If you run a cannabis delivery service in Los Cabos, you already know how much writing the business needs: menu updates, product notes, texts to regulars, weekend social captions, and answers to the same questions customers ask every day. Many owners have started asking AI tools for help, and the results are often bland or off-target. The fix usually isn’t a better tool. It’s a better request. The same discipline that makes good ai art prompts work for image generators, specific subject, clear style, and firm limits, applies to writing for a delivery brand.
Why generic requests produce generic copy
Typing something like “write a product description for our gummies” gives a model almost nothing to work with. It doesn’t know your customer, your tone, your city, or the rules you have to follow. The output tends to be upbeat, vague, and interchangeable with copy from any other shop.
Delivery businesses have a particular problem here. Your customers are often tourists, part-time residents, or locals who order from their phones between errands. They want to know what they’re getting, how to order, and when it will arrive. Vague marketing language doesn’t answer those questions.
What a working prompt contains
After testing a range of requests, most useful prompts for local service businesses share the same five parts:
- Role and audience: Who is writing, and who will read it? For example, “You are writing for adult visitors to Cabo San Lucas who order through a mobile app.”
- Source facts: Paste the real details: product name, package size, delivery window, service area, and pickup or ordering steps. The model should never invent these.
- Constraints: Word limits, banned phrases, and required disclaimers. State them plainly.
- Format: A title, three bullet points, a 120-character limit, or a text message under 300 characters.
- An example: One sample of copy you like, labeled as a style reference rather than something to copy.
A prompt built this way gives you a draft you can edit in a few minutes instead of a rewrite from scratch.
Three examples of tighter requests
Menu update
Instead of “announce our new menu,” try: “Write a 40-word announcement for our Thursday menu refresh. Use only the product names and delivery windows listed below. Friendly, plain language. No superlatives. End with our order link placeholder.” Then paste the list.
Customer text message
For a text to returning customers about a schedule change, specify the exact new hours, the reason in one sentence, and the opt-out wording your phone provider requires. Ask for two versions, one formal and one casual, so you can match the relationship you have with that customer segment.
Social caption
Captions are where generic AI copy does the most damage. Ask for three options, each under 120 characters, with no emojis unless you specify them, and no claims about effects. Then pick the one that sounds most like your shop.
Compliance guardrails for Los Cabos brands
Cannabis rules in Mexico are complex, and they have been changing. Some activities that are discussed publicly are still restricted, and enforcement can differ by municipality and by license type. Treat AI output as a first draft that a human must review against current local rules. Before publishing, check that the copy: To go deeper, explore The marketplace for AI prompts that actually work.
- makes no medical or health claims, including implied ones such as “relief” or “helps you sleep”;
- does not appeal to minors, and avoids cartoon imagery or youth slang in captions;
- does not promise delivery areas, speeds, or product availability you cannot guarantee;
- does not describe sales or delivery arrangements that your license does not cover;
- includes any warning or age notice that your permit or local advisors require.
Build these checks into your prompt as well. A line like “Do not include any health or medical claims, and flag any sentence that could be read as one” catches many problems before they reach your review step. A lawyer familiar with cannabis licensing in Baja California Sur should approve your standard templates.
Building a prompt library for your team
Once you find prompts that work, save them. Most small delivery teams benefit from a shared document with four sections: customer messages, product copy, social media, and internal notes such as shift handoffs. Each entry should include the prompt, the fields that must be filled in, the compliance checklist, and the name of the person who approved it.
Version your prompts the same way you version menus. When a rule changes or a product line is retired, update the template and note the date. This prevents a new staff member from using a six-month-old prompt that still mentions a discontinued item.
Measuring whether the effort is worth it
Don’t judge AI writing by how fast it produces text. Judge it by how much editing each draft needs and whether customers respond. Track three simple things for a month: minutes spent per piece of copy before and after your prompt changes, the number of corrections your reviewer makes, and any customer questions that suggest the copy was unclear. If editing time drops and questions go down, the prompts are doing their job.
Where to start this week
Pick one recurring task, such as your weekly menu announcement. Write a prompt with the five parts above, run it three times, and note what you had to fix each time. Revise the prompt based on those fixes, not on a general sense that the output was “off.” Repeat with one more task next week.
You don’t need a large system to get started. A handful of tested prompts, a short compliance checklist, and a habit of reviewing every draft will cover most of what a Los Cabos delivery brand needs. Keep your customers, your license, and your local rules at the center of every request, and let the tool handle the first draft.
Leave a Reply