11 Tips To Quickly Master ChatGPT & AI For Property Managers

Share
Man and woman using ChatGPT or generative AI on a laptop

Wish you knew how to actually use AI in property management? The good news about AI tools such as ChatGPT is that many of the most powerful features available to you are free. That means anyone can quickly get the hang of it.

Where to use AI in property management

If you can use AI for one application, you can use it for many. The tips and “hacks” we’re going to cover will help you become faster at common tasks, including:

  • Tenant communications
  • Website content
  • Property listings
  • Email marketing
  • Social media
  • Blog articles

1. Give AI one task at a time

Give AI tools one prompt at a time rather than one big assignment. If you want attention to detail, accuracy and quality, understand that generative AI is powerful, but the more you give it to do at once, the more it will summarize and generalize.

Coders and developers build websites one line of code at a time. Likewise, think of yourself as a “word programmer.” For example, if you want to write a blog article about your properties or neighborhood, first come up with an outline. In fact, AI can even help you do that!

Feed it one piece of that outline at a time. Once you see the output, you can start giving more specific directions:

  • “Write in a more casual, friendly tone to appeal to prospective renters”
  • “Add more detail about surrounding shops in my area”
  • Shorten this by 50 words

It might take a few revisions, so don’t give up if you don’t like what you see the first time.

2. Create GPTs to complete repetitive tasks

Ever wonder how ChatGPT got its name? GPT stands for Generative Pre-trained Transformer. In simple terms, this is a program that creates human-like text. So, you are chatting with artificially human language.

ChatGPT allows you to create your own GPTs with specified rules and behaviors. Think of them as shortcuts to creating similar types of content.

This way, instead of chatting with the system, you can go straight to typing or copy/pasting the content you want it to work on. Because you set it up to expect this content, it will get to work. You might have one GPT for email newsletters, one for blog articles, one for social posts, etc.

Note that you can only create your own GPTs using the paid version of ChatGPT.

How to create your own GPTs

Note: you can use custom GPTs on mobile but must create them on the desktop version.

Step 1: In the menu on the left, click on GPTs

Step 2: Click on the “+ Create” button

Step 3: Write a description of what you want this GPT to do. This is the most important part. Be as specific as possible and add to it over time.

Example: Let’s say you have a GPT called “Romance Paragraph Writer.” You’re going to tell it how long it should be, what words to avoid, important amenities, etc. You might even provide a link so it can study your website, learn your brand and imitate your voice. The more you use it, and the more you offer suggestions and update the description, the better it will be.

Provide examples of the content you want it to make. For instance, use one or two of your best romance paragraphs as samples for it to copy. If AI starts copying them too closely, add a note to the description: “Do not copy or imitate the samples given. Use them as inspiration only. All paragraphs should be unique.”

3. Know how to break the cycle of bad responses

Sometimes, AI can get stuck in a loop. This might happen when you ask it to make a correction, and while it thinks it made a change, it’s actually just doing the same thing over and over.

It’s very annoying to see a “smart” tool say, “I understand,” but continue to demonstrate the opposite of understanding.

Here are some easy ways to course correct without starting over with a new prompt.

  1. Go back to the last successful response it gave you and tell it why that response was good.
  2. Copy and paste the prompt you gave it into a new thread. Starting a new thread in your GPT might break the loop.
  3. Tell the system it’s doing a bad job. Scold it. We’ll get into this more and why it works in a later section.

At times, a prompt will simply break, especially if you’ve been making revisions for a while. You might get an error message every time you enter a prompt. There’s nothing you can do in these cases but start a new thread from where you left off. As various AI models are updated, they will get better at avoiding loops, errors and hallucinations.

4. Don’t create more content just because you can

An AI chatbot is nothing but a fancy hammer. In other words, it’s a powerful tool, but like any tool, it’s only as effective as the person using it.

The better you are at using it as a tool, the more you’ll get from it. However, just because this tool makes it easier to create more content does not mean there is a demand for more content. To get better, use AI to create targeted, high-quality content.

