What if you could understand when customers are not going down the right path, or when they are lost?

What if you could engage with them in real time — and even intercept carts on the verge of abandonment? The result would be a more successful web site and happier customers.

When carts are abandoned, leads are lost. The average ecommerce store loses over 75 per cent of its sales to cart abandonment, and some industries experience average cart abandonment rates as high as 83.6 per cent.

Your business systems can capture the data needed to stem this loss and engage with customers in a personalised manner — but connecting the moments as customers move through your channels is hard when you don’t know who to engage, when to engage, about which issue, or even how to connect with them.

Even when customers continue online self-service, a disjointed experience is a lost opportunity for businesses to offer premium service in a timely manner, present personalised offers, and increase brand loyalty.

Engaging in real time before frustration even begins saves valuable resources. Predictive engagement lets you guide known and unknown web site users and help them achieve what you have built your web site for: to buy, to sign up for offers, and to get answers. These types of insights let you see which shoppers are likely to buy or abandon, then engage with them at the best time — at their moment of decision.

Carl Jones, Predictive Engagement Lead ANZ at Genesys, noted that a customer’s online behaviours are a tell-tale sign of their intent.

“If they click on an advertisement, they’ve shown interest and intent. If they look at certain types of products, they’ve also shown strong intent. They really show intent when they start an application process, or place something in their cart.

“There are hundreds of data points available with each interaction. AI and machine learning are great at finding patterns, with the vast amount of analytics data available. They can then be used to decide when a customer is likely to buy or sign-up and when they are not.

“And then focus interaction efforts — such as chatbots, human agents and content offers to customers who are likely to convert, but who may have questions, who got distracted or need other assistance.”

Automation drives sales conversions

Applying predictive analytics to customer journeys better equips you to automate responses that shape the journey. Using proactive chat, proactive callback and content offers enables you to maximise customer engagement during real-time conversations. And, after visitors leave your web site, you can still route the complete journey context to agents and sales reps. This intelligent automation builds on an understanding of online behaviour that improves customer experience and boosts sales conversion rates.

Such mutually desirable outcomes require the efficiency of AI and the power of the human touch. For example, chatbots are very effective at handling repetitive, simple and mundane tasks. But customers in the throes of disrupted travel or other issues might need human intervention. The question is when to offer human intervention, and which agent is best to offer that support?

In addition, transferring from bot to human — and potentially back — should be seamless and based on who can do the task best to achieve the desired business outcome. Intelligent automation of tasks uses the best capabilities of human assistance, chatbots and micro self-service technology to do this, and builds stronger relationships with customers.

About the author

Brendan Dykes is the senior director solution and product marketing at Genesys who is a member of the Which-50 Digital Intelligence Unit. Members provide their insights and expertise for the benefit of the Which-50 community. Membership fees apply. 

DIU

LinkedIn
Previous post

Technology for good: fighting the growing problem of child exploitation

Next post

Digital Assets: Can you Afford U$90m for a Piece of Art?

Join the digital transformation discussion and sign up for the Which-50 Irregular Insights newsletter.