Ask any C-suite executive about their company's use of social media and you'll quite likely find yourself being referred to the social media manager for more information. It's high time executives understood that social media is so much more than a marketing channel but a game-changing way to grow their businesses.
Most companies don't doubt that they should be using social media. It's considered part and parcel of how they push products and services and engage with their customers. Most employ dedicated social managers to generate and schedule content designed specifically for their various social platforms. And while these roles are important in maintaining a presence, social media is not just a tool for talking to customers but for listening to them too.
Many executives may have heard this story before, that social media and the sentiment data it contains holds great promise. And they'd be forgiven for rolling their eyes. Until now, accurately mining social media for sentiment data and then applying it in a meaningful way to a business has proven challenging, but with the advances in machine learning, alongside human intelligence, this is now achievable.
Accuracy is probably the greatest challenge to overcome. The natural language processing algorithms used to analyse the data have not been good enough at understanding human conversation. Online conversation is full of nuanced language such as sarcasm, slang idiom and emojis that machines can't reliably interpret. For now, humans still understand each other better than machines do.
That's where integrated human intelligence with AI becomes vital. The data which highly advanced algorithms analyse are distributed amongst a crowd of trained and vetted human verifiers. Each individual mention is distributed to multiple crowd members who code it for sentiment and the topics driving that sentiment. By using humans in our methodology, up to 40% greater accuracy can be achieved over the multitude of AI-only methods, and in turn provide training data from which machine learning algorithms are able to learn and improve.
Accurate sentiment data allows companies to bridge the gap that exists between their promises and the experiences and expectations of their customers. With the use of crowdsourcing, one is able to accurately capture all the relevant customer intention data, where customers express a desire to either cancel or purchase with a business or its competitors. Each mention containing an intention either to cancel or purchase is an opportunity for a company to retain an existing client or acquire a new one. But routing this data to the correct customer contact person in a business is vital. If you don't connect a disgruntled broadband customer to the correct person in a company, the accuracy of the data is wasted.
Beyond retention and acquisition, sentiment data can be used to inform customer experience strategy, improve operations and predict customer behavior. With near real-time sentiment data, airlines can assess how customers feel about their online check-in process, while banks can quickly adapt their mobile app.
Social media is a rich source of unsolicited customer intention data that businesses can use to improve customer experience and ultimately grow their customer base. But to fully realize these benefits, a boardroom-level paradigm shift is required so that social media be considered a tool for listening as much as it is a tool for talking.
If AI and natural language processors are of interest to you, join us at the ExCeL London between June 12-14, 2018 for the largest technology expo in Europe – TechXLR8. Take an in-depth look at the future of AI at the highlight AI event of London Tech Week, AI Summit, taking place at TechXLR8 between June 13-14, 2018.
Find out more about TechXLR8 here. Booking for the TechXLR8 Premium Pass is now open; book your pass here. You can also register for your free TechXLR8 expo pass here.
— Nic Ray, CMO, BrandsEye