Our strategy involves two components, incentivizing users via prizes donated by companies and increasing the number of times a user’s name is entered into the lottery for the prizes. We plan to reach out to local Cambridge area businesses like Tatte, Urban Outfitters, Felipes, and Daedalus and incentivize them to donate $50 – $100 gift cards by advertising their companies on our survey page. Thus, the larger the value of the prizes they donate, the more likely people are to take the survey and see the advertisements, which benefits the advertisers and helps satisfy our aims. A similar approach was taken by Groove, a customer service software company with good results, so we want to apply this approach to our survey [Groove].
Our customer base is women 18 – 35 who are Harvard students or just live in the Cambridge area. There are roughly 31 million women in this age range in the US [Census], and by taking a sample of women in Cambridge, we are estimating that Cambridge is a fairly representative sample of the US population. Cambridge’s racial demographics mirror those of the United States as a whole, though the median income in Cambridge is higher than the national average [Cambridge Census], so our assumption is reasonably valid, at least for a short term experiment. Thus, we will use a sample of women in Cambridge to make conclusions about what women all over the country may like, though additional studies may be necessary in order to validate that claim. In order to have 95% confidence in our results, and to have a confidence interval of +/- 2.5%, we need 1,537 individuals to take our survey [Sample Size Calculator].
We plan on targeting our user base by using Facebook and Pinterest Ads. Both platforms allow us to choose the users who will see the post, and will let us filter users by geographic location and age, which will let us access our target demographic. We want to use Facebook because it is a general platform that will give us a broad range of users, and we want to use Pinterest in order to capture a subset of people who are more likely to be interested in fashion and design. Both options are also fairly inexpensive, and will work within our $100 budget.
We’ll encourage immediate participation in the survey with a tiered raffle system based on timeliness (see chart below). We’re increasing one’s motivation to participate precisely at the moment when one has the highest ability to do so — they’re already on the web with the survey link right in front of them. Better take this now and reap the rewards, rather than do it later or possibly forget. Moreover, participants will be able to earn an additional five entries for every person that they refer to take the survey. This kind of peer-to-peer or referral marketing worked wonders for Venmo and Uber in their early stages [Pando]. In 2013, each successful referral netted a Venmo user $5 dollars, allowing the platform to spread incredibly fast among friend groups.
*In the survey’s final hours, we can boost the number of entries awarded to participants.
[Sample Size Calculator] http://www.surveysystem.com/sscalc.htm#one
[Cambridge Census] http://www.cambridgema.gov/CDD/factsandmaps/demographicfaq