

I am fortunate to have moved to a climate where the heat is less severe and when it is hot it tends to be dry-ish. My house does not have AC so we put a big exhaust fan on the top floor and crack a window downstairs. Works so far, but we have some small portable AC units for the bedrooms just in case we need them.
I design contact centers for a living. I have done so for almost a quarter century now, until very recently I only had worked for Fortune 200 companies (moved to the public sector which is a nice change of pace).
A quick bit of jargon definition: We refer to various means of communication as “channels”. A contact center is multi-channel if you can reach it by more than one channel (i.e. phone, SMS, chat, email, etc.). It is considered omni-channel if you can switch between these channels (supposedly seamlessly, but see below).
This article gets several points dead on and misses several more. Here is my professional take, make of it what you will.
Call centers are expensive. Licensing and software costs are very high. There are few vendors who offer scalable omni-channel offerings and the licensing costs end up being exorbitant. And you need omni-channel contact centers because:
Phones are the least efficient way to service customers. An agent can only be on the phone with a single customer at a time, but they can staff around 6 chat or email sessions simultaneously. For a customer, this devoted attention is a boon, but for a company it’s very costly because Agents, even poorly paid ones, are the most expensive part of your contact center if you are paying benefits, and if you aren’t you will not get good agents.
Agent turnover is very high. Agents are poorly paid and their job sucks. They are driven by metrics that are poorly thought out, intended to drive efficiency but ultimately create poor behavior; the article gets this very correct. A lot of poor service you get is caused by agents trying to hit impossible metrics. Don’t blame the agent, the managers are the problem here.
The technology has gotten better - and worse. VOIP infrastructure radically reshaped contact center design and the migration to CCaaS reshaped it again, with some good sides and a lot of bad sides. Telephone technology is an aging tech with a substantial demographic issue. I am consistently the youngest member of my teams and I have been doing this for almost 25 years. Expertise is aging out of the field and taking a lot of knowledge with them. Further, the number of disciplines you need for expertise has dramatically increased. It is no longer enough to just know CCNA-level networking, wiring, PSTN tech, linux and windows servers administration, codecs, basic related legal knowledge (wiretapping laws, Ray Baume’s Act, TEHO laws in India, etc.), design and infrastructure theory (like Poisson distribution), but now you also need to know Kubernetes, docker, ESXi (or equivalent), AWS, Azure, etc. It’s a lot and nobody can know it all, the complexity of modern design and no education program to get there means there’s just a lack of comprehensive understanding of the technology at a pretty fundamental level for most people trying to design and maintain this stuff. The result? A system designed around 99.999% uptime is now failing to meet that SLA, hell some vendors won’t even promise it anymore but most will just lie and claim that they do. So there are reliability issues.
AI. This one hits pretty closet to home for me because of a personal experience so a quick anecdote: at one job, I had a spirited discussion with the head of our IVR technology group over how effective AI would be at reducing call volume into the center. He initially had great success, reducing call volume by ~30% in the 6 months. He received accolades and commendations, a big bonus, he was riding high and honestly he deserved to be. The problem, and what prompted my attempt to intervene, was his promise to continue that trend, predicting that his AI tech could reduce human-required calls by 60% within the next 2 years.
To me, this was madness. His initial success was because he moved the payment system into the IVR instead of having agents do it. This is a no-brainer. Computers are quite capable of taking payments or listing basic account information, but more complex tasks involve a much greater up-front cost in technology development and we didn’t have that budget, it was a massive over promise and I told everyone who I could to not take his estimation seriously. Unfortunately, he had a PhD and I am a college dropout, so they listened to him and cut 50% of their agent count via attrition. The results were predictably disastrous and the company hasn’t yet been able to fix it years later (thankfully, I left that place).
I don’t think this is intentional per se. Having been in numerous meetings with leadership about contact center issues, I can say that they are just as upset by poor customer service as you are. There is no top-down effort to make your life suck. But line must go up and contact centers are always cost centers which means companies hate them, they don’t view customer service as integral to making money despite understanding that angry customers will leave them so there is a constant budget short-fall. The issue isn’t someone at the top thinking “If we treat our customer poorly enough they will stop calling and we’ll save money!” It’s just standard corrosive capitalism creating perverse incentives that make everything worse. It’s a systemic problem.
Anyway, that’s my view for whatever it’s worth. I am glad to be in the public sector now, which has its own issues, but at least everyone is focused on actually providing service because the service is the value.