JanRezab / Blog

CEO & Co-founder of Socialbakers.com

Clarifying the case of the Jeweler’s Facebook Page in Nice, France

Published: September 16, 2013

First of all, I am sorry this article is not in French, as a company from Europe, we are very committed to having local support in the countries we work in, including our office in Paris. However, I felt that, as CEO, I need to step in and talk about a story which broke in France over the weekend.

What happened? On Saturday night at 9pm CET I was in a weak mobile signal area at a friends wedding party, when suddenly I noticed that Socialbakers.com got an extra 12 000 visitors in just 1 hour. Thats a lot of traffic in an hour on a Saturday, so I just did a quick investigation on sources, countries, social media – and after investigating, found out that this is the case: A French jeweler had shot a robber and a Facebook support page had been set up, which was gaining a huge amount of fans in a short time. The page created grew to 1 million fans within a day, and the stats on our website for the first 24 hours showed that most of the fans seemed not to be from France.

Now I’m in no position, and don’t have the information, to comment on this story, but the news affected us as people started to use the stats on our website to make assumptions that the site might have fake fans.

Just to be clear, we have no way of determining if users on Facebook are fake. There is no technical way of checking this, and only Facebook knows if the fans might be suspicious. Facebook says in their quarterly filings the percentage of fake accounts is up to a maximum of several single digit percentage points (its definitely more managed than on Twitter).

Now it doesn’t happen every day that a local page, jumps from zero to 1 million fans, and this created an unusual situation when combining this data, that caused some confusion. This is because some of the numbers are not real-time, they don’t update right away (the local fan counts specifically update approximately every 48 hours). Yesterday, they already smoothed out and you will see tomorrow, they will smooth out even more.

For the purpose of transparency: Platforms need to aggregate the statistics, and it can sometimes take a while to refresh data. Just to give you a few examples of real-time vs. non-real-time: the Facebook page Like count is real-time, Twitter search stream is real-time, but YouTube channel views are aggregated (on a daily basis).

Unfortunately, in this case, the fan count was updated but the local fan distribution information took some time to catch up (around 24 – 48 hours). This is just how we receive information from the platforms themselves. In the meantime a large proportion of fans were listed as coming from ‘minor countries’ even though they were not yet attributed. The high volume of Likes in such a short period created a 1 in a million situation that we have never seen happen before.

Socialbakers always does its best to be transparent and to publish clarification where possible so we put text on our site as quickly as we could to clarify how this data works.

Not all data is the same, and for example if you see a big jump down on your channels’ YouTube views on the 1st of April, its because YouTube changes the way it counts its upload views (does not count hidden videos). We know this, but you might not – and we realize it’s our job as a company to inform you. Because we have a deep commitment to keep doing this better, we will soon display this info as data notifications that will be very unique in the business of Analytics.

As Socialbakers, we are now taking further measures into product & data integrity. Every study we develop we try to add a data footnote on what sample, what size. If you have any questions about our data, research, studies, or anything we do, always feel free to tweet us, email us at press@socialbakers.com, let us know in any way possible. We will always try to help.

Do You Like This Article? Tweet Linked In