X Close

Global Social Media Impact Study

Home

Project Blog

Menu

Does Targeted Advertising Work?

By Daniel Miller, on 5 February 2015

Photo by Mike Licht (creative commons)

Photo by Mike Licht (creative commons)

As Ethan Zuckerman noted in The Atlantic (14/08/2014) even though many groups and initiatives really didn’t want to go down that route, targeted advertising has become the default funding model for the internet, as people failed to find an alternative. A combination of developments such as big data and mining information from sources such as search engines and social network sites means that today it is possible for ads to be honed quite precisely to the interests of individuals as revealed by their online activity.

It is not at all surprising to find that English people who, as many of my blog posts have argued, are hugely concerned with privacy and keeping people away from their homes and intimate worlds, vociferously complain about the development of targeted advertising. The two most commonly quoted examples are Facebook and the supermarket chain Tesco. A typical complaint was ` Google will change your settings on your cursor, so that every time it goes back and tells them what you are using it for. Then they send you certain adverts….If you join Tescos, every time you go through the till it records everything you’ve brought. And suddenly they start sending you vouchers to buy meat… or this persons a drinker. Everything you do.’

In our project we anticipate cultural variation and it was interesting to read an article in the Financial Times recently (28/01/15) that suggested in China customers of WeChat felt personally insulted when they were not included in a targeted advert for BMW. This leaves us with at least two interesting possibilities. The first is that people say they resent the advertising but actually find them convenient and use them, which is why they continue to spread. Alternatively corporations tend to follow technological advances and do this simply `because they can’, even if in actuality these adverts did not in fact work. When I studied businesses (Miller, D. 1997 Capitalism: An Ethnographic Approach) I found that fear of what the competition might do was much more important than evidence for what customers actually do in understanding business practice around advertising. The academic work on the topic is still slight, and it is starting to look like targeted adverts in some combinations might actually be sending people away from companies rather than building their profits (e.g. Goldfarb, A., and C. Tucker. 2011 “Online Display Advertising: Targeting and Obtrusiveness.” Marketing Science 30.3 (2011): 389-404). In the meantime I have been faced with some of the most egregious examples of such advertising through my research with hospice patients. As one put it `I’ve joined the moving-on group now, since I’ve finished treatment, try and move on. Sometimes I get a lot of feeds and it does get a bit much. Don’t want it in your face all the time, keeps coming up, so I had to stop a lot of the feeds, otherwise every other thing was cancer cancer cancer and I’m not moving on. Think I’ll get rid of these off my Facebook.’

A talk at Oxford Internet Institute

By Xin Yuan Wang, on 18 November 2014

by Shriram and Xinyuan

Our research Shriram and Xinyuan giving an talk about social media in India and China at Oxford Internet Institution

Our research Shriram and Xinyuan giving an talk about social media in India and China at Oxford Internet Institute

It is always good practice to exchange knowledge and ideas with scholars from other fields, as it adds immense value and vision to the research at hand. New Media and Social Media are fields that attract scholars from various other areas of study. The bridging of interdisciplinary ideas that this area of research induces is phenomenal. For example, while social media can be approached from an anthropological point of view, it can at the same time also be approached from a Big Data perspective. In a previous blog post Xinyuan talked about the difference between ‘Big Data’ and digital ethnography.

A couple of weeks ago, two of our researchers (Shriram and Xinyuan) were invited by Professor Ralph Schroeder, a Big Data theorist, from the Oxford Internet Institute (OII) to give a talk on social media in China and India, from an anthropological perspective. This opportunity, which is one of our earliest since our return from our intense 15 months of fieldwork provided great value in understanding and evaluating aspects related to disseminating our research results.

One of our first significant points of contribution was to address the need to study social media in a more traditional anthropological way. This also flowed from the strength of ethnographic evidence that the nine field sites generated over the last 15 months. Ethnographic evidence needs to be presented in a context based ‘thick description’, which anthropology does, allowing researchers present the field site in a rich descriptive and detailed format. This allows the readers/audience understand the situation/context better. The process of communicating such detailed description also brought us another opportunity to transition to a multimedia based approach. Xinyuan showed a 4.5 mins documentary film clip about her field site, Good Path town, a small factory town in southeastern China and the way she did ethnography among Chinese rural migrants who work and live in the factory town as factory workers. The result of using a visual platform to describe one’s field site was extremely positive. When the audience really saw the population and the field in the short clip, the subject no longer stayed as alienated or abstract but became far more humanized and engaging. The positive feedback received has also further encouraged us in our long-term dissemination plan for the project, which also involves a multimedia approach to report on our analyses and findings.

The discussion section was also inspiring. The questions varied from censorship in China to the influence of gender in the use of social media in India; and a few questions touched the very core issue of this comparative project, which was on how to draw conclusions from the ethnography of nine different sites.

Even from the two field sites (India and China), let alone the total nine field sites of the whole project, our audience had already strongly acknowledged the huge differences in the way people use social media as well as the impact of social media usage in local peoples’ daily life. Though the presentations were not intended to be comparative, the format in which they were delivered played a role in giving rise to a few interesting questions that leaned towards a comparison of general issues between China and India. For example: after listening to our presentations, students found that the use of social media in India was strongly influenced by the reality of offline life, however in China among Chinese rural migrants, social media offered a platform where people can simply set up a new world where they can enjoy an ideal life where their offline lives (including their social status are largely irrelevant. Although it’s always risky to over-generalize claims of totally different uses of social media by lower socio-economic groups in India and China, the ethnographic evidence gathered from our 15 months field research allowed our study to showcase the diversity of social media usage in different cultures and societies. Social media itself is by no means a unified or universal concept and its meanings are way more diverse than we can imagine. In short, this opportunity helped both the parties (OII and GSMIS scholars) to explore and understand social media in a much deeper context through an anthropological perspective that is contextually and fieldwork based.

