She asked which parts a model wrote, and my answer was worse than it should have been

Question

Twenty-two retainers, average $740 a month, mostly listing presentations and property packages with a monthly email. My workflow for eighteen months has been: I draft structure by hand, a model writes first-pass property copy and email variants, I rewrite roughly half of it, and I do all the pricing narrative and the neighborhood positioning myself. Nobody asked. Then one of my better agents asked directly, on a call, and I said "I use tools like everyone does," which is true and also sounded like I'd been caught.

So now I have to decide what the standing policy is, and the room splits on this in a way I didn't expect.

The case for saying it plainly up front: agents are shopping providers hard right now, and half of them are running the same tools themselves at $20 a month. If my pitch is that I charge $740 for judgment and market positioning, hiding the drafting step undercuts the pitch. It also protects me when a listing description has a factual error in it, because responsibility for the claim was discussed before it shipped rather than after. Some brokerages also have their own policies on this, which is the agent's compliance problem and mine by extension.

The case against: the agent isn't buying a process, she's buying a finished package her seller reacts to. Telling her the first pass came out of a model invites her to price the first pass, and the first pass is the cheap part. I've watched two competitors go from $900 packages to $300 packages within a quarter of putting "AI-powered" on their site.

There's also a middle version where I disclose on written client-facing copy and not on internal materials, which I can't decide is honest or fussy.

jessreyes12 · Accepted Answer

I'd have asked the same thing she asked, so let me answer as the person on that side of the call. What I'd actually want to know isn't which sentences came from a model. It's who checks the facts and whose name is on it if a square footage or a school boundary is wrong in a description my seller signs off on.

So put the answer in the engagement terms rather than in a disclosure line. Something like: drafting uses software, every factual claim is verified against the source before it ships, you review and approve before publication, and fair housing language gets a human read every time. That last part matters more than the tooling question and it's the part an agent's broker will care about.

Doing it that way also kills the pricing problem you're worried about. You're not announcing a tool, you're describing what you're responsible for, and responsibility is the thing the $20 subscription doesn't include. The two competitors who dropped to $300 didn't lose the price because they mentioned AI. They lost it because "AI-powered" was the whole offer, so buyers priced the software.

The split disclosure idea is where I'd stop. Different rules for internal versus client-facing material means you'll get the line wrong once, and it'll be on the piece that matters. One policy, said the same way every time.

brandonlee36 · Answer

I buy content as a small portfolio owner rather than an agent, and I've never once asked a provider what drafted the first version. I have asked twice whether anyone verified a number, and both times the answer was slow enough to tell me nobody had. Compass is right that the verification promise is the sellable part. I'd add that you should say it in writing, because a verbal answer on a call is exactly how you ended up with the awkward moment in the first place.

bmoreno23 · Answer

Starting from scratch here, my guess is that everybody's using the tools by now, so silence reads worse than saying it. If asked and you hedge, I hear something being hidden.

She asked which parts a model wrote, and my answer was worse than it should have been

3 replies