Will A.I. Ever Pass the Turing Test for B2B Sales?
And is that really what we want?
The Turing Test, for those of you who don’t know, is a test of whether a robot (artificial intelligence) is advanced enough to trick a human into thinking that it is also a human. But, more on the Turing Test later…
Life is moving pretty fast in the B2B sales arena, and companies are always looking to accelerate and optimize anyway they can.
In a world of diverse corporate buying teams, lengthening sales cycles, and demanding customers in a competitive marketplace, those same companies are having to lean more on artificial intelligence (A.I.) to solve complex sales issues and increase efficiency.
But, will there ever be a point when we no longer need an actual salesperson in B2B Sales? Will A.I. ever be so advanced, that in certain sales scenarios, you won’t be able to tell the difference? Are there A.I. programs out there right now that are on the brink of passing the “Turing Test” and replacing humans entirely?
We went looking for real answers from real people (not robots).
First, What Is the Turing Test?
In 1950, English scientist Alan Turing proposed a test of computer intelligence or artificial intelligence. He said it was meaningless to ask if computers could think, but it was important to ask what it would take for a computer to fool us into thinking that we were talking to another human. This test of artificial intelligence is now known as the “Turing Test.”
To pass the Turing Test, a computer would have to pass a test where an interrogator asks questions on a computer screen and receives answers from two subjects – one a person and another a computer program (illustrated to the right). Turing estimated that by the year 2000, artificial intelligence will have advanced to a level where an average interrogator will not have more than a 70 percent chance of making the right identification after five minutes of questioning.
Turing came pretty close with his estimate. A version of the Turing test was definitely passed by a robot named Eugene in 2014.
Is A.I. Closing the Gap on Passing the Turing Test for B2B Sales?
There is some evidence that A.I. is almost there. Consider these two data points:
In 2016, CenturyLink bought a program called Conversica, which creates and manages an A.I. sales assistant. The program’s avatar, Angie, sends out thousands of emails each day, then evaluates human responses and turns over the best leads to a human salesperson. The prospects who receive emails from “Angie” have no idea they are interacting with a program. Clearly, CenturyLink is not alone.
Google has even taken the limits of A.I. one step further by creating a program that speaks to people in real time. At Google I/O in 2018, Google CEO Sundar Pichai played a recording of a computer program named Duplex that called a hair salon to make an appointment. Apparently, the people at the salon on the other end of the call couldn’t tell that they were speaking to Google’s digital assistant.
While these are impressive, these are both instances where A.I. was used to accomplish a specific task with qualifying variables. Greet the customer, translate their responses, and book appointments for the salespeople or customer.
If the question is whether an A.I. chatbot or similar program can do transactional processes for sales today, the answer is pretty clear, “Yes.”
But, can A.I. handle all the deal complexity, situational fluidity, and nuances of a B2B buyer-seller relationship? Probably not. But progress is being made and some of the most innovative ideas and companies are working to get us closer to an A.I.-driven sales process.
The next big question of course: do we really want that? Well, your customers will tell you. The future of business growth is providing the best experience for your customers because that’s the clearest way to differentiate yourself. So if A.I. can help with that task, then it will be something we want to see in the sales process.
Here’s what business leaders are saying:
- Dana Therrien, sales operations expert and leader of SiriusDecisions’ sales operations strategies research practice points out that, “Sales transactions can be grouped by how buyers make decisions–as individuals, groups or by consensus. For individual buyers deal sizes are usually smaller, non-complex, sales cycles are short, and terms can be reached without much personal interaction. A.I. has already and will continue to replace salespeople for these types of transactional sales. A.I. will not replace salespeople who sell to groups and consensus buyers. Instead, A.I. will coach them and it will make them more effective and productive by reducing their administrative burden (i.e. no more manually updating the CRM) and it will empower them with guided insights and suggestions.”
- Ilan Kasan, Co-Founder, and CEO at Exceed.ai, believes that, “High touch B2B sales will never pass the Turing Test. [Deals] are simply too complex and every sale has its own nuances and specific details. However, sales development will, and already has, started to be replaced by robots who will work alongside reps and perform manual, error-prone tasks.”
- “The goal of artificial intelligence in B2B sales shouldn’t be to replace, but rather supercharge your sales reps,” says Oleg Rogynskyy, Founder, and CEO of People.ai. “We’ve found that A.I. can be leveraged to drive predictive insights for both sales and marketing, allow leaders to manage their teams with data, and help remove repetitive tasks like data entry into their CRM and reporting so much so that we’re able to give salespeople more than 20% of their time back every week. Sales leaders should be looking for creative ways to empower their sales and marketing teams, not replace them.”
- Meanwhile, FloydHub represents the next generation of deep learning platforms for data science teams. Their customers are already using deep learning algorithms to automate their sales and marketing pipelines. Co-Founder and CTO of FloydHub, Narendran Thiagarajan, believes that, “The biggest advancement in the area of deep learning is natural language understanding and processing. Since most of the sales and marketing happens over email, this area will be first to be fully automated.”
- “The gap would come on the ability of the A.I. to really understand the interplay of a deal and what’s happening behind the scenes in an organization,” explains Casey Hill, Senior Account Executive at Ontraport. Casey gave the example of having, “A marketing person excited about the product or service, but feeling pressure from engineering to have a seamless integration. Furthermore, a salesperson might get a simple vibe from a prospect that it could help their chances of closing the deal if they go above and beyond with a special gift or taking the client out to dinner.”
It’s certainly worth keeping a notification alert set on developments in A.I. and B2B buyer surveys, but it appears that smart programs just aren’t smart enough yet in terms of displacing strategic enterprise B2B sales experts. For lead development and appointment setting sales reps (SDRs and LDRs), though, the dream of relaxing while a robot does your job for you might already be a reality.
Hear from Rajit Joseph, People.ai’s VP of Product, on this topic via TechBytes here.
Find out how People.ai helps organizational leaders gain visibility into their sales and marketing teams, drive predictive insights, and increase the overall productivity of their company by scheduling a demo of the People.ai solution today.