Coding and analyzing open-end responses have challenged marketing researchers for decades. AI text analytics automates the chore of analyzing text, saving researchers time and money. Here’s how to pick the right AI text analytics partner for your needs.
Artificial Intelligence seems to have burst onto the scene in the past few months. Generative AI is evolving rapidly, building on and going beyond previous AI technologies (e.g., text analytics, Natural Language Processing (NLP), machine learning, and large data models). Additionally, many products today use various related technologies, and many more firms say they are incorporating AI into their platforms and products.
In market research, while there is no replacement for manual coding in some cases, AI is getting very good at analyzing open-end data. There is no substitute for the speed and quality with which AI makes sense of open-end responses. However, businesses are currently buried under a veritable tsunami of text data from social media, customer reviews, and open-end answers to survey questions, and the problem is only growing. Many businesses simply ignore that goldmine of insights because they can’t invest the time or money in analysis.
AI can change that.
Open-end question responses and unstructured text allow respondents to freely express what they want you to know. Because open-end responses are more powerful than close-end question data, they can lead to new and unexpected insights. The ease of use and productivity advances gained by using AI text analytics enables you to rely more on open-end responses and unstructured text. In addition to getting better insights, this can help improve your surveys overall.
For example, you can replace several close-end questions with one well-constructed open-end question, reducing survey length. Because you let respondents tell you what they want to share up front, you reduce respondents’ fatigue and frustration with going through a lengthy questionnaire waiting to be asked the critical question for them. Since you can minimize survey incompletes and break-offs, you save time cleaning them out of the data when the field is complete. Overall, with AI text analytics, you improve survey response rates, a goal of every researcher.
How to choose the right AI product for you
With all the chatter about AI, it is easy to get confused about choosing the right product for market research. But only a few AI-based analysis products work well for marketers. Here’s what to look for in an AI text analytics partner:
Experience
With many companies claiming to incorporate AI into their platforms, the truth is that AI has been around for a long time. Therefore, look for a partner with deep experience in applying this technology in the market research industry. A partner who has experience with many companies’ use cases in different sectors will be able to guide you to the correct AI application for your dataset and research projects.
Ascribe has provided verbatim coding solutions to the MR industry for 20+ years. With numerous large customers, including many leading market research providers, they have thousands of customer use cases, giving Ascribe the experience you can rely on for your project.
Customer service, support, and training
You are new to AI, and AI is rapidly evolving. You will have questions, and you will need your partner’s support. You want to be sure someone is there to help and advise what tools to use to help your specific project and the best way to use them. Ascribe has an experienced staff of experts to provide high-quality and personalized training and onboarding to their products and services. Additionally, they have 24/7/365 global customer support (via online, phone, and email).
What’s the output? Themes Vs. Topics
Two types of results return from AI text analytics: topics and themes. Some AI-based analysis solutions produce a list of topics, which are one-word responses that result from counting their frequency in the text. Topics need a lot of cleaning and manipulation to understand the big ideas.
Other solutions deliver themes. Themes, conversely, put the topics into context and look more like sentence segments or complete thoughts. They provide broad ideas of what respondents said in response to various questions. Themes are better than topics in providing critical insights from the text, meaning you can act on them more quickly.
Here is an example to illustrate the difference between themes and topics. Analyzing a restaurant’s CX survey might result in the topic “food.” An analysis returning themes, conversely, would result in “the food is very good,” immediately letting you know that customers are satisfied with your food quality.
Ascribe’s latest innovation built into its solutions is Theme Extractor, which delivers results with meaningful themes. Additionally, Ascribe can apply sentiment analysis to its results, letting you quickly determine what’s working and where to look for what’s not working.
A range of products for any use case
One AI-based product does not fit every company or every research project. Market research has a broad range of users and usage occasions, so we need various product choices. Do you need to analyze 100,000 survey responses quickly? Or do you have 2,000 customer reviews each quarter you are tracking?
Or do you need help with a survey of 500 consumers with two open-end questions? Do the results have to be highly accurate, or are directional results sufficient in exchange for a very quick turnaround? Or would you really just prefer that an expert help you decide which approach is best, and then run the text analytics and coding analysis for you? Ascribe’s solutions and Services experts help you in the way that fits your situation best.
Ascribe’s Coder is a verbatim coding management system that leverages AI to improve productivity while still allowing coders to maintain control of the results. Coder is a valuable project and team management tool for coding operations, as tasks can be distributed to individual coders to deliver the project more quickly. Coder also includes automatic translation and multi-lingual capabilities as well as AI Coder with Theme Extractor which instantly creates a theme-based codebook with nets and codes the responses. Your coding staff can then leverage the other features of Coder to edit and finalize the project.
CX Inspector is the product for you when you need quality insights quickly and want to analyze and visualize the results. The full-featured advanced text analytics solution can easily process large datasets of open and closed-ended questions and display results on an interactive and customizable dashboard. Click on results to see individual responses, apply sentiment analysis, create crosstabs, filter by variables, select your preferred charts to uncover the key insights, and easily export to share with your stakeholders.
The recently introduced CXI Go with Theme Extractor is Ascribe’s DIY text analytics platform, which is excellent if you are looking for the top themes from a dataset quickly. CXI Go is easy to use – load your file and charts immediately identify the top ideas, then you can easily export your results or fine-tune them. CXI Go instantly delivers top ideas from open ends with minimal effort at a lower price.
Don’t have the time and motivation to use Ascribe’s tool yourself? Ascribe also has a Services team of global experts who use Ascribe’s solutions to analyze your open-end comments. Their goal is to deliver projects first-time-right, on time, with excellence. They work with you to determine your goals and needs, and then produce the finished product quickly.
