Five copy-paste prompts that research a venue and draft your walk-in pitch, cold email and follow-ups in about five minutes. They run on the free tier of ChatGPT or Claude, so there is no second subscription to buy.
“I have the list. I just have not actually contacted anyone yet.”
That is the most common sentence in this business and almost nobody says it out loud, because it sounds like laziness and it is not. You did the research. You have twelve venues in a spreadsheet with names and addresses. And every one of those twelve requires you to sit down and write something — an email, an opener, a follow-up — from a blank page, for a building you know almost nothing about.
So it slides. Not for a day. For weeks. The list ages, the enthusiasm that built it drains out, and the honest reason nothing has happened is that writing twelve different first emails is twelve small unpleasant tasks and you have a job.
The workaround everyone reaches for is one template, sent twelve times. It is a reasonable instinct and it performs badly for a reason that is easy to miss: a property manager reads three or four unsolicited emails a day and can identify a template in about two seconds. Not because the writing is bad. Because it does not contain a single fact about her building.
The pitch that lands at a 24-hour gym is not the pitch that lands at a law office, and it is not because the tone is different. It is because the argument is different. A gym has members who arrive hungry and leave thirsty at predictable hours and a front desk that already fields the question “do you have anything to drink?”. A law office has staff who will not leave the floor at 3pm and a facilities manager who cares whether you are insured. Those are different problems and a template pretends they are the same one.
Writing twelve genuinely specific pitches by hand is the correct answer. It is also, realistically, four hours you are not going to spend, which is why the list is still sitting there.
Five prompts. You paste one in, fill four or five bracketed slots with what you know about the venue, and the model does the researching and drafting that was blocking you. Venue research, the 45-second walk-in pitch and its 15-second version, a cold email sized for someone who skims, a three-touch follow-up sequence, and an objection rehearsal that role-plays the venue back at you and grades your answers.
It runs on the free tier of ChatGPT or Claude. A paid model writes slightly nicer prose; it does not give you a different answer, and we are not going to pretend otherwise to make the pack look better.
The prompts themselves are about a page. The other eleven pages are the difference between a prompt that works and a prompt that produces confident nonsense. Every one of the five has a section on exactly what to put in each bracketed slot, a section on what a good output actually looks like so you can tell, a list of the seven or eight specific ways the model gets this wrong, and the exact correcting sentence for each one.
And every prompt carries a guardrail. Left alone, a language model will happily write a revenue figure, a commission percentage or a guarantee into your pitch, because that is what persuasive sales copy looks like in its training data. In the United States, an income claim about a business opportunity is a regulated statement. The guardrails are written to stop it, and the pack tells you plainly that you still owe every output ten seconds of your own judgement before it leaves your outbox.
One instruction document plus a plain-text file. Page count measured from the generated PDF.
Takes a venue name, address and type, and returns who is likely on site, when foot traffic peaks, and who most plausibly signs off on a placement. The instructions cover what to fill in, what a good answer looks like, and the failure modes — chiefly that the model will state assumptions as local facts unless you make it flag them, which this prompt explicitly does. Treat the output as a well-prepared guess to confirm on site, not local knowledge.
Writes both lengths from the research output, because the 45-second version is useless when the person behind the desk is mid-task and the 15-second version is what you actually use more often. Annotated with what to fill in, what good looks like, the common failures, and a guardrail that stops it inventing numbers about what the location will earn.
Subject line and a body sized for a property manager who skims, built around a single ask rather than an explanation of vending. The failure-mode section is the useful part here: models default to long, warm, and multi-ask, which is exactly the email that gets archived, and the correcting sentences are given verbatim.
Writes the three-touch sequence, each message shorter than the last, each with a different reason to exist rather than three variations of “just checking in”. This is the prompt that matters most for anyone whose actual problem is that they send one email and never send a second.
Role-plays the venue back at you, throws the objections that venue type actually raises, and grades your answers. It is the closest thing in the pack to practice, and it is there because reading a good answer and being able to say a good answer out loud under mild pressure are different skills.
The document closes with how the five prompts feed each other in sequence, an explicit section on what the AI cannot do for you, and the compliance note on the ten seconds you owe every output. The plain-text file is just the five prompts with no commentary, because once you know how they work you paste from it rather than read it.
Verbatim from the opening of the pack, including the guardrail every prompt carries. No summary, no paraphrase — this is the actual text.
This is five prompts. That is the whole product. You paste one into a free ChatGPT or Claude account, fill the bracketed slots with what you know about a real venue, and it writes the thing you were going to spend forty minutes rewriting badly. Everything else here exists to make those five prompts work harder.
What it is not: not software, not a course, not a lead list. Nothing installs, there is no login, and there is no second subscription — every prompt is written to run on the free tier of ChatGPT or Claude. What you need is a free account, one real venue, and four to six things you noticed about it with your own eyes. A full run takes about 20 minutes the first time and under 10 once you know the slots.
One rule sits above all five and is not negotiable. Never let a model state a revenue figure, commission percentage or guarantee on your behalf. In the US, income claims about a business opportunity are regulated. Every prompt below carries a line telling the model not to do it. Leave that line in.
Most operators typed one question into ChatGPT, got back something bland, and decided the tool was useless. The tool was not the problem. A one-line question gets a one-line-quality answer. Five things fix that, and all five are already built into the prompts below.
| The other way | What it costs | What you are left with |
|---|---|---|
| Writing twelve pitches by hand | $0 and about 4 hours | The right answer, and the reason your list has been sitting untouched for three weeks. |
| One template, sent twelve times | $0 and 20 minutes | Twelve emails that contain no fact about the recipient’s building, read as bulk in two seconds, and are archived. |
| A copywriter or VA | $50–$150 per venue | Better prose than the model, written by someone who has also never been inside that building. |
| This pack | $27, once | Five prompts you reuse for every venue forever, on a free AI tier, with the failure modes and the income-claim guardrails written in. |
The comparison that matters is the first row, not the third. This is not competing with a copywriter; it is competing with the version of you who was going to write them all properly on Sunday. It costs $27 once and then costs nothing per venue, which is the only reason it survives contact with venue number nine.
This is five prompts, not a course and not a piece of software. If you are comfortable writing your own prompts you do not need it, and we would rather you kept the $27.
Email [email protected] inside 30 days and we refund the whole thing. We will not ask you why and we will not make you prove you tried. We would rather have your goodwill than your $27.
Cheapest first. Everything here is a one-time purchase — nothing on this page renews.
The kit is yours forever once you buy it; the VendBuddy app is the separate monthly thing that does the list-building continuously, if you ever want that part automated.
Payment, licensing and refunds for this product are handled by Gumroad; the checkout opens right here on this page. VendBuddy sells documents and software. We make no representation about the income, profit, or number of placements any operator will achieve — results depend on your market, your effort, and factors outside anyone’s control. Nothing here is legal advice; have a local attorney review any agreement before you rely on it.