In order to find adequate online communities to collect data from, we developed a list of search terms. Using these search terms would ensure that we find communities that directly address our trend and our research question.
We developed a mind map in order to show our search progress, from broad to specific searches. Since the sub-question that we are trying to answer with netnography is “How do people react to the Frappuccino unicorn trend?” we decided to have our broadest search term be “Unicorn”. Then, depending on the community, we continued to add search terms and narrow down our results. Different communities needed a in depth search for the correct community than others.
For example, “Starbucks” was a general community on Reddit that had over 20,000 users. We choose to analyze the only discussion on this community specifically about the unicorn Frappuccino. This meant we already had a community familiar with the trend with a good quantity and variety of comments.
YouTube on the contrary, had an endless supply of videos and comments about the Starbucks unicorn trend. Very specific search terms were required in order to find a video with a comment section worth analyzing.
Facebook is a online social media platform used to share media, events, opinions and lots more. With 1.4 billion active users, Facebook is by far the largest social media platform. With roughly 510 thousand comments being posted every minute, Facebook also has some of the most active users online. (Zephoria 2017)
The majority of users are still between 13-25, however Facebook has been gaining users from all ages. For online forum research Facebook is can be very useful because most of the time comments are linked to real profiles. This ensures more truthful and traceable comments.
YouTube is a online platform were anyone with an account can upload videos and comment on videos. There are over a billion users registered for YouTube and 100 million of them take a social action, such as liking, commenting or sharing videos.
Youtube is being used by all types of people all over the world. These videos are easily accessible to anyone. Youtube videos also have a massive amount of comments which gives it a high variety and amount of data to gather even with specific searches.
Reddit is a popular online platform for people to create their own sub-communities (subreddits) dedicated to a certain topic and share media, questions and other posts on that topic. Certain subreddits can also have around 500 thousand unique visitors each day. Due to some popular subreddits, Reddit has gained a reputation for breaking stories and having users be the first to comment on them.
Seeing that subreddits such as unicorn or starbucks have upwards of 20 thousand dedicated users, the interactions within these communities should yield thoughtful comments by people who are passionate about the topic.
Our role as ethnographers
Active or passive
We have decided to take on a passive role as enthnograpers in the online communities in which we conducted research. This means that we did not participate in any conversations nor did we start a conversation by posting a question regarding the topic of our interest. We have chosen to take on a passive role because there is no need for us to participate in the conversation, as we found that the people within the communities are already talking about our topic of interest, so there was no need for us to initiate the conversation. By solely observing them, we would be able to get the data needed for our research.
Overt or covert
When deciding on our role as ethnographer, we had to choose if we would be overt or covert. When an ethnographer is overt, they will let the people of the community know what their intentions are and they make sure that all members are aware of what is happening. However, when an ethnographer is covert, they do not inform the community on the reason for their presence and hide their true intentions.
For our research we have chosen to be covert. This is because we did not want to influence the participants by letting them know we will used their comments and quotes for research, as this may influence what they will say about the topic, how they will say it, or if they will even say something.
How we introduced ourselves
We did not introduce ourselves in the communities as we were already part of the communities ourselves. Besides this, as stated before, we have chosen to be covert, meaning that an introduction on who we are and why we are joining the community would not be present.
Immersion in the communities
As stated before, we have done research on our topic in three different communities, namely Facebook, YouTube, and Reddit. Down below a table can be found with all the information about our immersion into the communities and the specific threads/posts/videos used for our research.
As we have collected over 50 comments for each thread/post/video, we did not include those in the following table. They have all been opened and coded in ATLAS.ti. The results of this will be presented next in the data analysis section.
(For the entire Facebook link please copy past the entire link into your browser, the link has been cut-up to make it adjusted to this webpage.)
There are two main ways of collecting data when doing netnography. These two ways are elicited data collection and archived data collection. Elicited data collection is when you as a researcher interact with a human being in order to retrieve information from them. You evoke conversation and your will bring your topic of interest to light in the community in which you will do your netnography research (Data elicitation, 2014). Archived data collection on the other hand is when the researcher looks for data that already exists in the community in which they want to do netnography research (Rabinowitz, 2016).
