Why Is It So Hard to Reach a Person for IT Help?
Every time someone’s AutoCAD glitched, he'd open a chat window, type out the problem, and, in his own words, "hope the AI would help."
It rarely did. By the time we spoke, his voice had that flat, resigned sound of someone who stopped expecting real help a long time ago.
How did getting help become another task to work through? And how do you make sure your next IT partner won't leave you in the same position?
I was on a call with a principal whose architecture firm had launched barely a year ago and was already landing major wins. He'd been handling the tech himself until it stopped making sense.
He'd spent so much time trying to fix things himself, wrangling with AI support bots, that it sounded like he expected a similar experience with an IT partner. Real people probably existed somewhere, but reaching them was another story. He'd accepted that AI would sit in the middle and he'd have to live with it. That's the part I consider unacceptable.
I'm often the person recommending AI to help teams remove bottlenecks and make everyday work easier. Here, it had become one more obstacle between someone and the help they needed.
So when I told him there's no AI system between us and our clients, his eyes got wide. I explained that when he needs help, he'll talk directly with a person, in plain English. That's how we believe people should work together. He leaned into his webcam like he needed to make sure he'd heard that right.
You're Not Imagining It
I spent this past weekend at an IT industry conference, walking a vendor hall that felt like a preview of that exact future. Booth after booth. Tool after tool. Automation after automation.
In the vendor conversations I had, I heard plenty about software and AI but very little about good customer service or taking care of client relationships. Much of what I saw seemed designed to sit between you and the person who can actually help you.
The promise behind these tools is appealing: let AI handle simple requests so people can focus on harder problems. That's what persuaded us to try one ourselves about a year ago.
But our internal testing quickly left us frustrated and exhausted. We didn't even finish before telling the vendor no.
It used to be that you could keep pressing 0 to reach a person. With tools like these, even that feels harder. You're dealing with a smarter gatekeeper just to reach the service you're paying for. That isn't helping.
A few days before that conference, another prospect told me almost the same story, except she'd already lived through the switch. Her IT company had been solid for years, then somewhere in the last twelve months it turned into canned replies and long waits.
She wasn't ready to sign with anyone. She was still under contract elsewhere and wasn't shopping. She just wanted to know if real people still existed anywhere in this industry.
The next morning she sent one line: "Thank you so much for your time today. It was time well spent for me."
If you're the person your team turns to when technology stops working, you may recognize the feeling. You're chasing the IT company while everyone else is asking you for an update.
Why are you the one chasing a resolution when you've hired someone to handle it?
What Went Wrong in Our Testing
The product failed at every part of the job we needed it to do. It failed at solving the issue faster. It failed at suggesting a better approach. Most importantly, it failed at connecting someone with our team when it clearly couldn't help.
It didn't feel right, either. It felt like we were giving clients another tool to wrestle with instead of honoring the basics of IT support: one person helping another person.
Before we stopped testing, we measured the time it took to get an answer. In those tests, it took roughly twice as long when an AI system sat between someone and the help they needed, whether in sales, support, or account management.
Run the math on a $200 hourly billable rate. If support delays cost your team roughly three billable hours a week, that's $2,400 in potential billable time over four weeks.
And that doesn't capture the bigger cost: people who quietly stop asking for help because asking has become its own chore.
A support system isn't doing its job if your team would rather live with the problem than use it.
Tools Should Help the People Helping You
Think about construction for a second. Every trade has fancier tools than it did ten years ago, from tile cutters with built-in dust collection to laser levels. There's a higher tech version of just about everything.
But when you have a construction project, you don't go talk to the tools. You go to the contractor. That person talks to their crew, and the crew uses the tools. Tools serve people. People serve you.
That's exactly what you should expect from an IT company. Their tools should be working for you quietly in the background, never blocking your access to help.
To be fair, we use AI constantly, and I've built plenty of ways it saves our team time behind the scenes. When I use it to prepare something for a client, I review it before it leaves my computer. AI helps with the work. I stay directly connected with my clients.
The same expectation should apply to IT support. Fancier tools should never replace skilled, experienced people who respond promptly and have the time and capacity to care about your situation.
How to Find Out Before You Sign
You don't need to understand an IT company's software to find out what getting help will feel like.
Start with a few conversations:
Ask to speak with the people who handle sales, support, and your ongoing account relationship. In a small company, some may be the same person. What matters is whether you can meet the people you'll rely on and get clear answers.
Ask whether a short call with the owner is possible. At a small IT company, that can tell you a lot about how accessible the people in charge will be.
Ask this question: "AI is really interesting, and we're seeing some cool use cases. How are you using AI in your business?"
If their answer includes AI handling support requests or chat conversations, ask one more question:
"When I need a person, how do I reach one, and how long should I expect to wait?"
An optional AI tool that answers a simple question can be useful. Having to work through a bot before you're allowed to reach someone is a different experience.
Ask them to walk you through what happens when the AI's suggestions don't solve your problem. Can you ask for a person right away? Will that person see what you've already explained, or will you need to start over?
If they can't give you a clear answer, take that seriously.
What You're Actually Paying For
Picture the first version. You send in a request for help, and a person who knows your account picks it up. A question turns into an actual conversation, not a loop.
Now picture the version some of us have already accepted as normal. Real people exist somewhere, technically, but you rarely reach them, and every question costs you time you didn't plan on losing.
For the person putting out fires in your firm, that means more chasing and more fielding complaints from frustrated teammates. For the business, it means interrupted work and lost time that could've gone to clients and projects.
That's a meaningful difference to look for when choosing an IT company.
The tools matter. So does the team's ability to solve the problem. But neither means much to the person stuck waiting if they can't reach help when they need it.
When something breaks, can you reach a person who will take ownership of the problem?
You shouldn't have to wonder.
If you want to run this test on your current IT company, or you're sizing up a new one, I'm happy to talk through what to listen for. That's a conversation, not a pitch.