So what do you think?  Are Computerized A.I. Telemarketing and VRS systems on the rise?

We all know what Telemarketing is, but what is VRS?  VRS stand for Voice Response Systems, and basically is the automated response and menu system you get when you call your bank or other large companies.  It’s the front line of such systems.  While VRS systems are not horrible, most of us like to talk to a real human being.

Have you noticed lately that you’ve been receiving telemarketing calls for surveys and the like where the agent sounds human, but after the call ends you realized that you questioned whether the person you were speaking to was a live human?

The call actually sounds like a normal person at the other end but something was a bit to perfect with the exception of an almost undetectable pause in their response?  By undetectable pause, I’m talking about a pause that maybe be 1 second or less.

Usually followed by the pause is an answer to your question that may have interrupted the speaker/A.I. computer agent during the call.

Unfortunately, I’m rude like that and interrupt the speaker as they are speaking.  The typical sequence is a small, almost undetectable pause followed by a perfectly spoken response with that matches the speed and tone of the original message perfectly. My argument is that such a response is not normal and highly questionable.

So what’s wrong?

An unscripted response should not be followed by a slight pause and a perfectly executed response to what the original pitch was.  To me this makes me think of a computer voice response system, but more sophisticated than what I’m use too.  The latest VRS system responses sound more like Apple’s Siri than a real human.

This makes me curious if telemarketers, customer service and other call centers are the next human labor pool to be attacked by technology?

So how did I recently make the jump from an A.I. telemarketing Survey VRS to the potential attack on a labor?

Well basically because if technology can create an A.I response system that acts and responds so human-like, they can eliminate the human element of a customer/participant services system.  Of course if the customer that called in request to speak to a supervisor, manager or someone else, the system could transfer such calls to a skeleton crew of humans to handle the hypothetical heavy lifting.

So what is my statistical estimate/guess on who such a system could affect the hiring of humans for customer/participate services?  Well it would depend on the size of the customer service / participate services department (obviously).  But let’s take a simplified customer service model and say that it has 100 people in their human resource pool.  I would guess that at least 60 of those 100 jobs could be A.I. agent based and the number would continue to improve as the technology improves over time.  I could even imagine that at some point during the following 5, 10 or 20 years after such a system is introduced that the number could even go as high as 90 A.I. agents instead of human agents.

Think 90% of a VRS as A.I. agents is too high?

It might be, it really depends on what the A.I. Agents are representing and the complexity of the responses.  But I wouldn’t be surprised that in the future A.I. Agents will query their responses from some type of private IT cloud VRS system per company.  I also would be surprised if the entire agent presence would be cloud based too.  Why have a cubicle for an A.I. agent when the entire A.I. agent instance is somewhere out in the cloud?

Such a cloud mechanism will have a direct impact on existing offshore Voice response systems.  After all the voice could be tailored to the area that the phone call is coming in from.  What do I mean by this?

Let’s say that you are from Scotland, and you call into a VRS.  Well, if the company has the bucks to buy the best computerized system, it should respond to the caller in the language or speech pattern that the native from the area would find the most agreeable.

What’s positive about A.I agents instead of human labor?

1. Privacy and security.  The A.I. agent won’t talk to friends about your account especially if you are even a little famous, but more importantly you are pretty much assured that your account information will remain private through the conversation with the A.I. agent.
2. No bias responses and pure systematic responses.  You could screen at the A.I. agent and you won’t hurt it’s feeling or make it do something nasty to your account.  Although at some level you must feel pretty silly yelling at a computer, no?
3. Cheaper products?  With all automation and computerization processes, such a change would mean that a company could lower their prices on their product a little.
4. Reduction of lawsuits and other problems with labor.  An A.I. agent doesn’t ask for a promotion.
5. Potential less errors and recorded sessions for review.  Some team will randomly sample conversations to continue to improve the responses by the A.I. agent.  Less error in that the A.I. system could eventually be better at deciphering what a hard to understand caller might be requesting.
6. Cost should be cheaper than the current offshore advantages.  With customer voice dialect and language, such a system is incredible versatile, even more so than the current offshore advantages.  As such systems become more prevalent and computer, A.I., cloud and robotics technology continues to become more powerful and cheaper at that same time such systems might become many more times cheaper than the cheapest labor pool.  I’m talking 1 to 100 or 1 to 1000…  Such a system that is could based could become scary cheap!

I have to admit, with the advantages that I outlined above, if I were a company, the appeal would be pretty obvious.

We are truly living in both exciting and depressing times!

Thanks for reading,

Don

If you are interesting in my original views on the way technology is hitting us, read: Pros and Cons of Automation.