Market research reports are based on numbers and most of them, they are either sponsored or the sample size is too small to make sense of them (which is why we at NextBigWhat are very hesitant about publishing these reports).
Few days back, Nielsen came up with a report on wallets and had high traction numbers for a particular company (Freecharge). Competitor, Mobikwik exposes the research flaws :
Couple of questions raised by Mobikwik team:
- The survey claims facts about the mobile wallet industry that caters to more than 100 million Indians by tracking just 6000 smartphone users!!! – How can a SAMPLE SIZE OF 6000 be even statistically relevant for a burgeoning country like ours? [source]
- Even the timeframe of this is questionable. For this particular comparison, Nielsen took timeframe for the data point between May – July 2015 while Freecharge wallet was officially launched in Sep’15. Isn’t it baffling that such a so called research institution will make this blunder?
Finally, Freecharge misled media by not disclosing that this was a paid custom report (and not a neutral third party research) where they controlled how the research was done. This is admitted by Nielsen themselves.
Update from Freecharge
- It’s syndicated track data and not custom data. Clear as sunlight. This data is from the Nielsen Smartphone Android Panel that covers usage of standalone payment apps and not payments as a service.
- But, it is no surprise that some folks are indeed trying to influence outcomes through custom reports (as inadvertently? Revealed in MobiKwik’s email)
- Six thousand is a very robust sample size by any standard. Somebody needs to revisit their lessons on probability and statistics. A primer on statistics is on the way.. Hope it helps! [via]