Wednesday, May 27, 2009

What do Macedonian Facebook users do with the ads by Macedonian companies?

A survey of 450 Facebook users from Macedonia indicates that very few of them—one percent—actually use the benefits of e-commerce provided by this environment, and that over a third of them do not click on the advertisements by Macedonian firms served by this social networking platform. Even though majority claim to click on the advertisements, most of them do not do that often, and almost two thirds of them have visited the advertised websites only once.


To gather the data, I posted a note (simple text article) available to all Facebook users, containing the question and the alternative answers (closed set). Users were invited to provide the response via private messages or to post them as comments with the number of the alternatives, and additional info at their convenience. The decision to use a note, as opposed to a specialized poll or quiz application, was based on the aversion experienced as a user, and corroborated by other users, to the constant influx of invitations to install new applications.

This user-friendly approach and the nature of the social media contributed to increasing the number of participants, some of whom shared the link among their friends. The data gathering took place from May 20, 2009 till May 30, 2009.

All communication and inviting took place strictly in Macedonian language, in order to provide a sort of filter for foreign friends of the surveyor and the respondents who might notice their activity regarding the note.

The responses are based on respondents’ recall and can be used to gain a rough sketch of the current state of the affairs, but cannot be taken as precise measurement of the effectiveness of the Facebook ads, or the level of use of e-commerce in Macedonia. However, it must be taken into account that Facebook users as a group can be considered to be potentially more wired than the general population of Macedonia, for at least they need to know to use the web and e-mail to start using Facebook.

In addition, six respondents, who reside outside of Macedonia, also took part in the survey. Their responses—that they do not notice such ads—confirming that the ads by Macedonian firms are targeted according to geographical location of the viewers. The responses by the Diaspora members were not included in the final tally.


Out of total of 456 survey respondents, 450 persons from Macedonia answered the question: “What do you do with the advertisements by Macedonian firms on Facebook?”. Cleared data from 435 respondents indicates that:
  • 27,4% do not notice those advertisements
  • 6,0% intentionally do not click on them
  • 37,0% click on those advertisements very rarely or never
  • 25,3% occasionally click on such advertisements, and
  • 4,4% often click on them
Several respondents who choose the first option also reported that they use Adblock Firefox Add-on, or other service which automatically prevents them from viewing ads. Future surveys on this topic should add this possibility as separate option.

In addition, the respondents were asked what they have done after clicking on such advertisements. The 294 survey participants, who first confirmed that they’ve clicked on such ads, responded as follows:
  • 67,7% - I have visited the advertised websites only once (then and never again).
  • 16,0% - I have returned to an advertised website several times.
  • 3,1% - I have placed an advertised website in the bookmarks/favorites.
  • 1,4% - I have bought a product or a service directly from such a site.
  • 0,3% - I have bought an advertised product or service via telephone.
  • 2,7% - I have bought an advertised product or service from a shop.
These results show a correlation with the results of a study of banner ads effectiveness published by the American company iProspect, in particular in regard to 36% of users who choose the option “Nothing. I haven’t responded to any such ads,” within this survey, which is quite close to over 33% of Macedonian users who selected one of the two options with the result of non-clicking on the ads.

However, over 66% of all respondents reported clicking on the ads by Macedonian companies, which points out that this targeted advertising on social networks indeed has a strong potential for reaching the intended audience. This notion needs to be confirmed by data on the actual Click-through rates (CTR), even though the CTR can take a back seat to building brand awareness by users who just see the ads and tend to get familiar with the advertised brand name.

The low level of actual online purchases done by the Facebook users can be due to the low level of development of e-commerce in Macedonia, which also indicates low user confidence in the existing domestic e-shops. Only one percent of the respondents who clicked on the ads confided that they’ve purchased an advertised product or a service online, as compared to three percent of them who later used a brick-and-mortar shop to conduct an actual business transaction.

Assumptions and Limitations

The main limitation regards the possible bias of the sample, which included 450 self-selected respondents, a number of whom are not surveyor’s Facebook friends, so even their self-reported data such as sex could not be accessed and verified. Estimated sex ratio for this group is 48% to 52% for males and females, respectively.

The assumption for this survey is that ratio of male to female users on Facebook reflects the estimate for the general population of Macedonia of 50% each. This could not have been verified, as Facebook estimates the ratio of males to females for users of all ages (by the rule: over 13) from Macedonia to roughly 51% to 49%. This estimate is prone to errors due to using self-reported sex info as basis, and taking into account that some users do not input this data. For instance, on May 26, 2009, the official estimate by Facebook for the total number of users from Macedonia was 170,360, with 84,660 males and 83,120 females, leaving 1240 unaccounted for. Even more peculiar is that the total number for males and females from Macedonia provided by Facebook is 168,800, while the simple sum of the above numbers is 169,120, with an unexplained difference of 320.

A rough estimate of the ethnic affiliation could be made, based on personal acquaintance and used surnames, yielding a ratio of 82% of ethnic Macedonians, 4% of ethnic Albanians, and 12% of people belonging to other ethnic communities (Serbs, Croats, Turks, Vlachs, Roma, Slovenes, etc.) or whose affiliation is undeterminable at this point. This does not provide a strong correlation with the general population. According to the 2002 census, the population of Macedonia consists of 64% ethnic Macedonians, 25% ethnic Albanians, and 11% others.

Facebook does not collect data on ethnicity or native language, and it cannot be verified if the above estimates point to a kind of digital divide between ethnic communities, indicating that a larger proportion of members belonging to one of these categories use the system.

Taking all the assumptions and limitations into account, no efforts were made in order to dig deeper into the data, such as cross-indexing of gender and a particular response. Even though the sample size of 0.26% of the total population of Facebook users from Macedonia is not negligible, to segment it even further would be too ambitious for a survey of this kind.