Open-end text elicits free consumer responses that offer rich insights and complement other quantitative data.

  • Brand Tracking: Thinking of value-for-money, which soap brands comes to your mind?
  • Satisfaction Studies: Why did you rate dissatisfied?
  • Post Shopping survey: Tells us about your shopping experience?
  • Ad-testing: What was your overall impression of the ad?
  • HR Appraisal: What could we do better?

Infopickle facilitates grouping of these text responses into interesting patterns and categories. Infopickle can help in verbatim coding and analyses of open-end text data. Connect with us.


Theme Codebook


Contextual Validation


Insight Locator

Self Service Effectiveness

Theme Codebook

Infopickle scans all the text responses to look for common recurrent themes. It uncovers and automatically catalogues 100+ themes. Users can validate emergent themes, merge with other themes or further break down into sub-themes.

Understanding Themes

Themes are about consumers and how they interact with the product in terms of perceptions, usage, and attitude. For example, how do consumers experience ‘Credit Card’ in terms of “Fees & Offer” and “Issues with Credit Card Statement”. Similarly, keywords such as superlatives and synonyms around a given theme, is automatically grouped. For example, Internet or Online or On-line. Similarly, Bank or Banks or Banking.

Connect with us.
Prioritising Theme

Themes are important if they can be integrated with business solution. Colour coding for sentiment and keywords helps to filter important themes. “Easy to navigate” can be separated from “Not easy to navigate”. In addition, theme can be prioritised based on the frequency of occurrence or how they combine with other themes.

Contextual Validation

Simple themes such as ‘Any Mentions’ of brand names, characters, celebrities, ingredients, claims, benefits do not need validation. Complex themes require additional validation. For example, Is ‘safe website” same as ‘secure website’? Similarly, ‘Bought Netflix with card’, ‘Bought Netflix with Citi credit card’ and ‘Bought Netflix are same’? Infopickle provides three outputs that validates definition used for themes: Contextual Quotes, Consumer Quotes and Unmatched Sentences.

Operating Effectiveness
Contextual Quotes

Contextual Quotes provide an easy read of consumer verbatims by themes. Contextual quote retrieves representative sentences to validate definition of theme. This acts as a true measure of accuracy. Contextual Quote around ‘Website Acclimatisation/Getting used to’ could assure or negate soundness of the theme.

  • Although the new layout took a bit of getting used to, it has a better look to it compared to the old setup.
  • It took me a few minutes to get used to the system (change is never easy) now I find it quick and simple.
  • Slowly getting used to it, question longer term if single factor authentication is enough.
Consumer Quotes

Contextual quotes refers to sentences, but sometimes that may not be sufficient. Researchers and analysts also need to know additional context such as preceding or antecedent answers.

  • Consumer Quote for ‘High Rental’: ‘TV rental amount is high. Rental amount is INR 767 per month. It should be less.’
  • Users may want to know additiontal details such as ‘What is the perceived current amount and how much amount reduction is expected’. Consumer quotes is exhaustive way to validate theme definition.
Unmatched Sentences

At the end of analyses cycle, mutually exclusive and collectively exhaustive themes extract most from data. Unmatched sentences are leftover sentences. Researchers and analyst can decide to club unimportant sentences as “others” or club with existing themes. Large number of unmatched sentences indicates that list of themes is not collectively exhaustive.

  • Non-Theme specific Sentences: ‘Nothing specific, ‘No comments’, ‘No issues’
  • Theme specific sentences: ‘Good approach to digital technology, but interest rates could be improved’, ‘I think the site is a bit clunky’, ‘The process suits me and I am now in a routine in which I can quickly and efficiently do what I need to do’
Marketing Effectiveness

Insight Locator

Theme becomes an insight when a researcher or analyst deems it to be significant. Insight is reading between the lines (thousands of words in this case). Insight connects up multiple themes across multiple scenarios. It improves understanding about products or consumers. Infopickle facilitates uncovering of relationship that that may go unnoticed with naked eye. Uncovering of relationship requires searching and retrieving different combination of text themes that may exist in data. Insight Locator is based on 3 key levers that facilitates combination of text themes.

Search Query

Infopickle uses a simple, but very powerful search engine. Connecting and defining the relationship between search terms is done in 3 simple ways: EITHER/OR, AND, NOT. When searching for open ends, search can be used to either narrow or broaden definition. For example, Insight Locator helps detect complex sentences such as “I like parle biscuit OR britannia biscuits AND I like oreo chocolate but NOT hot cocoa.”

Theme Segmentation

Insight emerges from combining deep consumer understanding and personal intuition. This requires combining basic themes to higher order concepts. This is also called Netting but done at multiple hierarchical levels. Multiple nets can be created by user to determine best conceptual fit with minimum frequency count. For example, Infopickle can combine basic sentence to higher order themes:

  • Net 1: Shopping is fun ~ Shopping is Fun, I Combine shopping with eating out.
  • Net 2: Value for money shopping ~ I try to get the best buys while shopping, you can save a lot of money by comparing prices.
  • Net 3: Anti-shopping ~ Shopping is bad for your budget; I don’t care about shopping.
  • Net4: Net 1 and Net 2
  • Net 5: Net 1 and Net 2 but not Net 3
Theme Sequencing

Key events in a sequence of words – continuation or breakages are important to record. Some of them, will be more frequent activity while others could be rare activity. This is relevant in certain scenarios such as customer satisfaction or ad-testing, when respondent answers refer to preceding response. For example, good refers to TV series, while quality refers to signal and NOT TV series. “TV services are good but they should improve signal quality”