ChatGPT Improves Lead-to-Lease Conversion Rates
Improve Lead-to-Lease Conversion Rates Using ChatGPT
Having a high Lead-to-Lease conversion rate is crucial for increasing the top line. With many potential tenants exploring various options, personalized communication is key. However, some property managers may notice their lead-to-lease rates decline when relying on lazy AI responses. This doesn't mean the tool is ineffective—it means we need to learn how to use it properly. Let's break down how to become better "prompt engineers."
1. Use ChatGPT to Refine, Not Replace Your Input
ChatGPT should complement your understanding of the tenant, not replace it. Leasing agents still need to pay attention to prospective tenants so they can create personalized responses—even with AI. Provide key details about the prospective tenant—what they liked during the tour, their needs, and any concerns they have. This way, you guide the message, and ChatGPT refines it.
Here's an example of a poor prompt:
"Write a follow-up email to Joe Smith after his tour of Apartment Standard."
This is a poor prompt because it provides no specific details, allowing ChatGPT to generate a generic, often robotic response.
2. Focus on Personalization
AI should enhance personalization, not automate it. Including personal details from the tenant's tour makes your communication more engaging and relevant. Key information might include:
- Apartment features they liked
- Family or personal details that influence their decision
- Specific concerns, such as price or lease terms
By adding this information to your prompt, ChatGPT can generate an email tailored to the lead. For example:
Improved Prompt:
"[Prompt from part 1] Joe is looking for a 2-bed, 1-bath apartment for him, his wife, and newborn daughter. He liked the complimentary parking, the new appliances, and the spacious living room. He is concerned about the rent being $50 above his budget."
3. Refine the Tone to Match Your Style
While ChatGPT can create well-written responses, it may not always match the tone you want. You can guide the tool by specifying the tone that fits your style, whether it's formal, friendly, or neutral.
Here's an example of how to specify tone:
"[Prompt from part 2] Keep the tone professional but approachable, without being too casual or sales-driven."
Remember that adjusting the tone may require some iteration. Try adding adjectives describing the tone until the message aligns with your voice.
4. Leverage Past Emails to Fine-Tune Responses
Fine-tuning involves giving ChatGPT examples of successful past emails. This helps improve the quality of the output, making it more in line with your style.
Here's how to use past emails for fine-tuning:
"[Prompt from part 3] Here are examples of emails I've used in similar situations: [Email Example 1], [Email Example 2].”
This allows the tool to refine the message to better reflect your style.
5. Create a Template for Repeated Use
Once you've refined the process and found a style that works, you can create a standard prompt template to save time. You'll only need to adjust specific details like the tenant's name, preferences, and concerns. The tone and past examples can stay the same, ensuring consistency while still personalizing each email.
Conclusion
Lazy prompting leads to poor responses, but with a little thought and refinement, your ChatGPT responses can be much more useful. Test it yourself by comparing the difference in responses between these prompts. We started with:
"Write a follow-up email to Joe Smith after his tour of Apartment Standard."
But with a bit of prompt engineering we finally have:
"Write a follow-up email to Joe Smith after his tour of Apartment Standard. Joe is looking for a 2-bed, 1-bath apartment for him, his wife, and newborn daughter. He liked the complimentary parking, the new appliances, and the spacious living room. He is concerned about the rent being $50 above his budget. Keep the tone professional but approachable, without being too casual or sales-driven. Here are examples of emails I've used in similar situations: [Email Example 1], [Email Example 2].”
You'll notice a significant improvement in response quality.
That's all for this week's AI update! I'm curious to hear your thoughts. Email me at suket.shah@propkeep.io.