improve as the dataset grows, and Mazaru’s recent London launch event is focused on identifying 100 pioneering customers to get on board and thoroughly explore the potential.
As Fish concludes,
“They could be our existing customers, or new customers, who come with us as we continue to build the application, knowing it’s only going to add to the value”
