ABOUT THIS EPISODE
My wife and I needed a new accountant. Ours had sold the practice, and we wanted someone who could handle more than just tax prep. So we did what everyone still does: asked friends for referrals. We got several names and reached out to three of them.
The first firm demanded that all inquiries go through email. I filled out the form, hit submit, and didn’t hear back. After a week went by, I tried again. No response. The second firm funneled you to voicemail. After two messages and no response, I moved on to the third firm.
And this is where I did hear back, and received the single best, worst response to an inquiry I've ever heard. I left a voicemail with details about who had referred me, our tax planning and accounting needs, and asked for a call back. Seventy-two hours later, I got one. This is it, verbatim:
"Yea, hey Tony. This is “Wayne”. Look, my partner and I are firing clients, and given you were a referral I don't want to have to fire you too. We're only looking for clients we can bill at least $5,000.00 per year. So good luck with your search. Bye."
Here’s the punch line: $5,000.00 per year is significantly less than we’ve spent with an accountant for more than thirty years. My wife was in hysterics listening to the voicemail and has kept repeating the “I don’t want to have to fire you too” line since. When I called my friend who gave the referral, he couldn’t stop laughing either.
Here’s where my professional life and my personal life converged.
My wife turned to ChatGPT with a prompt laying out exactly what we needed. We got a highly curated list, with a detailed analysis of each firm, why it was selected, and which partner to ask for by name, all mapped to our specific requirements. None of the word of mouth firms made the list.
The referral is still human to human. The vetting just went machine
So why tell a story about one rude voicemail? Because I've spent the better part of a year writing about how agentic AI is changing how B2B buyers find and choose the people they work with. I didn't expect to become the case study.
Referrals still built our long list. They stopped being the thing that got anyone chosen. Word of mouth isn’t dead, it’s been demoted to a scouting job. A friend’s name used to do two things at once: get you on the list, and earn enough trust that you stopped looking any further. AI split that job in half. The referral still gets you the introduction. The vetting and matching, the actual choosing, runs on a machine that never met your friend, owes him nothing, and doesn't care whose name he gave you.
In AI Has Changed the Way B2B Buyers Buy, I used a survey of 226 B2B buyers to show third-party sources gaining ground everywhere, vendor-produced content losing it. I wrote the piece. I had no idea I was describing my own kitchen table.
Turns out I was the textbook example. That same survey found buyers newest to a category trust AI’s answer without checking it at four times the rate of buyers who know the category cold, 56 percent versus 14. My wife and I were newcomers, not to accountants, to this specific mix of needs, a broader set of capabilities than our last advisor covered. We were exactly the buyer most likely to take the machine’s list and start dialing without a second opinion. We did. The list was good enough that I’d do it again.
If you run a business that sells on relationship, referral, or reputation, you are the accountant in this story, whether you know it yet or not. A warm introduction still opens a door. It no longer predicts who walks through it. That's the difference between discovery, which built the internet era, and matching, which is building this one: state your criteria, skip the search box, and something else hands you a shortlist already narrowed to fit. The firms landing on it aren't the ones with the deepest referral network. They're the ones whose actual differentiation, who you're built for, what you charge, what you're genuinely good at, is stated somewhere clearly enough for a machine to match it against a buyer's specific requirements. Visibility in these tools is table stakes now, and like most table stakes, having it doesn't win you the hand. Not having it just means you never get dealt in.
The old failure mode was a bad referral: someone vouches for you, you don’t deliver, word gets around. The new failure mode is quieter. Nobody calls to tell you they’re firing you. You just stop getting the call. Say what you will about the accountant who fired us before he’d even hired us, at least he had the nerve to say it to our voicemail.
Word of mouth isn’t over. It just stopped being the last word.
The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.
”Uphoff on Media” is published by Tony Uphoff, Founder and Managing Partner of Uphoff Advisory: a strategic advisory practice for founders, CEOs, and investors in B2B information, marketing, and technology. The businesses that drive business.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit tonyuphoff.substack.com