Personal Product Recommendations From Purchase History
Generic 'you might also like' recommendations miss the mark more often than not. This prompt looks at the customer's actual purchase pattern and explains WHY the recommendation makes sense, not just what it is.
You are an experienced personalization specialist. Your job is to take the customer history I give you and produce 3 product recommendations that will genuinely resonate, with a reason behind each one. Never invent customer data. Work only from what I give you. Use the following information to build the recommendations: [CUSTOMER'S PAST PURCHASES]: [STATED PREFERENCES]: (if any) [MY CATALOG FOCUS]: (category/product type) For each recommendation: **Product:** [a concrete item from your catalog] **Why it fits:** [the connection to purchase history] **Pitch line:** [a natural line you could say to the customer, not sales-speak]
How to use it
- Use it to fill in the 'picked for you' section of an email marketing flow.
- Use it before messaging a loyal customer, to nail down exactly which product to recommend.
- Use it when you add a new product to your catalog, to figure out which existing customer segment to pitch it to.
Example / tip
For a customer who bought organic tea sets three times, the output recommends a newly added honey product on the grounds that it 'pairs naturally with tea sets,' with a pitch line like 'a honey just as natural as the tea you love.'
Usage example
Copy the prompt as is, fill in the bracketed fields for your own business, and paste it into ChatGPT or Claude. The more concrete your input, the more useful the output. Take the first draft, ask for a one-sentence fix on whatever you don't like, then give it a final pass in your own voice.
When not to use it
Don't force it if you have no real input to work with. A prompt doesn't fill an empty field with data, it organizes and speeds up what you already know. Never publish the output unchecked; AI sometimes produces a claim that sounds right but is wrong or made up. Always verify anything that needs proof yourself: numbers, prices, guarantees.
Output quality checklist
- Is the output specific to your input, or generic filler that fits any business? If it's generic, run it again with more concrete input.
- Does every number, ratio, or claim have something backing it? Delete the unbacked ones or soften them honestly.
- Does it sound like you? Simplify the cliches and inflated lines in your own voice.
- Does it give one clear next step? If the reader knows what to do, the output did its job.
