IA que trabaja

Spanish AI prompt examples: how to write them and real-world uses

Skip the theory and use these Spanish prompt examples in your business today: copy them, replace the text in brackets and see the typical mistake and how I write each one.

By David García Published: 6 min read

If you already know what a prompt is and want to get straight to it, this guide is for you. If not, first take a look at what a prompt is and how to ask good questions, where I explain the structure. Here, we’ll focus on practical examples in Spanish, grouped by task, that I use or have seen work in real businesses.

They all follow the same idea: tell the AI who it is, what information it has, what it needs to do and how you want the result presented. Replace anything in brackets with your own details.

Why do my Spanish prompts produce worse results than English examples?

Most of the time, it’s not the language but a lack of context. Current models write very well in Spanish. The issue is that we tend to be more indirect in Spanish (“Could you do something about...?”), and the AI fills in the gaps with generic text. Two things that work for me:

  • Ask for “Spanish from Spain” (or your country) to avoid expressions you don’t use.
  • Give it an example of your own writing: an email or a post you’ve written. It’s much easier for it to copy your tone than to guess it.

Prompt examples for emails and communication

Responding to an upset customer

Not this: “Reply to this customer.”

Try this instead: “You’re in charge of customer service at [your business]. A customer has sent us this email: """[paste the email]""". Write a reply in Spanish from Spain, using the informal ‘tú’ form, that acknowledges the problem without making excuses and explains what we’ll do ([what you’ll do]) and when. Maximum 120 words. Avoid phrases like ‘we apologize for any inconvenience caused’.”

Requesting a quote from a supplier

“Write a short email to [supplier] requesting a quote for [product or service], quantity [X], for delivery on [date] at [location]. Ask about lead times, payment terms and whether the price includes shipping. Use a professional but friendly tone. End with my signature: [signature].”

Summarizing a long email thread

“I’ll give you an email thread. Summarize it in 5 bullet points: what has been agreed, what’s still pending, who needs to do what and by when. If anything in the thread is unclear, say so rather than making assumptions.”

Prompt examples for customer service

Answers based only on your information

“You’re the assistant for [business]. Answer customers’ questions using ONLY this information: """[opening hours, prices, terms, frequently asked questions]""". If the answer isn’t there, reply: ‘A colleague will confirm that for you shortly’ and don’t make anything up.”

This is the most important prompt in the whole guide. It’s the basis of any AI-powered WhatsApp chatbot: limiting what it can say so it doesn’t promise something that doesn’t exist.

Classifying incoming messages

“Classify each message into one of these categories: quote request, issue, appointment, invoice, other. Examples: ‘How much does it cost to replace the screen?’ → quote request. ‘My order hasn’t arrived’ → issue. Now classify: [messages]. Return a table with the message and category.”

Prompt examples for sales and lead generation

Preparing for a sales call

“I’m going to call [business name], which operates in [industry] in [city]. I sell [your service]. Prepare: 3 questions to understand their situation, the 2 most likely problems a business like theirs has that relate to what I sell, and a one-sentence explanation of the benefit. No salesy phrases.”

Following up after sending a quote

“Write a follow-up message for a customer I sent a quote for [service] [number of days] ago. Keep it short and low-pressure, offer to answer any questions and include a question that’s easy to answer.”

Analyzing why you lost a sale

“Here are my notes on 5 quotes that didn’t lead to a sale: [notes]. Look for patterns: price, timelines, lack of follow-up, type of customer. Give me 3 conclusions and one concrete action for each.”

Prompt examples for social media and content

TaskPrompt
Post ideas“Give me 10 post ideas for [social network] for a business in [industry] in [city]. Include a one-line hook for each idea. Make them useful to customers, not advertisements.”
Adapting a post for another platform“Turn this LinkedIn post into a 30-second script for a vertical video: [text]. Put a hook in the first sentence, use short sentences and end with a call to action.”
Writing in your tone“Here are 3 of my posts: [texts]. Pay attention to how I write (length, expressions, tone) and write a new post about [topic] in the same style.”
Responding to comments“Suggest short, friendly replies to these comments on my latest post: [comments]. If any are complaints, suggest moving the conversation to private messages.”

If you want this to happen automatically every week, without opening the chat, that’s automating processes with AI.

Prompt examples for Excel and data

Formulas

Not this: “How do I add numbers in Excel?”

Try this instead: “In Excel, I have amounts in column B and status in column C (‘paid’ or ‘pending’). Give me a formula to add up only the pending amounts and explain it in one line.”

Cleaning data

“Here’s a list of customers with names, phone numbers and email addresses mixed together and in different formats: [list]. Return it as a table with columns for name, phone number (format +34 600 000 000) and email. Flag any rows with missing information.”

Drawing conclusions

“Here’s the sales data by month and product: [data]. Tell me which 3 products are growing, which are declining and whether any month looks unusual. Explain it for someone who isn’t a numbers person.”

Prompt examples for documents

Reviewing a contract

“Read this contract: [text or PDF]. Make me a list covering: duration, how to cancel, penalties, payments and deadlines, and any clause that could work against me. This isn’t legal advice: I just want to know what to look at with my adviser.”

Turning notes into a document

“Using these meeting notes: [notes], write meeting minutes with attendees, topics discussed, decisions and tasks, including who’s responsible and the deadline for each task. Use a list format, with no filler.”

What if the answer isn’t useful?

Don’t start from scratch. Tell it what’s wrong, just as you would a person: “make it shorter”, “remove the adjectives”, “this isn’t true, use only the information I gave you”, “put it in a table”. Two or three rounds like this are usually enough.

And if you repeat a task every week, save the prompt that worked for you. That’s the difference between using AI now and then and having it work for you. If you want to bring this to your team, our AI training for businesses focuses on prompts for your real tasks, not generic examples.

Keep reading

Guides

AI advisors: how intelligent assistants work for businesses

An AI advisor uses data and context to recommend actions. Learn how to choose one and put it to work in your business.

Read · 10 min ·
Guides

Claude Opus, Sonnet or Haiku: practical differences and when to use each

Compare Claude Haiku, Sonnet, Opus and Fable by speed, quality and cost, with a task-by-task guide and common mistakes to avoid.

Read · 5 min ·
Guides

Multimodal AI: What It Is, Examples and How to Use It in Your Business

Everyday AI assistants can read photos, listen to audio and understand PDFs. Learn how businesses use them and what to consider before adding them to your processes.

Read · 6 min ·