Why AI Needs the Human Touch in the Post-COVID Service World

Print Friendly, PDF & Email

AI has been threatening to replace many of our jobs for some time. And there’s no doubt that it holds tremendous capability in a range of fields. But for now, there are still limits to what algorithms can do. And there’s one thing humans still do better than AI: understand other humans text-based interactions, especially on social media, and the feelings we try to convey. In a post-pandemic world, as firms grapple to understand and serve their customers, AI-only systems cannot replace the human touch.

Since the Covid-19 pandemic began, businesses have been adopting plans to ensure they can continue to service their customers despite new social distancing norms, increased customer uncertainty and restrictive lock-down rules.

These factors likely contributed to the increase in customers seeking service on digital channels. Research has shown a steady increase in customer complaints and queries on social media in 2020. In an analysis of some 950 000 social media posts involving six South African industries between 1 March to 2 April 2020, the rate of response to customer posts that require attention has dropped by 26.6% across banking, retail, telcos, insurers, pharmacies and ISPs. Similarly, in an analysis of telcos, banks and insurers in West, East and Southern Africa, the research identified that only half of customers who reached out to their service provider on social media received a response.

The value of unsolicited feedback that customers are providing firms on social media is increasingly being acknowledged by leading organizations. Those who remain unconvinced have probably been reliant on AI-only tools.

AI is still unable to accurately analyze our conversation. Sentiment analysis, the automated algorithm-driven task of determining the meaning and intent of a tweet or a comment on Facebook is still not yet consistently accurate. The algorithms that many organizations use for social media responses and analysis can pick up on keywords and identify trends but are not able to reliably determine sentiment or the topics driving sentiment. These algorithms are not able to grasp the nuances of our language like sarcasm, local slang, and emojis.

To accurately identify which social media posts require a response and uncover the sentiment and topics driving sentiment, a combination of humans and artificial intelligence is required. Humans are able to verify the sentiment contained in individual social media posts, and code each post for relevant conversation categories, and priority.  To do this effectively at scale, social media posts can be analyzed by people using crowdsourcing platforms, including proprietary crowd platforms made up of trained and vetted human verifiers. 

Contributors are remunerated for executing micro-jobs that include the verification and categorization of social media posts. 

Organizations, like banks and telcos, can then use this human verified data to make informed improvements on the customer journey.

For example, a large fast-food chain was able to streamline its menu by monitoring feedback of a new menu item on social media. While sales of the new product looked promising, as consumers were curious to try the new item, social media feedback said otherwise. The overwhelmingly negative feedback saw the chain act quickly to remove the item from their menus and avoid costly roll-outs and further damage to their customers’ experience. Typically, this feedback about the product would only have surfaced months later in traditional survey data or via focus groups. The real-time monitoring of social media, with human verified data, provided them with crucial cost-saving insights.

Having real people verify individual mentions is vital for delivering responsive social customer service. Humans can identify the priority of an individual post and unlike an AI-only tool, can identify prospective customer’s intent to purchase, or existing customer’s intent to cancel, seek service and gather valuable feedback on the customer experience.  This prescreening and prioritization of the most important customer interactions allow customer service agents to save time as they no longer need to spend hours identifying the important tickets in a messy queue. Instead, they can focus on responding to each customer’s specific request. This process dramatically improves the volume of customers firms can serve and respond to quickly.

With AI alone, organizations cannot understand their customer’s needs and service them responsively. Those organizations that have understood this will likely be able to deliver rapid and responsive customer service on social media and identify the valuable feedback customers provide. 

In a post-Covid world, organizations will need to ensure they are using tools that allow them to go beyond simple keyword matching to find the customer conversation that matters. Digital service should by no means spell the demise of the human touch, on the contrary, it reminds us of its importance. 

About the Author

Nic Ray is the CEO of BrandsEye and was part of the founding leadership team that built and sold an African digital agency group (Quirk) to WPP. Today, he is helping grow BrandsEye into a global customer data business that is changing the way companies understand and engage with their customer base.

Sign up for the free insideBIGDATA newsletter.

Join us on Twitter: @InsideBigData1 – https://twitter.com/InsideBigData1

Speak Your Mind