Our client is a US-based company that conducts market research based on a set of capabilities and tools provided by Qualtrics for conducting marketing research and deriving targeted answers, backed by various scientific algorithms. It uses different technologies of Artificial Learning and Machine Learning to gain customer insights on time and well in budget. Deep insights into the thoughts, sentiments, cognitive, and emotional processes of the consumer across different platforms are utilized to find hidden meanings in segments and documents.
Since market research is a pretty vast field of study, a proper understanding of the industry’s social, cultural, and economic frameworks is essential for the overall success. They wanted to know their targeted customers in the areas they are dealing with, and what their competitors are doing. The technologies they were using with their tools, needed more convenience, different interfaces, storage of responses, generating reports, preferences for permissions, among others. It was getting a bit overwhelming for them because they were not able to identify and separate complex challenges.
They were facing problems in developing a process for analyzing data about a market, brand, product, or service, and the ability to help understand their customers. Their expectations, identities, and wants can only be analyzed through efficient research. They also found it challenging to establish an open-ended line of communication with their customers. Hence, they needed to understand the customer's needs and then tailor their business according to them.
They needed customers who can share genuine opinions for the brands with which they were dealing or involved. Finding honest opinions is an intense field where feedback can be interpreted in various ways through meaning and emotional engagement. And yet, these sources had to be less time-consuming, customer-centric, and dynamic in response gathering. They needed to conduct market research where they targeted their customers and gained responses on the basis of audio, video, ranking and re-ranking of brands and products on different scales. In a nutshell, they needed to determine the territories for translating the customer surveys under multiple interfaces. They also wanted to get the best possible results on video-feedback, video insight vendors, among other things, by connecting with respondents.
Hence, they needed better segmentation and representation of the chosen locations and audience to understand them better. They were seeking a system for agile market research that enables them to collect data from any audience and integrate open-ends into existing surveys. They wanted to set up a structure of customer-centric studies and through survey responders. They wanted to have all the options available for conducting such research, right from the fields with an array of tools to capture, transcribe, translate, and analyze.
Based on their requirements, our solutions mainly focused on integrating an innovative multi-channel journey that consisted of designing dynamic interfaces for each stage. Hence, we decided to develop a mobile web interface that would have integration with all the advanced Machine Learning and NLP techniques. To capture the consumer insights accurately, we started the journey to integrate everything from item selection, familiarity with brands for analysis, capturing videos, audio and image interaction, and many other unique features. We wanted to develop a system where quantitative agility can be achieved by different types of behavior for asking questions, setting random responses, and deriving an action-based path on such behavior.
We integrated the default and custom survey distribution channels by detecting device capabilities and special behaviors to create a responsive design. An interface that is accessible across any platform was enhanced by adopting human voice and data to collect and capture by storing, transcribing, analyzing, and reporting. With such tools, we also enabled our client to carry out their research activities for large consumer groups. A single intuitive platform for distributing several surveys enabled a systematic and accessible way to manage each survey related operation.
Reporting is another essential task after survey distribution that needs a proper analysis of research projects and the responses received. Open-ended text responses are one of the main challenges faced by the client that needed to be analyzed on text-basis. Hence, we added different preferences in the fields among interfaces with permissions to choose from video, text, audio, etc. For this, we also utilized the text analysis tool that makes the cumbersome task of analyzing open texts a lot simpler. With NLP, it can analyze sentiments, emotions, and behavior behind each response and comment on the document. The analysis is automatically compared with trending feelings and topics to get accuracy in the insights.
Our future plans include integrating multiple languages for text-analysis through NLP, so that we can reach a wider global audience. Apart from that, we plan to develop a client dashboard where admins can track each aspect of analysis, research, and reporting from a single intuitive interface. We also plan to integrate report generation features that cater to each stage of the journey right from online item selection to interaction with customers through images, audio, and videos. And additional feature we plan to develop is the auto-tagging of brands and other attributes for identification through NLP and ML techniques.
Our approach to solving our client’s issues aimed to focus on providing a flexible service model that is capable of inculcating the right support levels to enhance the maturity of their market research. It started by designing a curated content platform for research analysis workflow that is seamlessly automated. We wanted to make this incredibly simplistic and flexible by personalizing the experience, utilizing advanced branching methods, display and skip logic, piping, A/B tests, among others. Hence, for embedding such integrations, we used Qualtrics - a well-known platform for assisting clients in knowing which products or services they need to build by delving into deep learning methods of Natural Language Processing (NLP).
With advanced Machine Learning and Artificial Intelligence technologies, we leveraged the highest potential to check surveys and screens for the quality of responses. Our system helped us approach the behavior and feedback received in responses by intervening with the respondents before they churn. The text-based analysis was used to analyze and gather sentiment and emotion scores. And the audio responses which are stored were analyzed by Google Transcribe that makes it easier to convert speech to texts. With various in-built models, we helped our client optimize the screen recognition with video transcription, indexing, and subtitling through ML.
Demographic and psychographic information is the current way of how companies extract information from their target audience and communicate with them. So, this was our base to develop a solution based on technologies for building interfaces, and mobile web. Our team of developers that consisted of 6 members, worked on different techniques ranging from JavaScript, Laravel, and Python for writing seamless and single codes. Apart from ML, NLP, and Google Transcribe. We utilized cloud functions of serverless architecture that integrated programming interfaces (APIs) for hosting the cloud storage data. We ensured that we chose the technologies in which we have gained years of experience and expertise.
With an intuitive video insight platform, our client started availing an extremely vital benefit with the in-built editing tools. These tools enabled them to subtitle responses in any language and create snippets by overlaying a common one and made research surveys more engaging. Adding new tools in search filters like thematic and sentiment analysis, helped them explore responses based on keyword variables and word clouds. Analyzing the real emotions behind a statement or response automatically assists them in identifying essential and common themes that respondents prefer. A corporate partner of our client who avails its services for getting customer insights recently told them that “Transcribing and translating so many videos through sentiments helped us gain insight through several brands worldwide. It has really increased efficiency and delivered maximum impact.”
Our client also saw great results in reporting and analyzing these results in the form of actionable and evidence-based data insights. Market research services provided to industries like pharma and health, energy, chemicals, and mining saw a significant jump of around 34% - 38% in their sales generated through such analysis. Big Data analytics have increased their efforts by almost 43% towards online market research surveys and is yielding great responses.