Identifying potential clients using Social Networks

Identifying potential clients using Social Networks

I constantly receive phone calls from mobile carriers. Usually, they offer deals to switch my current carrier. In my case, it's a waste of time and money since I have no intention of doing it. The worst part is that they call all the time which annoys me a lot and make me have a bad impression about the company.

An efficient way to improve this process would be to identify customers who are very unhappy with their current carrier and offer services to them. The question is how do we identify them?

An easy solution is to use social networks to make this identification. The idea is simple: Very dissatisfied users tend to complain and when this is done on social networks (facebook, twitter etc), it is possible to get data from the user (messages, images etc) and use algorithms to understand their mood. After identifying them and their specific needs, it is easy to offer specific deals.

As an example, I have identified some clients on Facebook who are disappointed with a large carrier in Brazil, called VIVO. The whole process of getting the data, processing, analyzing and identifying angry/dissatisfied people was done using Facebook API, Machine Learning/Natural language processing (NLP) methods and R. Check a sample of the results:

For example: an user called Alisson Albanez is very upset about his internet:

"Até pra reclamar e dificil no *8486 ninguém atende fiquei mais de meia hr e nd, no chat também n?o respondem t? com problema no meu chip número XXXXXX uso a Internet diária até parei de colocar crédito pq n funciona, tenho um plano pós pago de Internet no número XXXXXXXX e também tá mto ruim gastei uns 5 minutos pra carregar a página pra reclamar isso quando carrega alguma coisa até whatsap n tá querendo enviar mensagem zem falar nas liga??es que est?o péssimas estou na torre localizada na comunidade de baioes cep XXXXX na cidade de formiga fico no aguardo"

And the user Ant?nia da Costa is angry because can't talk to VIVO's costumer service:

"Parnaíba, 10 de maio de 2017. Hoje, por volta de 12:30, o celular da minha irm? foi furtado. O número dela é XXXXXXX. Até às 17:30 a mesma n?o conseguiu ser atendida na Central de Atendimento Vivo. Que empresa imunda é essa. Ant?nia da Costa Silva Cliente Vivo, n?o sei até quando."

So, If I'm one of VIVO's competitor (such as Oi, TIM, Claro etc), I would definitely explore this information offering plans and solutions to these potential clients using the same social network (Facebook).

#socialmedia #machinelearning #bigdata #statistics #midiassociais

要查看或添加评论,请登录

Tiago Dantas, PhD的更多文章

社区洞察

其他会员也浏览了