Companies do everything they can to keep you from talking or texting to a real person, and the automated assistants literally (and I’m using that word literally) are never able to handle whatever my problem is, and it can sometimes be nearly impossible to talk to a real person.
I, on the other hand, have had multiple cases where the AI I was talking with was, in fact, able to fix the problem I had. And let’s be fair, most of the humans they could hire wouldn’t be able to fix problems, either.
I’ve found they help about 50% of the time, but it’s the truly automated AI that are less than useful. If it’s a chat “bot” that’s a real person, they can be useful.
The Spectrum chat I had yesterday was useless. What I needed answering wasn’t in their script.
But if it wasn’t in their script, then the human couldn’t have helped, either. What do you think humans were doing, back when they did this job? They were following the same script. Except that now, the script can be more complicated and comprehensive, and the “customer service rep” never screws up on following it, and they usually have a more understandable voice, too.
It seems like whatever my particular problem is it’s never in the script. Generally what’s in the script is stuff I could have figured out myself.
The problems I run into are often forms that won’t accept valid information or a system that suddenly doesnt recognize my account. The automated systems often tell me to do things I’ve already done.
In the most recent issue I’m trying to get AT&T to unlock my phone. It’s telling me that it can’t because my phone isn’t compatible with AT&T. I’ve been using the phone with AT&T for months plus it’s on their list of compatible devices.
All the automated scripts do is lead me to the same dialog that tells me they can’t unlock it because it’s not AT&T compatible!
Their automated scripts don’t have choices for this kind if inconsistency but they won’t connect you to a real person.
I finally found a form where you could actually enter a problem description which wasn’t vetted by a script. Hopefully I’ll hear back from them.
Edited to add: Chronos, if I talk to a person then usually they can either solve my problem or pass it on to someone who can. You make it sound like if it’s not in the script then it’s unsolvable.
And I take issue with the idea that a bot is just as likely or even more likely than a person to solve a problem. Someday yes. But not quite yet.
I like automated assistants, with one caveat. As long as there’s an obvious way to talk to a real person, it is all good. Many times they can’t help me and I need to talk to a real person, so I need a way to do that.
But otherwise they are great. If I’m able to do what I need to do with a phone tree, I can generally get it done much faster with an automated system than talking to a real person.
Case in point, at my work I deploy many iPhone devices. I need to activate them by calling a phone number. The automated process takes less than a minute. If something goes wrong and I have to talk to a real person, that’s probably 30 minutes to an hour.
I’ve found that just pressing the zero button repeatedly works most of the time, even if the IVR says “we need more information before we can connect you”.
It depends on whether or not the person has any common sense, and/or whether or not they have any actual tech knowledge, and/or whether or not the company allows any deviation from the script.
There was one time where I was having some trouble, and I had already rebooted the device in question. The human rep that I got insisted that I reboot before he would move on to the next item on the list, even though I was careful to tell him that I had already done so.
Oddly enough, the only thing that the extra reboot did was waste our time.
I’m guessing their statistics say that people claim to have rebooted often enough that insisting you do it is an overall time saver for them.
On the other hand, I’m pretty sure the last time I had to call Google Fiber (which is rare), they believed me when I said I’d already rebooted everything.
Keep in mind the person on the phone could lose their job for not following that script absolutely, and that in my experience doing tech support, it’s not at all unusual for callers to lie.
Google Fiber TV - one of my TV’s wouldn’t connect. I was on the phone with them for maybe 40 minutes (I love that Google will spend all the time necessary on calls). They never actually asked me if everything was plugged in - I think I exuded competence initially, explaining everything I’d already tried.
Eventually, it got to “let’s try rebooting everything again, again”. When I went to reboot the particular TV box, this time I noticed that it wasn’t warm to the touch, and then I noticed it wasn’t plugged in (old crappy outlet that didn’t ‘grab’ the prongs). Plugged it in, and, amazingly, it worked.
In my defense, the TV interface boxes had a very subtle red light that indicated they were powered on, and the box was in a location that was hard to see (bluetooth connectivity so didn’t need line of sight).
I just tried to use the AI for Spectrum the other day because I needed the answer to a specific question. One of the first questions to me had three choices. My question didn’t fit any of those choices exactly, and the AI wouldn’t let me proceed without choosing one. I chose the closest, but it had no subsequent options that fit my question. No way to enter text either. So, I had to find a way to talk to a person. Had my answer in seconds.
Do you guys think those companies don’t want to have a fully staffed highly competent help desk?
Those people simply do not exist.
Some drones that walk you through a script is the best they can do. And even that is getting harder, AI for such stuff is the only way to go. Bettter get used to it.
IME Google Fiber has a fully staffed, competent help desk.
My company has one for internal (not sure what the external is like).
The companies that don’t have decided to not pay for one, having determined that whatever service they are providing is sufficient to not drive away too many customers.