The 10th Bi-annual Conference of Market Research Industry & 20th Anniversary Ceremony of CMRA was held in Beijing Landmark Towers on November 1-2. Almost 300 people including famous economists, marketing experts, scholars from universities and scientific institutions and business representatives gathered together to communicate with each other on the topic of “Beyond Boards · Co-creation”.
InsightWorks was given two awards at the evening party to commend our years of contribution to CMRA and great innovation & entrepreneurship.
Once again we’re tremendously encouraged and honored by the recognition from peers.
On the scene
InsightWorks CEO Bill Zuo gave a speech titled “Truly know your consumers by AI + targeted survey”. He explained how to use artificial intelligence and behavioral targeting to get through data barriers between online & offline, behavior & attitude.
- Let’s start from AI
From my standpoint, the applications which AI can actually come into play in market research includes:
① Voice Recognition: automatically turn voice memo, say from focus group, into text document.
② Image Recognition: mass image data from social media can be used to identify brands, face recognition in malls, facial coding, etc.
③ NLP: draw a valuable conclusion from user reviews on e-Commerce website or social media
- The evolution of panel: mobile & behavior tagging
The collection and integration of different kinds of data have already been accomplished with our mobile panel. Mobile panel has a lot of undoubtable advantages including prompt response, more flexible time schedule, the ability to upload images, videos and location information, etc.
With that being said, mobile panel is not omnipotent at all. We still face the problems of surveys lasting too long, content too complicated, topic irrelevant or boring, screening rate too high... So the next generation of panel must solve all these problems: no screening, lasting less than 5 minutes, an interesting topic, and that’s what we call Tag Panel. We exchange fluid data through IMEI, phone number and cookie to provide more detailed user profile and behavior data.
- The ultimate solution based on the integration of big and small data
When we evolve to the Tag Panel mentioned above, NLP technique can acquire numerous authentic text data intelligently and behavior targeting survey will make up for big data’s lack of meticulousness.
Generally speaking, the demands for survey data may decline in the future but it will still play am important role in market research, while the integration of multi-source data is the general trend. All we can do is to adapt to the challenge and seize the momentum.
