From Cost Saver to Revenue Driver: Conversational AI Comes of Age

Recently Amazon launched its ‘shopping assistant’ Rufus in India. A GenAI powered Agent that can browse Amazon’s (rather overwhelming) product catalogue and help customers find what they are looking for.

Along similar lines, both OpenAI and Google have touted their visions, or even launched Agentic AI products or features to help customers buy what they want, how they want.

For years the business case for Conversational AI has focused on reducing costs. Mainly the cost of human customer service. But what about generating revenue? 

CDI’s Jasper Klimbie points out: “While it’s true that there has always been the promise that a chatbot could be a marketing and sales channel, success stories were rare. Truth be told, the technology wasn’t really ready.”  

Sales flows seem predictable, but human behaviour can quickly knock a journey off track. A company with a diverse array of products would often struggle in realising successful conversations about those various products. Training a language model to understand all the nuances of the ways customers describe what they are after is very, very difficult.

But with the rise of Agentic AI, and especially with the ability to search databases and access pre-defined knowledge, this is changing rapidly.

A great example of this can be found on XXL nutrition’s website. This CM customer sells nutritional supplements. They are fully aware that customers are often bewildered by the many options - products that are similar yet different. Their Agentic chatbot helps customers get what they need quickly: avoiding the dreaded ‘death by browsing’. 

As CM’s Saskia Dekkers explains: “It’s not a RAG solution, the Agent uses a specific lookup tool to produce more detailed and accurate information.” 

At the point of writing, the customer still needs to complete the actual transaction through the website purchasing flow, but it’s on the roadmap to complete sales in the chatbot - and it is already technically feasible. 

Saskia expands: “Only a few years ago, without the power of Agentic AI solutions, building a chatbot like this would have been an effort of heroic proportions - if possible at all.”

Proactive marketing and sales

Akin to a website based sales assistant, we have talked about proactive Agentic possibilities for years. But if you are going to reach out to (potential) customers, you need the experience to be perfect: you only get one chance to make a good impression.

As Saskia explains CM’s vision: “Be where your customers are. The best channels for pro-active Agentic marketing and sales efforts are those that surface on mobile phones. WhatsApp, RCS and SMS. And, because Agentic AI is so good with language and can make a conversation feel natural and effortless for the customer - proactive use cases are showing more success than ever before.” 

Here’s some use cases to consider, that CM has deployed with success for their customers:

• Sending offers and deals - classic marketing outreach with behaviour design to increase conversion.
• Advertising and brand engagement - building brand awareness and creating loyal customers.
• Customer updates - proactively reaching out to customers regarding order status or pending deadlines for example.
• Assist the buying journey - being able to quickly resolve doubts and hesitancy during the buying journey.

Recover abandoned carts - reminding customers with well designed nudges that they have products waiting for that final decision.

One startling statistic that CM shared can be seen below. Who wouldn't want to benefit from this kind of conversion?

So, if your business has been considering growing revenue through Agentic AI, rest assured, the moment is now. A final word of advice on behalf of CDI: make sure that you thoroughly design the experience you hope to offer your customers - even if, or especially if you are letting an Agentic AI do the talking.