In 2000 You Asked If You Needed a Website. Now You Are Asking If You Need an AI Employee.
2026-11-17

In 2000, the question a business owner asked themselves was whether they needed a website. A few early adopters built one, everyone else watched and waited to see if there would be an ROI.
Ten years later you could not open a business without one. Not because a law passed. Because customers stopped looking anywhere else.
I think about that shift a lot right now, because I am having the modern version of it about three times a week. Somebody asks me whether they actually need an AI employee, or whether it is something the big companies do until the fad passes.
It is the same question, about the same kind of change. And I have been on the wrong side of it before. I did not connect two systems in my own bike shop until 2012, when I hooked QuickBooks up to MailChimp through Zapier and realized independent software could talk to itself. I was late, by the standards of the day, and I only did it because a problem got bad enough to annoy me.
Every business technology goes through three stages
Look at what happened with websites, and you can see the shape of it.
Stage one is optional. A few businesses do it early, mostly the ones who like technology. Everyone else watches and waits. Nothing bad happens to the people who wait. This is the comfortable stage, and it is comfortable on purpose.
Stage two is expected. Customers start assuming it. Not asking for it, assuming it. If they hear about you and cannot find you, they quietly move on and you never learn why. This is the stage where the waiters start doing it badly and in a hurry.
Stage three is required. Now it is table stakes. Nobody gets credit for having one. You just cannot play without it.
Here is the part people miss. The move from stage one to stage two does not come with an announcement. There is no day where the market tells you that the thing you skipped has become the thing you need. It shows up as a slow leak: fewer calls, prospects who already know a competitor, a sales conversation that starts further behind than it used to.
Where AI employees actually are
I would put AI employees at the end of stage one and the start of stage two, depending on your industry.
Most owners have played with ChatGPT, gotten a decent email out of it, and gone back to doing admin at nine at night. They have not had a machine answer their phone, draft their proposals, or chase a follow-up without being asked.
In companies with a real operations function, it has already moved. The ones I work with are not debating whether AI does useful work. They are arguing about which process to hand over next.
So no, it is not mandatory today. That is exactly what makes now a good time.
Why waiting is more expensive than it looks
There are two costs to waiting, and only one of them is obvious.
The obvious cost is the hours. If a machine can take three hours a day of quoting, scheduling and follow-up off your desk, and you wait two years, that is roughly 1,500 hours you spent doing something you did not need to do. You cannot get those back and you cannot invoice for them.
The less obvious cost is the foundation. This is the part that actually decides whether you will be able to adopt this later.
An AI employee is only as good as the information it has about your business. Your pricing rules. How you write a proposal. Which customer gets special handling. The list of things you check before a job goes out. Most owners carry that in their head, and it has never been written down anywhere.
The businesses that start early end up with that written down. The businesses that wait until it is mandatory have to build the system and document their own business at the same time, under pressure, with a customer watching.
That is the real gap. Not the technology. The two years of written-down knowledge.
The honest counterpoint
I do not want to sell you on a trend, so here is the other side of the website story.
Lots of businesses launched websites in 2005 that did nothing. Brochure pages nobody could find. The website alone bought them nothing. What they needed was a page that earned its keep.
AI employees will go the same way. A lot of money is going to be spent on agents that answer questions nobody asked, in a chat window nobody opens, with no connection to the actual work. That will produce a lot of expensive disappointment and a few very loud skeptics.
So here is the question worth actually asking. Do you have one piece of real, repetitive, checkable work that a machine could take off your desk and do consistently?
If you do, start there. One workflow. Measured. Not a transformation project.
If you do not, wait. You are not missing anything yet, and buying a chatbot to feel current is how you end up bitter about the whole category.
What I would do this month, either way
Write down the three things you did by hand last week that you have done a hundred times before.
Not the interesting ones. The boring, repetitive, rule-following ones. The quote you rebuild from scratch every time. The follow-up you send when you remember. The fifteen minutes of checking before anything goes out the door.
Then pick the one where a mistake would be easy to catch. That is your first candidate, because it is the one you can hand over without lying awake about it.
That exercise costs you twenty minutes and it does not require buying anything. It also happens to be the exact work an AI employee needs in order to be any good.
What I would say to that owner now
Get the agent, but not because the market is about to force you into it. Get the one workflow done because the exercise of describing your own business precisely enough for a machine to run part of it is worth doing on its own. I know that from the 2012 version of this. The integration itself took an afternoon. Working out exactly what I wanted it to do took longer, and that was the useful part.
That is where I land on AI employees too. The machine is the easy part now. The hard part is being able to say, precisely, how your business runs.
Start there and the rest is a few weeks of work.
If you want to talk through which workflow to hand over first, I offer a free hour. We map the repetitive work in your week, and you leave with a written list of what could come off your plate, whether or not you hire me. If you would rather see the whole thing before booking a call, the offer and the price are published on the site.