Member Since: 26th Jul 2016
Alyona Medelyan, PhD, is an expert in Natural Language Processing and Machine Learning. Alyona has extensive experience working with customer feedback data, such as surveys, social media data, call centre logs, and public forums. She is also CEO and co-founder of customer insight company Thematic.
13th Jul 2018
The score I give depends on my particular mood at the time of asking. I've given a range of scores to Confluence, our internal knowledge base, when being asked in-session. The focus shouldn't be on collecting scores but the actual actual customer feedback in their own words. What are the important improvements Confluence can do to make my experience better and why? Answers to such questions will provide more insight than scores in my opinion.
CEO at Thematic (http://www.getthematic.com)
5th Jul 2017
With regards to NPS, I'd like to add that customers are likely to mention critical touchpoints and their views on how to measure them in response to the open-ended "Why did you give us this score", and you can use NPS data to measure the effectiveness of these touchpoints. Here is an article on this topic for those who are interested: http://www.getthematic.com/post/net-promoter-score-analysis/
27th Oct 2016
Hi Simon and Steven! I found both your comments to Seth's interview very useful. In my experience, customers rarely mention emotive words, but rather explain what happened during their interaction with a company. Lana solves this problem by probing customers to pinpoint the specific emotions, and for those who can express them, she can capture them using her tools. I propose an alternative way of addressing this: Sticking to detecting themes that aren't exactly emotional, but which translate to human behaviours that trigger emotions in customers. Here is a link to my post describing this in more detail, if you are interested: https://www.linkedin.com/pulse/missing-link-emotional-analysis-customer-...