Where can operators get target profiles so they can attract advertisers to shift their budgets from traditional media to mobile media? Herein lies the advantage of operators over OTT players who have been dominating the online advertising cake. If OTT providers rely heavily on data feeding by users on their service sites, operators have been sitting on mountains of gold so far, because they have data from data traffic on all mobile internet activities that go through their network. However, it seems that, with the recently introduced mobile advertising mechanism, only the surface of this golden mountain is ‘scratched’, where advertisements are displayed after the system detects access to certain addresses. Then, the system still performs an even rate on the types of advertisements that appear if there are visitors to site X, the ad that appears must be A, and so on. Fortunately, ott advertising is capable of doing this more accurately.
Furthermore, if you are sniffing data packets passing on the operator’s network, you will get a lot of information that can reflect user behavior. In addition to the site address, which is very clear, from a telephone number that accesses the mobile internet, you will also get the type of smartphone, the type of browser, the frequency of visits to certain sites, what activities are carried out on the site, even if the line used is not encryption, can get the ‘content’ of the contents of the data that is passed.
Of course, all of this can be used as the basis for profiling to match what types of ads will be displayed to certain users. The data flow must be analyzed ‘on the fly’, so it is necessary to have a big data processing system to do so.
There is still one issue left in processing mobile internet traffic data for advertising purposes, namely the issue of privacy. As is known, recently, several internet providers, including one of the largest operators in the country, were accused of spying on user data. As long as what is done on the data is behavioral analysis related to commercial activities without explicitly revealing the identity of the customer, then that activity should not be categorized as spying.