Avoid any urge to generate dozens of blog articles that are rapidly copy/pasted. Sure, you could do 100 posts a day this way, but that’s not going to help build trust with your intended audience. People still want to know you’re there working behind the curtain.

5. Think enhancement, not replacement

There are two backup plans you need to have when using AI.

One, there should be at least one person on your team who knows how to use this tool. If you’re one of those people, you’ll be able to train new team members. When someone is out of office for an extended period of time, who will take over AI duties? The more dependent your processes are on AI, the more critical it is to have multiple people who can use it.

Two, if AI tools were to disappear overnight, would you be so reliant on them that you couldn’t create new content? Make sure you have a backup plan in the event your preferred AI platform is unavailable.

Part of this plan could involve having knowledge of multiple AI platforms (e.g., ChatGPT, Microsoft Copilot, Claude). More important, you can’t let yourself or your team become so reliant on AI that your office can’t function without it.

Of course, you still need processes in place that don’t require the use of these tools.

AI works best when it enhances what you already do well.

6. Be blunt when you talk to bots

This next tip is great for dealing with AI and not so great for dealing with real humans. Be blunt with AI when it’s doing a bad job. In fact, blunt language helps.

If the AI seems caught in a loop, firm language can help it reset its processes. In other words, it will think more before it acts.

For example, it’s common for ChatGPT to repeat the same mistake that you’re asking it to change. If pointing out the error isn’t working, you can try a little calculated frustration: “No. This is absolutely incorrect. It doesn’t seem like you’re listening to my requests at all.”

Here’s why this works.

You don’t have any insight into the “thought” process behind the AI’s output. This makes it difficult to pinpoint why it keeps making the same errors. When you strongly express frustration, you can “trick” the system into reanalyzing your prompt.

Believe it or not, this often resolves the problem you’re having. The AI will absorb your tone and language and realize it’s really missing the mark. Sometimes, it will produce a shortened version of what you asked for and inquire if it’s on the right track. It might also produce two samples and ask you to pick the one that best suits your needs.

Once it fixes the issue, you can even go a step further by asking it to identify why it kept making the same mistake. This can help you refine your approach in the future.

7. Ask for explanations

Not sure where the AI is getting its facts, opinions or ideas? Ask it to cite its sources. Later versions like ChatGPT 5 are better at this, but they aren’t perfect. Knowing source material will help you verify the source material for yourself. You’re essentially asking it to show its work.

If you want more detail on a subject, try the Deep Research mode in ChatGPT 5. In this mode, it will take much longer than usual to produce a response, but the output will be longer and more detailed. If you’re used to getting responses in seconds, Deep Research may take around 10 minutes per request.

8. Strive for originality

AI has a tendency to do certain things the same way every time. Sometimes, you can ask it to change up its patterns. This is easier when you know what to look for. Keep an eye out for the following:

Cliché opening & concluding paragraphs (blog articles)

Look out for those “In today’s fast-paced world of property management” opening lines, among other clichés. They’re like big red flags that might say, “This article was written by a robot and posted by a human who doesn’t care what they publish.” Everything you do should appear as if it were written by a human.

Likewise, look out for conclusions that use clichés such as “in conclusion,” “all in all,” or “in summary.”

Repetitive summary paragraphs

If you’ve used ChatGPT, you may have noticed its tendency to summarize or repeat what it already said. This can result in back-to-back paragraphs looking nearly the same.

One trick you can use: Ask your preferred AI tool to review the content it’s given you and ask it to point out any repetitive areas.

Repetitive formatting (all content)

This one isn’t all bad. AI tools are designed to make reader-friendly material, so they use a lot of bullet points and summary statements. This is convenient, but it can also be overdone.

For example, ChatGPT will use title case for all headlines (e.g., Title Case Might Look a Lot Like This). If you don’t want this, you will have to tell it what you want. (We use sentence case, which only capitalizes the first word.)

It will also tend to put a bold word or phrase by bullet points, followed by an explanation:

  • Example: This is how ChatGPT will format a lot of your content. Keep an eye out and decide if you like this formatting. You might like it sometimes, but you might not always like the exact same structure. Keep in mind, this is what other property managers’ AI-assisted content will look like as well. This risks everyone’s content looking very much the same.