For our research we have decided on using archived data collection as our data collection method. This is because, as explained earlier on, we have decided to take on a passive and covert role as ethnographers. This means that we would not participate in discussion and therefore not evoke conversations, generate questions, and record the answers, which is what one does when using elicited data collection as their method of data collection.
The coding process
After collection all the quotes for our research from all selected communities, we have put them in ATLAS.ti to code them. We have coded all five data sets separately, as the program did not allow us to put them together in one file as we were using the trial version of the program, which does not allow more than 100 quotes in one file. After all the quotes had been coded, we compared all the codes in order to see which codes appeared in multiple data sets and to see which codes could be merged into one category. Once we finished comparing the codes from the different data sets we decided on two main categories, namely “positive codes” and “negative codes”. Both categories had a total of six different codes under them, which were as follows:
Positive codes Negative codes
- Excited Health
- Positive joking Disappointed
- Fun Negative joking
- Recommendation Restricted to U.S. and Canada
- Pretty Expensive
- Not concerned about health Surprised
In the following table a brief explanation will be shown for each code, as well as some example quotes that were labeled with the particular codes. Also included are the main results for both categories (positive codes and negative codes), which are the frequency of each code within the categories.
Explanation of codes, related quotes, and frequency of codes
|Code||Description of code||Related quotes||Frequency|
|Excited||Quotes that got this code showed the excitement of the person to try the unicorn frappuccino.||“if this is real I will be working out of Starbucks for the next week, please join me!!”
“I see a much needed date in our near future drinking magical drinks!!“
|Positive Joking||The positive joking involve jokes about how one’s friend would love the unicorn frappuccino.||“I can imagine you running to get this at your nearest starbucks”
“found your Starbucks”
|Fun||This code is about how people think the unicorn frappuccino looks fun.||“For all the craby people, it’s only for a week and it looks fun!”
“This might be a bit on the sweet side, but we definitely need a selfie with it“
|Recommendation||This code covers quite in which one person recommends the frappuccino to a friend.||“you need to get one of these while you’re in Miami!“
“Now may be the time to become a frappuccino drinker!”
|Pretty||This code covers all quotes in which people praise the appearance of the unicorn frappuccino.||“will have to get us this sometime looks beaut”
“they look so pretty, I nearly just want one for the look of it”
|Not concerned about health||This code covers quotes where people do mention how bad the unicorn frappuccino will be for their health, but that they want to try it anyway.||“All these people whining, “It’s unhealthy”. Well, I’m thin and eat well the majority of the time so what’s wrong with
a “treat” every so often? Won’t kill you. LIGHTEN UP! Looks fun! Gimmicky but, FUN!”
“people need to lighten up about sugar. I’m guessing more people die from stress than sugar lol”
|Health||This code covers every quote in which people talk badly about the unicorn frappuccino because of it’s unhealthy ingredients.||“With over 70 grams of sugar….equivalent to 3 snickers bars!! We wonder why everyone in this country has diabetes!”
“Why are you posting this?…there is nothing Healthy about thissss….coloring, lots of sugar, fat, etc!!! “
|Disappointed||Quotes that got this code showed people disappointment in the unicorn frappuccino.||“not really into how these tasted… “
“I got to try it yesterday…it’s not as sweet as their other Frappuccinos”
|Negative joking||The code negative joking groups together quotes in which the person jokes about the unicorn frappuccino in a bad way.||“That….Looks like a car wash…..”
