Vypr Data Robustness

Panel size & weekly actives

The Vypr panel size is as big as, or even bigger than, many long-established players in the market. It’s completely proprietary and unique to us. Total installs on the UK Vypr community across iOS and Android are just over 60,000. Weekly actives are just under 9,000 and our daily actives are just under 4,000. This gives us a broad & deep range for unique respondents, far more than is needed for a statistically significant effect to be measured. A sample size of 500 gives a 4-5 confidence interval at 95% confidence level in a UK-sized population.

Nat Rep

We deliver nationally representative steers by splitting the response profile into age/gender groups with a quota. The quotas are derived from the 2021 ONS Census information on UK population distribution. Each steer that is targeted at Nat Rep guarantees a fixed % of respondents in each of these quotas. The 12 individual quotas are;

This results in steers that have a very balanced response population every single time.

Demographics

We use exclusive single-answer demographic filtering for our public demographic profiles. A respondent can only give a single exclusive answer to join a group for further targeting. Therefore a respondent can’t belong to both “vegan” and “meat eater” groups. This ensures that targeted respondents can't game the system to maximise their inclusion in surveys they're not personally eligible for.

Respondent authenticity

To ensure the Vypr community is made up of authentic respondents who are who they say they are, we have a number of functions in place. Not only does this prevent bots from participating on Vypr, it also ensures that our respondents are not out to game the system.

For example, the Sign-Up Process for individuals requires:- Name, Age, Gender, UK Postcode and unique email addresses (no two accounts can share an email, either registered or PayPal). We process all PayPal payments manually, reviewing respondent account activity at the time of payment, to ensure industry-leading integrity levels. The mobile application only allows one account creation per device install which is a key deterrent for any bots or click farms. The app is only available on UK app stores. We also run “Captcha” steers on a regular basis which require a factual answer. If these are failed then the account is banned. And finally the monetary reward is very low; we offer minimum 0.5p and maximum 5p per response. This is a reasonable deterrent to anyone interested in gaming the system, as the effort to reward ratio is really quite high.

Customer experience

We have numerous long-standing customers who have tested the robustness of Vypr over the years. A quote from one of our larger clients Muller is below;

“When at Müller, I tested the reliability of Vypr by asking the same question (“Would you buy this product?”) on a Yes / No steer to 500 consumers on three products (we kept it to milk, milkshake, and cream) twice a day at exactly the same time (morning and evening) for seven days. We kept everything the same only varying the day. The results were all within the margin of error at 99% confidence. I also tested the relative results of a range of 30 SKUs in four categories against Rate of Sale and Penetration, and found a correlation of between 0.7 & 0.9 respectively. All of these results gave us a high degree of confidence that the Vypr tool was a strong tool in understanding consumer motivation.”

Bill Harte, former Innovation Controller at Muller Milk & Ingredients