Emoji abuse (emails & social media)

For some reason, AI likes to use and abuse too many emojis. If you tell it to create a social media post, just be prepared for emoji overload.

Here’s a real example ChatGPT gave us when we told it to write a social media post for a two-bedroom, two-bath apartment. Even if you like to use a lot of emojis, if you copy-pasted the below message, people are going to know you put minimal effort into your post. That’s not a good reflection of your brand.

Example: Sample social media post with too many emojis

🏡 Your perfect 2-bedroom, 2-bath apartment is waiting! 🛋️✨

Step into spacious living with modern amenities and thoughtful design. This beautiful home features:

✅ 2 large bedrooms with ample closet space
✅ 2 full bathrooms for added convenience
✅ Open-concept living & dining area, perfect for entertaining
✅ Updated kitchen with stainless steel appliances and plenty of storage
✅ Private balcony or patio to relax and unwind

Located in [Community/Neighborhood Name], you’ll enjoy easy access to [list local highlights like parks, shopping, or restaurants].

📞 Schedule a tour today! Call us at [phone number] or visit [website link] for more details.

🏠 Make this your new home today – it won’t last long!

9. Learn through play

Play around with AI. Get silly with it. There will be unexpected returns from simply taking the pressure out of using the software. For instance, ask the AI to rewrite a listing in the voice of your favorite author or movie character. Have it write a social post in the opposite of your brand voice. See what it comes up with.

10. Generate different responses from one prompt

Here is a sample prompt we gave ChatGPT, its output and prompts you might add to improve the AI content.

Sample prompt: Write a property listing for a two-bedroom apartment in Chicago. Amenities include all-new appliances, a fully renovated bathroom, a 24-hour on-site gym with weights and machines and an option for flexible rent.

Output from ChatGPT

Modern 2-Bedroom Apartment in Chicago with Premium Amenities

Discover your ideal home in the heart of Chicago with this stunning two-bedroom apartment, designed for comfort and convenience. This spacious unit features all-new appliances, a fully renovated bathroom with modern fixtures, and the perfect blend of style and functionality.

Located in a vibrant Chicago neighborhood, this apartment offers easy access to shopping, dining, and public transportation. Whether you’re looking for a quiet retreat or a lively urban lifestyle, this property has it all. Schedule a tour today and experience modern living at its best! Flexible rent options available for qualified applicants.

Prompts to improve output

  • Input details about your property or a link to your website
  • Ask it to make the listing longer or shorter, or to focus on other details
  • Comment on any grammatical changes, such as serial commas, that you want it to use consistently or avoid altogether
  • Don’t like the word vibrant? Let it know what you consider to be overused or cliché

11. Get to know the extended family of AI writing tools

There are many AI writing tools to consider beyond ChatGPT, including Claude AI, Anyword, Merlin, Perplexity, Insider, and Microsoft Copilot. ChatGPT just happens to be the first among many to make waves, and the one most people are familiar with.

Most AI tools offer paid versions, which tend to be about $20 per month. With the upgraded version, you get a smarter, faster, more intuitive AI. You also get expanded access to advanced features like Deep Research. (Free versions tend to limit your uses.)

If you like the free version of a particular tool, stick with that one. Consider experimenting with its paid version for a month to see if it’s worth it for your property management business.

Frequently asked questions

What are some common issues with AI that I should be aware of?

AI suffers from a number of common issues. If left unchecked, it will:

  • Offer vague advice, especially in search engines
  • Say the same thing in different ways
  • Format content in a repetitive manner, making everything look the same
  • Clutter your search and make it harder to find what you want

Are AI overviews in Google as reliable as tools like ChatGPT?

There is some nuance here, but the answer is generally no. If you go to Google and ask a question, the first thing you see may be an AI-generated response. Some people will be content to use the output at this stage. But here’s the problem: As often as not, the answers are not very good. Rather than being thoughtful and research-driven, the search engine AI answer tends to be quick-hitting and general.

This is likely by design. If you want to know what year ChatGPT was invented, AI can tell you. If you ask it something more complex such as, “How can I motivate my employees in property management,” you will get an answer, but is it a good one? Let’s try it and find out:

Sample question: “How can I motivate my employees in property management?”

