“Rabbit R1 is an interesting gadget. In any case, the marketing is a success.” But whether Rabbit meets the definition of a machine customer is debatable, says Mick Costigan, head of Salesforce Futures. Should customer service already prepare for the arrival of the machine customer or is it still too early?

 

Salesforce Futures is part of the Salesforce CEO office. The team of Mick Costigan helps the company with future issues: anticipating, imagining and shaping the future. To truly understand the concept of the machine customer, we need to go back to devices that capture that environmental information and send it back to the manufacturer or vendor, Costigan says.

“Think of the connected car, but also all kinds of industrial machines or white goods, although things are a bit slower there. The developments in AI mean that companies are not only receiving data from sensors, but also that machines are starting to act on behalf of a customer. So they become agents. There are still a lot of challenges and uncertainties in this area before these types of applications are fully mature. But still: go back to 2006, for example. Back then we had no idea what would happen in the field of interaction between customers and companies via mobile apps and social media. I think the impact of machine customers could be similar.”

 

Rabbit is different from Siri or Alexa: you can ask the latter two questions, but not give commands. However, Rabbit can perform actions across interfaces. Does that make Rabbit a real machine customer?

"Rabbit has a very simple interface. I don't think you can execute complex commands with it; I'm a bit skeptical about the use cases. Rabbit has shown a use case in travel, which can be complicated processes: do I fly today or tomorrow? Do I want to arrive in the morning or evening? What am I willing to pay for a flight? Which airline do I like to fly with? In which areas am I flexible and in which areas am I not? I have my doubts: can a simple interface handle this? Our devices can already do a lot, it is mainly the software that is not yet able to make all the considerations and reasoning in the background. But we are definitely moving in that direction.”

 

Do you think that consumers will ultimately dare to delegate processes such as purchasing decisions to devices and software?

"I think so. The signals are already there that this will happen to some extent. Look at Amazon: that platform offers customers a subscription model, especially for almost all small recurring purchases. With 'subscribe & save' you get a discount and delegate autonomy. This works better if products are commodities or basic needs. The Amazon Dash button was a precursor to this. You could say that this gives consumers back time and attention, after it has been taken away by social media and mobile apps.”

“Delegating decisions about and carrying out transactions requires trust, especially if delegation means that you hardly have to worry about it anymore. Large platforms have already won that trust – consumers do a lot of business with them and the platforms already manage a lot of data about their customers. Apple is also in an interesting position. They are trusted because people use their hardware. They have a strong reputation for privacy. I think Apple is in a great position to leverage Calendar and other apps to power their own personal AI agent. Whether they also make that functionality available to others remains to be seen.”

 

What are the consequences for customer service when machine customers enter?

“Customer service is already hugely impacted by generative AI. I think when machines become customers, you won't have customer service talking to a machine in natural language anymore – that would be crazy. We're probably moving toward even more machine-to-machine interactions. Brian Arthur, an economist from the San Francisco Bay Area, wrote about this in 2011 as 'the second economy': When you go to an airport and put your credit card into a machine to check in, all kinds of different data starts flowing through machines about luggage, passengers, airplanes.”

“The machine customer will raise new questions about investing in customer service: in technology or in people?”

“However, machine customers could make companies' services even more efficient and better. Because no matter how much we complain about customer service, customers are often difficult for agents to understand. We have to tell them everything. Imagine all that information being conveyed almost instantly, precisely tailored and in much more detail. Where clarification can be requested on both sides. AI has the patience to pay attention to details, something humans don't like to do. I think it can be very beneficial from an effectiveness and efficiency point of view. The machine customer will raise new questions about investing – in technology or in people in customer service roles. In addition, protocols will need to be developed so that CRM applications can safely interact with machine customers. We will also have to play a role in this.”

When will the machine customer make its appearance?

The rumors that OpenAI in collaboration with former Apple designer Jony Ive completely new kind of mobile phone would develop would give food for thought, according to Costigan. Above all, he predicts that it will become even more important for companies to strengthen the relationship of trust with their customers. Many other things – infrastructure, privacy rules – are already well developed.

When will machine customers arrive? Hard to say, Costigan thinks. “But did we expect a few years ago that we would be here now with GenAI? When it comes to the future, there are some dystopian images. But I think there is also something to be gained in terms of regaining attention – which has been hijacked by social media during the last wave of technological innovation.” (Ziptone/Erik Bouwer)

Featured, Technology Salesforce