Battling the Bots – Chanel case study

  • 28 May 2018
  • Written by: Emmalena L Ellis

How to Combat the Black Market of Social Media’s Popularity Game

Mention fraud detection


In a world that fixates so strongly on popularity and status, one of the ways that companies monitor success is through their audience on social media. Whether it be likes, retweets, pins, ratings, views or a new follower, the overall buzz for a business can ultimately grow your community of networking opportunities, customers and keep you ahead of the competition all in a single click.

Correctly using analytics, you can better target your digital footprint taking the opportunities best presented by the demographics and patterns of your customer behaviour. But with millions of fake followers targeting celebrities, politicians, pundits, athletes and popular brands, how can you grow your audience organically? And how can you direct your focus on the key issues your network expects? With Analytical tools from Bat Radar, businesses can begin to separate fake profiles from their target audience demographic.


What is a Bot?

The term ‘bot’ relates to a series of artificial intelligence used to automatically perform a task or series of functions to distribute content. However, with many social network accounts updating information in real time, the appropriation of ‘fake profiles’ and ‘fake reviews’ by third-party scammers and spammers can ultimately damage your digital presence and can create false content or steal information from your website that can detract from the important real-time leads and results necessary to improving a positive environment for both yourself and your clients, search engine credibility and overall website traffic.

Due to the work of ‘bots’, companies can often find their analysis is polluted by false and misleading data. Implementing the correct strategies that have been encouraged by their network is in the best interest of any company should they wish to gain rapport, which will not be possible should they be misdirected by non-human responses and reactions.


How Does Bat-Radar Help?

Using Bat Radar’s tools, you can easily detect the bots. Working with a fine-tooth comb for our clients, our unique platform can easily verify the page of any follower and classify it accordingly as legitimate or as a Fake.



Mention fraud detection. Bat Radar detected 1498 mentions @chanel as of 17 May 2017


Mention fraud detection. Bat Radar detected 2117 mentions @chanel as of 18 May 2018

Mention fraud detection. Bat Radar detected 2125 mentions @chanel as of 21 May 2018


Chanel case study

Using the example by searching for Twitter mentions of both Chanel and its Creative Director, Karl Lagerfeld, Bat Radar were successfully able to determine some of the patterns of a bot which was using the handle @rosapatrica23 who between 2016-2018 had mentioned the company over 2000 times, which is a staggering number of a normal follower’s interactions, approaching at least 1000 references a year; although many of these make little sense and provide clickbait links and fake information.



@rosapatricia23 Twitter account


Ways We Spot a Bot

Many fake accounts and bots have similar characteristics:


  1. Lack of Relevant Information – When searching for a bot the team carefully consider whether the page’s content provides information of value. The page of a legitimate account will usually provide the continuation of a discussion or develop a question or debate. With this account as an example, Bat Radar could distinguish a pattern of word-clusters that did not relate well to their chosen targets.
  2. Bots in the Family – The bios and followers will also be bots. The biography provided by the bot to confirm their ‘identity’ lists various other bots as members of their family and an irrelevant website as their home site – in this case, a generic family tree.


Rosapatricia23 bio

3. Brands are the Name of the Game – This profile generated mentions for brands and celebrities alongside Chanel and Karl Lagerfeld. The bot constantly mentions Tom Cruise, Scientology and a fake version of Queen Elizabeth II.

4. Hiding in Plain Sight – Every now and then the profile changes its name so it can avoid being picked and suspended. Whilst the handle (rosaPatricia23) remains the same, the account has changed names and identities, including Rosa Patricia Devoy, Romero Devoy Burgos, and most recently GuirolaMcKinonRomero. At the time of publishing the article appears a new name: GuirolaBurgosRomeroCarballo i French ElSalvador. It also appears a new profile picture, certainly drawn by chance from the account of the unaware lady in the picture.


Rosapatricia23 tweets


What Can I Learn from This?

Having studied this bot, we can consider a few questions both about the bot’s overall purpose and how the industry can evolve and grow.

  1. Who and why has the bot been created? i.e. Who gains traffic from its purpose. Is it the brand itself? A social media agency? Or a market competitor?
  2. With the growing emphasis and evolution of social media and analytical tools, how can businesses better work to encourage relevant traffic back to existing brands? And how might they find ways to better pinpoint and eliminate spam accounts from their data?
  3. If companies do not appropriately filter out the fake mentions, fake profiles, bots and any artificial operations, how can brands correctly interpret their data and share these with their stakeholders and develop their overall strategy? Companies analyse data, to help determine which direction they need to project their business. Social media when used correctly is a soundboard in which to listen to the consumers. If a company collect false data, how the company itself (Chanel or Karl Lagerfeld in this case) could use them in function of sales? In function of profiling their customers? Unless we think that the whole thing quickly becomes meaningless and will not help overcome the competition.


Social media today is one of the most important parts of any online business because social media put the customers at the centre. But bots continue to succeed because they work to better leverage themselves at adapting the needs of their programming and overthrowing the overall reception of the brand; despite the efforts of many companies and moderators. Unlike customers who can independently lose or attract interest, bots show little remorse beyond the confines of their code.


With that in time let’s see how long it will take to the inverted genius to delete the @rosapatrici23 account, which has existed now for over two years, I’ll start to count: 1, 2, 3….


Contact Bat Radar today to learn more about our bespoke analytical programmes including our Fraudulent Mentions Tool.




Leave a reply