Word of Mouth Is the One Channel AI Can't Commoditize

Something I’ve been chewing on this month is what actually survives when everyone gets the same tools. AI has made a lot of marketing cheap. Ad copy, landing pages, subject lines, product photos, even a passable video ad. The stuff that used to take a team a week now takes a prompt and a coffee break. And when something gets that cheap, it stops being an advantage. It becomes table stakes.

So the question I keep coming back to is simple. If everyone can generate infinite content, what is the one thing that can’t be generated? My answer, after a lot of arguing with myself, is a real recommendation from a real person who has nothing to gain by lying to you.

Why AI floods the channels it can reach

Look at what AI is good at. It is good at producing volume in any channel where the message comes from the brand. Paid search, display, cold email, social captions, blog posts at scale. Those are all channels where you control the words, which means AI can control the words too. The predictable result is a flood. When production cost drops to near zero, supply explodes, and attention gets harder to win, not easier.

Consumers feel this. There’s a strange paradox in the research on AI-generated content: people will say they trust it in the abstract, then quietly disengage the moment they realize a message was machine-made. Researchers at the Nuremberg Institute for Market Decisions found that simply labeling an ad as AI-generated made people rate it as less natural and less useful, even when the content was identical to a human-made version. Transparency didn’t buy trust. It cost it.

That tells me the brand-controlled channels are heading toward a trust discount. The more the audience suspects a machine wrote it, the less weight it carries. And they will suspect, because most of it will be machine-written.

The one channel that runs the other direction

Word of mouth works on the opposite physics. Its whole value comes from the fact that the sender is not the brand. When your neighbor tells you which moving company didn’t scratch her floors, the message is credible precisely because she isn’t selling anything. You can’t fake that at scale without getting caught, and getting caught destroys the very trust that made the channel work.

The numbers have been stubbornly consistent on this for years. Roughly nine in ten consumers trust recommendations from people they know more than any form of advertising, and that figure has held steady even as every digital channel multiplied around it. AI didn’t move it. If anything, the flood makes a trusted human voice more valuable, the same way a handwritten note stands out in an inbox of automated ones.

This is the part people miss. AI doesn’t commoditize word of mouth. It raises its price. When the cheap channels get noisier and less believed, the expensive-to-earn channel becomes the moat. I wrote more about the credibility side of this in the trust recession, but the short version is that proof from other people is the currency now.

What AI actually is good for here

I’m not an AI skeptic. I use it every day. The mistake is pointing it at the wrong job. AI is bad at being the trusted voice. It is excellent at removing the friction around the trusted voice.

Think about everything that stops a happy customer from actually referring you. Nobody asks them. The ask comes at the wrong moment. The process is clunky. The referral gets lost and never gets credited or thanked. Those are the real leaks, and they are all logistics problems. That is where smart technology earns its keep: remembering to ask at the right time, making the referral take ten seconds instead of ten minutes, tracking where it went, and closing the loop so the referrer feels appreciated and does it again.

That’s the model I believe in, and it’s how we’ve built the Snoball Engine. Real people create the trust. Technology creates the scale. The technology never pretends to be the recommendation. It just clears the path so more real recommendations actually happen. If you want the psychology behind why customers refer in the first place, we broke it down in human behavior for referral programs.

What this means for home service companies

If you run a moving company, this should be freeing news. You were never going to out-spend a national brand on paid ads, and now AI is about to make paid channels even more crowded and less trusted. That is not your fight to win.

Your fight is the one you were already built to win. You do dozens or hundreds of jobs a month, and most of your customers are quietly happy. Every one of those is a potential recommendation to a friend who is about to move. The homeowners you serve are already researching this way. Our look at how customers research home services shows they trust a real person’s experience over anything a company says about itself.

So here’s where I’d put my energy in the AI era. Stop trying to win the content arms race you can’t win. Systematize the one channel machines can’t counterfeit. Ask every satisfied customer, at the moment they’re happiest, and make it effortless. Track it, credit it, thank them, and do it again. That is a growth channel you own, and it gets more defensible every time AI makes everything else more generic.

The companies that win the next few years won’t be the ones generating the most content. They’ll be the ones whose customers do the talking for them.