“It looks like someone liquified an outfit the 1970s and turned it into a drink….”
|Restricted to the U.S. and Canada||This code covers quotes in which people complain about the fact that they cannot get the unicorn frappuccino seeing as it is restricted to the U.S. and Canada.||“I asked in Starbucks today about this and they said it was only in America “
“Omg…i want this…cant get this on vitality (or in the UK)”
|Expensive||This code covers quotes in which people complain about the price of the unicorn frappuccino and Starbuck drinks in general.||“The only thing magical about this is how it will empty your wallet and make your teeth disappear. Awful”||5|
|Surprised||This code covers quotes in which people seem negatively surprised about a certain aspect of the unicorn frappuccino.||“It’s almost taste like cough syrup to me wtf! “
“t tells what it’s made of and its mango and other stuff. Never would have guessed that”
Summary of findings
By means of the following graphs and tables we will present a summary of our findings. We have collected a total of 274, and have coded following the coding process as described earlier. The following graphs will illustrate what parts of the collected data were grouped into the same categories, as well as the division of data within those categories.
In graph 1.1, an overview of the frequencies of all codes has been displayed. We can state that the most frequently occurring code is the code “excited” (30%), followed by the codes “positive joking” and “health” (both 16%). The code that occurred least frequently was the code “surprised” (1%). The rest of the codes are all pretty close together as their frequency ranges from 2% to 10%.
In graph 1.2, the division of all quotes into the two categories has been displayed. In total, 193 (70%) of the collected 274 quotes fell under the category “positive”, and the remaining 81 (30%) could be grouped under the category “negative”. The graph clearly illustrates how the category “negative” took up most of the collected data.
As seen in graph 1.1, there were 12 different codes that each included a specific amount of all data collected. In that graph, we were able to see what the most used codes were when looking at the collected data as a whole. In graph 1.3 however, the frequencies of the codes that fall in the “positive” category are displayed. We can clearly see that the code “excited” (43%) occurred most frequently in this category, followed closely by the code “positive joking” (23%). The codes “fun” (14%) and “recommendation” (12%) come in third as they are used almost equally. Lastly we can see that the codes “pretty” (6%) and “not concerned about health” (2%) both did not occur frequently.
The frequencies of the codes that fall in the “negative” category are displayed in graph 1.4. In the graph we can see that the most frequently used code in the category is the code “health” (53%). This code is followed by the code “disappointed” (17%), after which come the codes “negative joking” (11%) and “restricted to U.S. and Canada” (10%). The codes occurring least frequently in this category are the codes “expensive” (5%) and “surprised” (3%).
Starbucks started the spring season by introducing the unicorn Frappuccino on April 19th. This was their attempt to tap into the unicorn trend. A focus for our research was set by developing the sub-question “How do people react to the Frappuccino unicorn trend?”. We used Netnography in order to gather data and eventually answer that question.
The data we gathered came from online communities in the form of comments and other interactions. In order to find adequate data, we focused on three different online communities, namely: Facebook, Reddit and YouTube. We based these choices on a variety positive aspects of each community.
We researched into each community and made assumptions based on that research on how valuable each community will be for us. However not all of these assumptions turned out to be accurate.
Facebook had delivered many more useful comments of people actually commenting on the product/trend itself, while YouTube had a large section of comments not addressing the product/trend at all.
Reddit had a very engaged and interested group of people. Unfortunately Reddit’s algorithm places “liked” comments all the way to the top. Unable to code all of the comments, the most liked comments were coded first, while others were ignored. Seeing that liking a comment is also a form of interaction and voicing an opinion, this shouldn’t corrupt the data.
Nevertheless the data revealed evidence showing a lingering excitement and enthusiasm towards the trend. Starbucks has recently discontinued making the product, yet many comments still expressed an eagerness to try the product. By looking at our data we found that 30% of all comments had been coded as “Excited” and a 70% majority of comments being positive. The absurd nature of a unicorn frappuccino has also directed a great deal of positive and negative humor at the product. Yet it is this absurdity that is also creating free publicity as well as demand for the product.
Data elicitation. (2014). ELICITING DATA ?? . Retrieved 06 06, 2017, from Data elicitation: https://dataelicitation.com/about/eliciting-data/
Rabinowitz, P. (2016). Section 7. Collecting and Using Archival Data. Retrieved 06 06, 2017, from Community Tool Box: http://ctb.ku.edu/en/table-of-contents/evaluate/evaluate-community-interventions/archival-data/main