Talk2Me: questfox Digital Research Voice Assistant goes onlinePosted: 15. August 2017
We are happy to announce, that a whole new category of breathtaking question types is out now and available in questfox.
In the CREATE section you will now a find a new category called “User Media”.
Besides the function of a Photo upload we have fantastic new question sets ready to be used.
questfox Digital Research Voice Assistant:
Automation in market research with speech2text technologies
Talk with devices becomes normal
Alexa, Cortana, and Ok, Google are currently the standard voice commands to ask for a digital assistant for help. Now a digital voice assistant is available for the first time in support of a market research interviews.
The team of pangea labs is dedicated passionately to the task of developing mobile methods for market research. Our software-as-a-service survey solution questfox now progressed to a milestone in the development of the survey interaction. It is with great pleasure that we announce today the implementation of a significant step in the development towards mobile methods. We present the questfox voice assistant Talk2Me: a fully integrated speech-to-text solution for verbal survey responses in the context of digital interviews.
MP3 to written text
A customer approached us with the idea to integrate original voice comments from customers in into market research using voice recording. So far not a new idea. As there are already similar methods available, we did not only wanted to meet this requirement, but converted the desire directly into a technical context of digital marketing research under the premise of increased automation. In the age of automation of market research, we did not only want to work with spoken text fragments work as MP3 file, but offer an integrated solution that solves the problem of analyzing large amounts of text. To deliver these verbal answers directly to an evaluation, we have in addition to the voice recorder technology implemented a speech-to-text feature, which transforms the spoken language into written text.
Three new audio question types for market research available in questfox
Total questfox now has three new types of questions for language market research:
|questfox question type||Functionality|
|Audio Recorder:The Audio Recorder records spoken language and allows later to listen to the recorded language. This type of question is suitable for sound bites from consumers. In a classical research process, this type of question would have to be transcribed later.|
|Talk2Me: This new type of question translates the spoken language into text on the fly. The user can then adjust and correct the text before continuing. The recording can be triggered multiple times within this issue in order to give longer thought into sections on record.|
|Background Speech2text: The first tests with live converting spoken language into written text has frightened some users. To avoid this effect, a variant has been implemented in questfox. With this question type the transformation takes place in the background. Here, the interviewee does not see the transformation of his spoken language while answering.|
All new questfox question types, the original audio format, and therefore are available for later purposes, such as integration into a client presentation.
questfox opens up new worlds of interaction in market research. We do not yet know all the consequences, but expect massive changes in the interaction of research interviews.
Next step: sentiment analysis via API
The text data is ready for analysis in questfox with the existing evaluation tools. Creation of tagclouds etc. are right away possible. The next step is to integrate an API for semantic analysis of the text data. The intensive cooperation with the Cologne expert for textual analysis, insius, it will be possible to analyze the spoken texts later in form of semantic analysis.
Thus, we offer a total new way of data collection and automated analysis methods for the market research industry.
Market research internationally and soon globally
This seemingly small idea could lead to an increase of the use of consumer opinions. The written statement of the open text responses has always been a question of education of the interviewee and his ability and willingness to use a keyboard to express himself. In the future there will be fewer socio-demographic differences in the degree of filling of open answers. The digital voice assistant could even interview audiences who do not feel very comfortable in the written comments. They can now give their opinion openly and perhaps more spontaneously.
This new development will be perhaps more important, when we look in the direction of less developed countries with a low degree of literacy. With the digital language assistant in market research, it will be possible to interview people, who cannot read or write. The new technology will lead to larger target groups can be developed that were not accessible to market research so far. This is very advantageous for the questfox Digital Research Voice Assistant automatically detects the spoken language and transmits the text. Interviewing in English, German, Spanish, Italian and French are not a problem today. Other languages will follow.
questfox Digital Research Voice Assistant Process Overview of
Requirements for the Digital Research questfox Voice Assistant
When searching for device-agnostic solutions, we strive to open the technology to as many people as possible without any additional effort needed. We have therefore targeted technologies that already allow interaction with language. Not all even modern browsers will be able to make speech recognition available in the context of a browser interaction. The questfox digital research voice assistant can therefore only be used when the browser technology supports a voice interaction. At the moment these are the browser from Google (Chrome), Firefox, Microsoft Edge, and Opera. To our great regret Apple Safari is not yet allowing such technology. With this technology setup it should be possible to carry out mobile speech questionnaires without installing an app.
Support of mobile Apps
We work in the next step in implementing the browser software into the questfox mobile research app, in which we ask the user’s permission for the acquisition of speech, so to use the microphone before interacting.
Disadvantages and fears
The first voices are raised, are afraid of digital assistants and suggest a permanent monitoring by activated microphones. We take these concerns very seriously and want to grant full control of their own responses collected through a survey questfox, the user. That is why with Talk2Me a two stage process available, where the user can first listen to his speech again and can make corrections even in the converted text form before he submits it. We will explore in usability testing whether this variant is acceptance.
We believe to be the first to offer this in German. Other languages are possible. It is then just a matter of the organization of the interview, if multiple languages are to be mixed in a study.
The costs are negligible compared to the currently performed manually transcribing voice files almost. The cloud-based model takes into account the time actually used, we provide our customer. This can vary greatly depending on the project. To avoid the first fear of contact, we plan to offer flat-rate packages that vary depending on the size, number of languages and requirements used in the evaluation.
For existing questfox users the access is for free for the moment. We expect you to make a fair use of the new functions.