How we source data for the world's largest purchase journey study.
Survey data
Powering gfknewron Consumer is survey data sourced exclusively by GfK from over 30 online panel partners. Fresh samples of around 500,000 consumers are screened globally each quarter.
We approach a nationally representative sample based on age, gender, working status and region. We screen within this to identify 16+ adult consumers who have purchased a qualifying product in the previous quarter. Survey invites are sent in batches to control for early response bias.
The 12-minute survey is designed mobile-first, making it easier to complete on any screen. 40% of surveys are completed on a mobile device.
Sample quality controls include comprehensive panel duplication, robot, speeder
and over-claim checks with any "suspect" interviews deleted from the sample.
A quarterly sample of around 500 to 2,000 recent buyers are found per product/country. These
are in part a function of purchase frequencies, so there may be some seasonal variation.
Data is calibrated to align with GfK POS data on key brands, price band and sales channel.
- Shoppers buying from retailers not covered by GfK’s panel market are not calibrated to PoS data, but merely sized to collectively represent the part of the market outside coverage (as estimated by local GfK sector exports annually).
- Shoppers buying from retailers covered within GfK’s panel market are additionally weighted and sized to increase alignment with PoS data on key brands, price band and sales channel (online/traditional). Data is NOT weighted on retailer shares. Any retailer shares reported purely represent the fall-out of the survey data.
A series of survey vs PoS data checks are also incorporated in both questionnaire scripting and analysis stages.
Review data
Where indicated, platform insights can also be sourced from passive scraping of online review data. This data is sourced using Etilize (part of the GfK family), which crawls the websites of large, trusted retailers, gathering consumer reviews for different tech and durable product categories.
Very high volumes of reviews are collected from the biggest retailers in each market, between tens of thousands and hundreds of thousands of reviews per product category/market.
Category-level, in-depth human coding is used to identify topics mentioned in reviews, with the ability to identify multiple topics and nuances within a single review.
For more information, speak to your local GfK Account Manager or contact GfK support.