In designing our engagement strategy, we decided to focus on a target population of women aged 15-65 who live in the most fashion-forward cities in America: NYC; LA; Seattle; Chicago; Miami; Austin; Dallas; Columbus, Ohio; and Washington D.C. The decision to target women was based off of the fact that the survey only shows women’s outfits. The decision to have a larger age range is also based on the survey – which showcases outfits on both younger and older women – and on the fact that consumers in the fashion industry span many generations. In deciding our target cities, we were faced with two questions: do we want to appeal to a niche population that understands and cares deeply about fashion? Or do we want to appeal to a more general audience? After some discussion, we assumed that targeting a general audience – with a much more diverse taste in fashion – would lead to inconclusive results. By focusing on a niche population that cares deeply about fashion, we would be more likely to obtain a concrete conclusion. As a result, we limited our target population to cities that extensive research has shown to be invested in fashion [1-5]. Many of these cities are also home to America’s most notable fashion schools, which gives us a large sample of fashion-forward students who may have greater financial incentive to complete the survey but are still invested in fashion.
From this target population, we were able to determine an optimal sample size using the sample size calculator . We calculated sample size by researching each city’s population for 2013, assuming that population should not have significantly fluctuated in the last 4 years (this information was generated by Google’s search engine). From there, we obtained the male to female ratio for individuals in the United States aged 15-65 . We used this information to find that 269 survey responses would be optimal.
After we found our target population, we analyzed them based on motivation and ability factors, using B.J. Fogg’s behavioral change model . Our core action here would be to get our target population to complete the survey. We assumed that we were dealing with a group with mid-level to high-level motivation—considering the fact that we are targeting a group of people that are highly conscious of fashion and invested in it—and potentially different levels of ability. To ensure engagement, we deemed that our engagement strategy needed to elevate both motivation and ability.
To increase motivation:
- Provide a rewards system
- Suggest that their response is crucial in setting fashion trends for the future (this idea would be based on the assumption that the women we are targeting care deeply about fashion)
- Suggest that the response will not be time-consuming while heaping a fair share amount of rewards
To increase ability:
- Make the survey appear short and quick to complete
- Make the survey easy to understand—instructions should be clear
After making these considerations and reading the Medium article on Amazon’s Mechanical Turk , we deemed that using mTurk would be the best strategy. mTurk is the most cost-effective and time-effective method. It allows us to easily administer our survey to a very specific population. This specific population can be accessed immediately without any huge effort on our end. All we need to do is to create the survey and add it into mTurk’s register. Meanwhile, it also allows us to provide a monetary rewards system to our user, thus easily increasing their motivation. We determined that 30 cents would be adequate compensation for a 5 minute survey, especially as this is higher than the median pay for a 5 minute survey on mTurk (median hourly wage is $1.38, which comes out to 12 cents pay for a five minute survey) [10-11]. Our presentation of the survey in mTurk’s system will allow us to further increase motivation and ability. A preview of what we hope to do can be seen in our modification to the HTML code shown in the Medium article :
<h3>Please follow the link for the survey</h3>
<p>Please follow the instructions for each question. Questions are image-based and ask you to determine whether certain images appear appealing or not. </p>
<p>This is a survey about fashion tastes.</p>
<p>The survey consists of just 22 questions and should take less than 5 minutes to complete.</p>
<p><strong>We are looking for responses from women aged 15-64, who live in NYC, LA, Seattle, Chicago, Miami, Austin, Dallas, Columbus (Ohio), or Washington D.C.</strong></p>
<p><a href=”LINK TO SURVEY”>Take the survey here</a> and then enter the code that you receive after finishing the survey into the text box below. Thank you!</p>
<p><textarea cols=”80″ name=”comment” rows=”1″></textarea></p>
The code that users will input after finishing the survey will be: FASHION FORWARD. Our survey title will be “PAID – quick Fashion survey – Women aged 15-64”.
- B.J. Fogg, A Behavior Model for Persuasive Design