Small (random) thoughts on Big Data

By Shriram Venkatraman, on 12 December 2012

Photo: hisperati (Creative Commons)

A casual search for the definition/description of Big Data can throw up results that define/describe this phenomenon in various ways. Though most agree on size (as the term itself implies), there are other dimensions applied to this term, that seem to be on the increase based on the nature of the industry that defines this. Definitions range from using 3V models to 4V models; single dataset to multiple datasets; single database with multiple datasets to multiple databases with multiple datasets; size of each dataset from gigabytes to exabytes (very relative); nature of each of this dataset; complexity not only in terms of types of data sources but also with respect to the relationships that these data points share; speed (or velocity) at which the data is produced, so on and so forth. Other than the dimensions of size and complexity, it looks like the definition of Big Data is as big as the data itself.

From a universal perspective, most of these definitions that speak about the size of the dataset proclaim that humanity creates 2.5 exabytes of data every day. However, one has to remember that this is only tracked data defined based on the technological storage capacity. So, what happens to the untracked data? So are these exabytes of data our data generation, or production that can be tracked by technology? Though, this will definitely grow in size as technology advances with data storage capacity, can technology reach out to every nook and corner of this world? It seems like a major portion of Big Data description is limited to the digital space alone.  Though, the definition of Big Data seems to grow in a non-linear fashion, the growth of Big Data itself seems to be linear based on its dependency on digital and or technological growth.

Data can be processed and does have the potential to turn into information, and information can be broken into data – so processing of this information is, in a way, producing more data, which is again processed to produce more information which is data again – in a way becoming a vicious cycle of production, storage and processing.

It will definitely be interesting to see what comes out of Big Data research; it might produce big definitions, bigger philosophies and biggest profits too.

‘Big data’ or ‘Data with a soul’?

By Xin Yuan Wang, on 8 November 2012

Image: Thegreenfly (Creative Commons)

What is big data? In the digital era, the data produced by people on an everyday basis is myriad. There is always more data coming into being, and it is growing at an unimaginable rate. People believe that big data will lead to big impact, claiming that big data opens the door to a new approach to understanding people and helps to making decisions. At the 2012 World Economic Forum in Davos, Switzerland, big data was a theme topic and the report Big Data, Big Impact by the forum claimed that big data should be considered as a new class of economic asset, like currency or gold. People who are masters at harnessing the big data of the Web (online searches, posts and messages) with Internet advertising stand to make a big fortune.

I love data, so big data sounds brilliant! However I am not a ‘big fan’ of big data. Partly because, for me, big data sounds more like a marketing term rather than analytical tool; partly because, being trained as an anthropologist, I am very cautious about going too far out on a limb to make such assumptions. For me, it will be a great pity to see people who fancy formulating big data with brilliant statistics, however ignoring the little stories happen in daily life which have been taken for granted

For anthropology, to some extent story is the date with a soul, or contextualized data to be exact. There is always a danger that data without a context would be confusing and very misleading. For example, in my previous study on the appropriation of Facebook among Taiwanese students in the UK, one thing I discovered is that the Taiwanese use the function ‘like’ on Facebook much more frequently compared to UK Facebook users. For a Taiwanese who have 150-200 friends on Facebook, 20-50 ‘likes’ for each status or posting is very commonplace, and the average amount of ‘like’s’ which people give to others is 15-35 daily. Such considerable amount of ‘likes’, per se, could possibly lead me to making some superficial conclusions, for example, that Taiwanese are more predisposed to admire others online, so on and so forth. However, it was only after long-term participant-observation and several in-depth discussions with each of my informants, that I start to realize that both the Chinese normativity of proper social reaction (save face, reciprocity, renqing) and moral responsibility taken by individuals in the negotiation of real life communication practices shape the pattern of Taiwanese online performance.

 “For most of the time I ‘like’ people because I have nothing to say about their updates, but I want them to know that I care about them, I follow their lives.”

“Liking is polite, just like saying hello when you meet your friends. Nothing to do with the content which you like.”

“…I kind of think that, the more I like a certain person, the less I want to be really involved into his/her real life. ‘Like’ is easy and safe. You know you still need to give a face to people.”

Also, according to the principle of Chinese “Bao” (reciprocity), people who have been ‘liked’, will try to find all the means to pay off debts of the “Renqing” (favor) to others.

“I would expect ‘likes’ from others on Facebook, you know, which makes me more engaged with them and I will like their posts as often as I can. For those who like or leave comments on my profile, I will reply to them with careful preparation to show my sincerity.” as the other key informant said.

It’s so interesting to explore the ways in which “Being Chinese” and Facebook appropriation have been mutually constituted. Facebook is to some extent re-invented by the Taiwanese. If I just count how many ‘likes’ and analyze it without looking into the online content and offline context, I will miss the point no matter how big and sophisticated the data is.

So, the question is whether we are looking at ‘big data’ or ‘data with a soul’? Of course, these two are not necessarily mutually exclusive to each other, even though there are some things you can only do with Big Data or ethnographic data. The point is how can we take advantage of the best parts of the both and contribute to the understanding of our human society as a whole, which is also a big question mark for all the researchers in the digital age.