BPO stands for "business process outsourcing." In short, it's a business practice we see implemented when an organization decides to outsource activities like payroll, human resources, billing and customer service. The best example of this is customer service because we all have experienced speaking with someone from a different country when we've called a bank or had an issue with a credit card and needed it resolved.
We will not spend any more time discussing BPO, but our technology conversation in this article will be focused on improving customer service. Now, recall an incident when you called your credit card company. You were likely asked to press 1 for English, press 2 for Spanish and then several options were presented before you finally get an option to press a number to talk to a real human. Next came the verification process where you had to provide your first and last name, then your date of birth, then your secret answer, or pin, or maybe the last four digits of your social security number. Finally, a CSR (customer service agent) validates your identity and you have an opportunity to ask questions. At this point, the customer service agent may have full access to your call history and any other interactions that you had with them in the past.
So what's the role of machine learning in all this?
Now, imagine a smart system where you are automatically redirected to a smart agent (or a digital agent) who knows that you are calling in to talk to a customer agent because you were on the website or app looking for answers to a particular question. You even interacted with the chatbot, but your question was not answered. Your calling number and voice can be used to verify your identity to search your record instead of spending the time to look up your information. There are machines behind the scene ingesting, processing and analyzing this interaction in real time and predicting that you are about to call the customer service.
Machine learning (ML) takes the customer touch point, tracks the activity in real time and predicts the next best action based on user activity. Machine learning predicts user future needs based on the history which results in up-selling and cross-selling opportunities. The system even triggers hyper-personalized notifications to CSR to share with the customer while the customer is still on the call like new products or service offering because this customer searched for that particular keyword in the past.
This is just one way businesses can use ML to improve customer service. Here are a few other ways you can leverage ML to improve the customer service experience:
• Shorten times to resolution on your cases. Implement smart routing to the right queue for folks and also use chatbots for those easier, self-serve problems.
• Increase customer satisfaction by helping customers route to the most effective agent. Alternatively, you can help those agents be more effective by recommending resolutions, articles or topics depending on the need of the customer. And use ML to help surface relevant customer history to the customer service agent.
• Reduce cost by proactively emailing users who look like they're searching for things on your website.
• Perform root-cause analysis. Try to mine data or look into models to see if you can -- based on models that are able to predict something -- dig into what is most predictive and use it as a way to improve a product or process.
So now that we know how to leverage ML in a customer service setting, what does it really take to build a system that utilizes ML?
First and foremost, it takes smart people who follow processes and use technology to design and build smart systems to provide the best customer experience possible. Based on my experience, the process plays a key role and in this context, I am talking about businesses prioritizing the digital transformation by using the latest and emerging technologies.
From a technology perspective, the journey should start with open source tools and packages when it comes to designing your systems. The primary reasons for utilizing open source is because of the wide range of options and that it helps keep the costs down. Tensorflow, H2O.ai and Microsoft Cognitive Toolkit are just a few examples.
In conclusion, it’s important to bridge business and technology. Once, people, processes and platforms are connected together, then driving ROI is easier. The same fact applies when attempting to improve the customer experience by using machine learning.
