![\"Digital](\"https:\/\/www.clickworker.de\/wp-content\/uploads\/2020\/04\/Digital-Market-Research.webp\")
<\/p>\r\n<\/p>\r\n
Digital market research promises a whole lot of data, thanks to which we can see inside the \u201cconsumer\u2019s head\u201d. But what is there really to digital market research? What are its methods, goals and challenges? <\/p>\r\n\r\n\r\n\r\n\r\n\r\n
For a long time, telephone polling and surveys done on the street, in department stores or with invited groups were the preferred information source of market researchers to make marketing decisions. But in a time when truly all social groups are active on the internet and share their opinions there, much more comprehensive information sources are available. The only issue is tapping them correctly, and that is exactly what digital market research does. It searches social media, blogs, forums, comments and other sources for relevant information, to ascertain customer preferences and needs. <\/p>\r\n\r\n
Digital market research describes the analysis and evaluation of information that is gained on and from the World Wide Web. It is accomplished using digital data processing, which ranges from manual analysis of data sets to artificial intelligence (AI). <\/p>\r\n\r\n
People often associate digital market research with big data, machine learning and the accompanying artificial intelligence. However, it doesn\u2019t have to be as complicated as that. <\/p>\r\n\r\n
Even the owner of an online shop conducts digital market research when he evaluates which products sell best. If he knows what those are, he can add similar or complimentary items to his assortment. Similarly, the assessment of customer support requests is also a form of digital market research. Looking at product reviews is naturally also a part of that. <\/p>\r\n\r\n
Companies like Facebook and Google are constantly doing digital marketing on a truly grand scale. They do everything to capture the preferences and needs of their users to then create an advertising profile perfectly suited to him\/her. To do that they rely on diverse algorithms which they are always refining, as well as user tracking and of course artificial intelligence or machine learning. <\/p>\r\n
Machine learning specifically has enormously advanced AI in the last few years. It involves a method by which software (an algorithm) is fed with training data. Based on this information the software independently detects certain regularities or patterns in the training data. Armed with these learned patterns and regularities, the software can then analyze unfamiliar data. <\/p>\r\n\r\n
Example:<\/strong> \r\n\r\n AI is also now capable of not only searching for keywords in a text, but also of delving deeper into it. It comprehends at least in a rudimentary way what the article, post or comment is about. <\/p>\r\n The results of this are practically unlimited possibilities for digital market research. In this way, forum article, posts and comments in social media, in blogs and even images and videos can be entirely automatically scrutinized for content. <\/p>\r\n\r\n Here, we primarily have the goals which traditional market research also has: <\/p>\r\n But there are also goals that only matter in the digital world: <\/p>\r\n\r\n Digital market research allows influencers who fit with your own product portfolio or marketing strategy to be located. With them, companies bring their products to the masses in a manner that is very focused on the target group. <\/p>\r\n\r\n Usability tests of websites or apps also take place via digital market research. Only thanks to end users are there enough different end devices (PCs in various configurations with assorted software, Macs, smartphones, tablets, etc.). Moreover, very different use scenarios and reactions are included that an offline test could never cover completely. Not even Microsoft engages large teams with diverse PC systems to test its updates anymore. That now only happens through users, which, however, often leads to problems. (see: Why does Microsoft Windows 10 have so many bugs? Ex-Employee tells you why!<\/a>)<\/p>\r\n\r\n Online surveys are also only possible in the world of digital market research, whereby they are extremely similar in composition and structure to the classic subject on the street-survey. However, the group of those surveyed allows itself to be expanded to include all internet users as market research participants<\/a>, at least theoretically. <\/p>\r\n\r\n Tip:<\/strong> Find participants for your online survey quickly and easily via the service “Surveys” at clickworker. <\/p>\r\n Find survey participants now!<\/a><\/blockquote>\r\n\r\n\r\n\r\n\r\n Poorly formulated questions, unsuitable tasks, imprecise definitions or interpretation mistakes do not only effect classic market research, but also the digital version. <\/p>\r\n In addition, particularly online surveys have the problem of sample bias. That means only certain people have the time and inclination to take part in such polls. The results are even more biased when specific prices or rewards are tied to the survey. Then \u201cbargain hunters\u201d, for example, participate who have no interest in the survey topic, but only in the incentive. (Additional issues with online surveys can be found in the article: Online Surveys: 10 Most Common Mistakes & How To Avoid Them<\/a>)<\/p>\r\n Poorly trained AI\u2019s are also a big problem. A Google algorithm once classified a dark-skinned person as a gorilla. That was due to the fact that the software had only been trained with images of light-skinned people. There was also software that purported to identify criminals based on faces. However, for its training a police databank of perpetrators was used, in which most of the criminals were photographed wearing t-shirts. Therefore, the software probably used the t-shirt as an attribute for recognizing criminals. (cf. Artificial intelligence on the wrong track<\/a>) This type of poorly trained program of course leads to errors with large data evaluations in the digital market research realm. <\/p> \r\n In addition, artificial intelligence is not protected against fake news. Results can emerge that are based on fake news and for example depict a manipulated trend. <\/p>\r\n\r\n There are many areas of the internet in which great legal uncertainty prevails, especially with respect to data privacy and the ever-present General Data Protection Regulation (DSGVO). <\/p>\r\n Anonymizing the data helps here initially, however, studies show that even anonymized data allows inferences to be made about individuals. And that again creates legal uncertainty. <\/p>\r\n\r\n And finally, there are also ethical problems in the field of digital market research. Are pollsters allowed to go as far as Cambridge Analytica and analyze personal data to influence an election? (cf. Cambridge Analytica<\/a>) Digital market research makes the detailed exploration of people\u2019s behavior and needs possible. Artificial intelligence specifically makes for an enormous workload reduction here, because it evaluates huge amounts of data in a very short time. Ultimately, digital market research creates a transparent consumer. <\/p>\r\n On the other side it also harbors risks. They can be systematic, like poorly trained AI\u2019s that bring to light completely incorrect results. Moreover, it presents legal and even ethical problems. At the end of the day digital market research encounters the same challenge as other sciences: It is not only a tool, but also an agent. For that reason, it carries with it just as much responsibility as the companies who use the results of digital market research. <\/p>\r\n\r\n\r\n\r\n <\/p>\r\n","protected":false},"excerpt":{"rendered":" Digital market research promises a whole lot of data, thanks to which we can see inside the \u201cconsumer\u2019s head\u201d. But what is there really to digital market research? What are its methods, goals and challenges?<\/p>\n","protected":false},"author":36,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[716,718,996,722],"tags":[],"yoast_head":"\nWhat are the Most Important Goals?<\/h2>\r\n
\r\n
Finding Influencers<\/h3>\r\n
Usability Tests<\/h3>\r\n
Online Surveys<\/h3>\r\n
Problems with Digital Market Research <\/h2>\r\n
System Errors<\/h3>\r\n
Legal Challenges <\/h3>\r\n
Ethical Questions <\/h3>\r\n
Digital Market Research Combines Opportunities and Risks<\/h2>\r\n