At the Insight Innovation Exchange conference in Atlanta yesterday, Jon Sadow of Google Consumer Surveys discussed the complimentary role of surveys in a research industry proliferating with new technologies and methodologies. With Big Data, text analytics, sentiment analysis, mobile research, video ethnography, eye-tracking, and more, where do surveys fit in?
Previously, Google was advocating the benefits of “less is more,” with shorter surveys designed with the user experience of the respondent firmly in mind. With microsurveys, survey research is faster, cheaper, more scalable, and more suited to mobile.
Now Jon says that cutting survey length alone just won’t cut it. “With these capabilities in place, we should be asking what more we can do,” Jon said. “Some of you already are. How are innovators using survey data?”
Survey Data As Variables For Big Data
Big Data is often missing key attitudinal information that surveys can best provide. Google eats its own dog food here. Google uses survey data to improve DoubleClick demographic algorithms, processing millions of data points to refine its models and defining behavioral categories using survey data. Additionally, microsurveys measure the user experience with Google ads (to improve relevance) and websites (as another input to improve search functionality).
Massive Survey Data Sets
Small surveys can produce big sample sizes:
- One client gathers “Consumer Rating annotations” using survey data to measure perceived value, customer-service quality and more.
- A high-growth transportation company gathered NPS data across dozens of metropolitan areas and was able to correlate the findings against its growth rates in those areas.
- A major telecommunication firm runs a monthly brand equity study on Google Consumer Studies, collecting almost 100,000 data points a month. The data is then analyzed against marketing, sales, and product data to measure the impact of various initiatives.
- A media measurement firm collects audience data at the level of the ZIP code, collecting viewing behavior, programming, volume and more across over 400 markets.
- A health-industry startup created the largest-ever pharmaceutical experience study, collecting 100,000 responses covering over a 100 prescription and over-the-counter medications.
Survey Data For Sophisticated Modeling
GCS clients are “using propensity and multivariate modeling to build election predictions and estimate future trends.”
Research firms are using “innovative design and question rotation models” to speed R&D research while reducing costs.
The Future of Survey Data
Jon concluded with a vision of the future, with:
- “Increased utility of Big Data, because of surveys.”
- “Researchers and analysts’ roles gradually converging.”
- “Global parity for survey research, driven by mobile.”
His final point: “What does the future hold? What will you come up with next?”