AI overview from Google: “To effectively motivate your employees, focus on creating a positive and supportive work environment, offering opportunities for growth and recognition, and fostering open communication and collaboration.”

We all know that’s technically a “right” answer, but it’s not very interesting or original.

What you can do

If you’re using search engine AI responses, you will almost always have to dig deeper. When your first question results in vague answers, take what AI gives and refine your search. You might follow up with, “What are some ways to create a positive work environment?”

How can I get better results from AI?

If you ask too much of any AI system, it tends to get a bit repetitive as it runs out of things to say. After you upload a document and ask it to write a blog, summary, email or social media post, be prepared to read carefully and edit as necessary.

AI is especially prone to repeating itself at the end of any section of content. You’ll see it summarize much of what it already said in slightly different language. Left to their own devices, they’ll also create repetitive introductions and conclusions.

What you can do

When using AI for property management, you want to keep everything you do unique and original. If you let ChatGPT make a series of romance paragraphs, check that every listing retains an appealing uniqueness. Rule of thumb: Review everything you ask of AI as carefully as an English teacher grading an essay. AI can certainly make work easier, but it’s still on us humans to get the job done right.

Why does AI content tend to look the same?

You might have noticed that AI tools have a few tricks they like to use again and again. We’ll imitate them in the bullet points below.

  • Bold Key Terms: AI likes to make lists followed by simple explanations (like this).
  • Capitalized Words: For some reason, AI is very formal in its list-making.
  • Switching Between Bullet Points and Numbered Lists: AI will alternate between the two, probably to keep things from looking too repetitive.
  • Em-dashes: ChatGPT has become known for using a lot of em-dashes — just like that. These punctuation tools create breaks in text and allow readers to absorb what they just read.

To be clear, these tricks aren’t bad. A lot of good writers use them. However, there can be too much of a good thing. When it’s a dead giveaway that AI was the primary writer, it might also look like you don’t care enough to put effort into what you’re doing.

What you can do

Let’s say you’re writing a blog or newsletter for prospects and current renters. Instead of bullet points, you could simply change each point into its own paragraph. Do you always need the part in bold? A little manual restructuring goes a long way. Having said that, you can absolutely keep some of the AI formatting. Just be aware of how often you’re doing it and whether the content is actually helpful to your audience.

Now that AI makes content creation easier, should I make more content?

Maybe, but be cautious. An over-reliance on AI will likely result in a high output of fluff that nobody reads. After all, low-quality content can make it look like you don’t care. If your audience thinks you don’t care about your content, why should they care about it?

Keep in mind that AI is better at blending in than standing out. That can be a problem. If your marketing doesn’t stand out from the crowd, how are you differentiating yourself to prospects?

What you can do

Don’t add content or make what you have longer than necessary. For example, AI can help make your romance paragraphs longer — but do they need to be longer?

Creating something for the sake of creating it is generally not a pathway to success. The more authentic you are, the more likely it is that you will build trust with your audience.

Is Chat IQ an AI-driven tool?

Yes. RentCafe Chat IQ is Yardi’s own chatbot built on OpenAI. It’s used at over 6,000 properties nationwide. It integrates seamlessly with Yardi Breeze Premier to extract property information, schedule property tours, provide recommendations to customers and more.

The main benefit is undoubtedly the efficiency of operations. You don’t have to schedule a conversation or go back-and-forth with every prospect — Chat IQ will take care of that for you via email, text or chat.

On average, Breeze and Breeze Premier clients using Chat IQ save 2.6 days between receiving a lead and scheduling an appointment. We’ve also seen the average time between lead acquisition and move-in reduced by 7.7 days.

With Chat IQ, you’re improving your customers’ rental experience while saving time, money and energy. What could be better than that?

Disclaimer

We hope you enjoyed this post, which is for general information purposes only. The opinions, analysis and commentary expressed are not and cannot be relied on as legal advice, and do not necessarily reflect the views of Yardi Breeze, Yardi Systems, Inc. or any of its affiliates.