Applied AI: Automating Cold Calling

TL;DR My new “Applied AI” series shares practical examples of how specialised AI systems can augment human workers. The series starts with the benefits of automating the dreaded “cold calling” process. You can jump to the end of the article to watch a video of an AI bot qualifying a lead and handing a hot prospect to its human master. 

Augmenting human workers with specialised AI systems will push humanity onto the next S-curve of productivity. It does not require big AI transformation programs or billion-dollar investments anymore. With the right strategies and tactics, leaders of small and mid-sized businesses can be at the forefront of this revolution.

This article is the start of my new series “Applied AI”. I will be sharing practical examples of AI implementations by my clients, partners and other AI practitioners. My goal is to demonstrate how AI can be applied as leverage to human capital by companies of any size. The target audience are my fellow CEOs and business executives. Don’t expect too many technical details (if they interest you, please message me privately). 

A mutually unpleasant experience 

Your phone rings, you run out of a meeting to answer, and here it goes: “Hello Mr Simon, I am from company X we are the premier provider of service 123 can I walk you through on how we can help you to achieve ABC?” Don’t we all love these calls? Guess what, the person calling you is equally excited. 

In my research, I collected feedback from over 400 sales professionals, contact centre agents, financial advisers and business development experts. 93 % of them consider cold calling the worst part of their job. They equally dread calling new prospects as well as up-selling existing clients. They have to deal with constant rejection and occasional verbal abuse. Cold calling is a driving force of emotional misery & human frustration. Why do we continue the practice?

The business case for cold calling

Conversion rates for cold calling and texting via popular messengers range from 2% to 20%. Lead database quality is the main success factor. For comparison: Email newsletters sent to subscribers produce less than 1% on average. It looks like there is a strong business case for cold calling.

Let’s look at the cost side of the equation: According to my research, a financial adviser can spend up to 60% of her/his time cold calling. He/she will connect with an average 30-40 leads on a good day. Assuming total managed cost (“TMC”) of $ 10k/month, this results in $ 8.3 to $ 11.1 per lead qualified (not customer converted).  

The next factor to consider is the opportunity cost. While the adviser is qualifying leads, he/she is not closing new deals and providing services to paying clients (=main revenue-generating activities). 

The AI-augmented financial adviser

How could a specialised AI system augment the sales process of the financial adviser from our example? To illustrate this, I had my team record an anonymised video of one of our conversational AI systems in action.

As you could see, the AI pulled the client’s portfolio information. It recognised the sales opportunity, confirmed that the client is interested in a conversation and booked a meeting with the adviser. It also integrates with the advisers’ calendar solution. When he/she starts working the agenda will be filled with pre-set client appointments. The dreaded cold calls have been taken care of by the dutiful AI assistant. 

Of course, there are also scenarios AI assistants will be unable to handle. Uncommon, complex customer enquiries and sarcastic responses are still beyond the capabilities of most commercial systems. Those have to be transferred to a human for review and processing. 

This last part is critical. I would strongly advise against having unsupervised AIs run any part of anyone’s business. Similar to a new junior employee, an AI assistant has to be trained and be able to escalate complex issues to their seniors. 

AI-augmented business growth

Advanced AI systems can call/text up to 3.45 million leads per day. An unaugmented human might reach out to 100-250 tops. The AI responds within seconds. 24/7, no sleep, no vacation, no sick leave. 

Across multiple experiments & production deployments, I found that AI-augmented cold calling/texting against publicly available databases produces an average response rate of 12.93%. On pre-classified data sets, e.g. existing customers, the rate goes up to an average of 68.5 %. These rates are below what a human agent can achieve. The AI gains an advantage with scale & cost factored in. In average we see the cost per call reduced by a full order of magnitude.

Having a team member that can do practically unlimited calls poses an interesting question for a CEO: How many new customers do you want?

Don’t start laughing just yet and shout “as much as I can get!”. Imagine your AI reaches out to 3 million potential clients, and only 1% come back to you with “interested, I like to meet you”. Can your human team follow-up and discuss deals with 30,000 individuals now? But as you might rightfully say: “This is a happy problem!” – I would agree. 

Conclusion

I trust this example illustrated how conversational AI provides leverage to existing human capital. With AI as a force multiplier, even small sales teams can achieve an outreach that dwarfs their better-staffed competitors. AI changes the game from “who has the biggest sales team” to “who has superior strategic agility & tactics”. 

If you enjoyed this article and like to learn more, please do not hesitate to connect and reach out to us. 

Share This Post

Share on facebook
Share on linkedin
Share on twitter
Share on email

More To Explore

Julian Seow

The Future of Online Petitions

There exists a problem with modern government that up until now has seemed near-impossible to address. In a traditional parliamentary democracy, citizens will vote on

Read More »