Hello, All!
If there's a thumb up / down available, please make use of them before providing feedback here in the forum.
Those little thumbs up/down buttons after Rezzy's responses aren't just for show - they're the most effective way to help us calibrate Rezzy to better match what you're looking for. For example, is Rezzy too wordy? Thumbs down. Perfect length? Thumbs up. Your clicks make a real difference!
Hey everyone! 👋
Thanks so much for all the awesome feedback you've been sharing in this thread—we really appreciate it! To help us stay organized and make sure we address everything, we'd love to ask for a small shift in how we handle feedback going forward.
Could you please start posting new feedback as individual threads in the Rezzy AI (Beta) forum? This helps us because:
No worries about what's already been posted here—we've got it all and will review it! Just going forward, individual threads would be really helpful.
Thanks for being such engaged members of our community!
I haven't played with Rezzy yet, but generally speaking, I have a really negative view of AI because when I encounter it in my own activity, I recognize it immediately, and it's frequently wrong and unhelpful, and only prevents me from getting the real, actual help I need because companies are (stupidly) relying on it to provide customer service, and all it does do in my experience is provide customer frustration. This is especially true for me because I almost never contact a company to ask simple, obvious questions (occasionally, I do make human mistakes, but they're rare). If I'm reaching out, it's almost always because something is really wrong, and the people responsible for the system I'm having a difficulty with haven't anticipated what I'm contacting them about. So there is no selection offered in the phone cue, or no AI prepared response that matches what I'm reaching out for. I need a real, thinking, responsible and responsive human, or I'm simply not going to get my problem solved.
So why am I posting here if I haven't actually tried the service yet? Because I was reading the information about Rezzy, and what I was really seeing was the introduction of two separate additions to OwnerRez, one I could really use, and one I have little interest in. They are:
1) Task Management
2) AI Message Responding (and task creation)
I encourage OwnerRez to separate these two features of the product. Task Management would be a great addition! AI handling of customer inquiries? For me (and many others I suspect) not so much!
I get that the genesis of these two enhancements may have happened hand in hand, but I suspect that separating them as features, and having Rezzy interact with task management as needed, would be a great approach. I'll edit this post once I've had a chance to look at it more, and see if I still feel the same way. Thanks! :)
We were just invited to the Beta, so looks like there is some functionality that wasn't available at the start that is now in the build.
GLOBAL FAQ ISSUES:
With that said, I am starting slow by building out a few FAQs regarding check-in/check-out times, service animals, and pet policy. These are all GLOBAL policies as defined in the FAQs; however, when Rezzy answers questions, it keeps asking for the property, stating that the policies are different in every case. I am using ChatGPT to analyze these problems, and we wrote a directive stating that if there is a policy on a global FAQ to go with that and not ask for the property, but that hasn't worked either.
Are others having any luck with the global policies? It may be the subtle difference between a service animal and a support animal that the system doesn't understand? But again it shouldn't be asking for a property.
Rezzy continually repeats ending phrases in all interactions for a certain guest, e.g., one guest said that he was going to celebrate his sister' birthday, and Rezzy ends all interactions with this guest with a reference to the birthday, like "have an amazing time celebrating your sister's birthday." , "Hope you and your group have an amazing time celebrating your sister's birthday", etc.
Some parts of a conversation are gone after replying...see this example, the guest asked where the trash bin is, I replied to the guest thru Rezzy, the reply was sent correctly to the guest (it shows on Airbnb), but is no longer to be seen in Rezzy (only the picture I sent after the text). This happened with both Airbnb and Vrbo conversations.
So far so good
I second these suggestions emphatically.
I want to be able to forward tasks to people via email and, as someone mentioned, have a way to track that and see if it is underway. When done, it should be archived. Otherwise, it is just a reminder and gets deleted.
We need a TRULY universal INBOX for these tasks to appear inside of.
Guest wants a early check in
But she is already checking in on the 25th
Rezzy should have caught that.
Hello, All!
If there's a thumb up / down available, please make use of them before providing feedback here in the forum.
Those little thumbs up/down buttons after Rezzy's responses aren't just for show - they're the most effective way to help us calibrate Rezzy to better match what you're looking for. For example, is Rezzy too wordy? Thumbs down. Perfect length? Thumbs up. Your clicks make a real difference!
I agree. It isn't a request if a "?" does not appear in the text.
Like many AI replies, it is too verbose.
Here are some preliminary notes and thoughts from the first week:
1. Send, Edit, Delete icons (and now thumbs up and down) are tiny and close together. It's too easy to accidentally hit send instead of edit.... Could there be a confirmation before sending?
I wanted to second all of these suggestions and also provide a screenshot from my iPhone. This illustrates how small these icons are.
An extra line feed between them might help in the short term and be easy.
Love this feature so far. Couple instant things I would like to see happen.
#1 VERY IMPORTANT: Let Rezzy reply automatically. If we are worried about her making mistakes, we can put a signature saying she's an AI, or we can select draft only, then choose what to send. Having an automation without automation seems redundant.
#2 Feature in inbox to use Rezzy to spell and grammar check what you wrote. This would definitely give hosts motivation to make the switch and pay for this service.
#3 Ability to use Rezzy when making property descriptions directly from the description box. Have a button that opens a pop-up box. Rezzy should pull from the amenities to populate a description; just have a note saying fill at amenities first, or the description will be inconsistent.
#4 Take this opportunity to incorporate Rezzy into the portals to help answer basic owner statement questions and reports, like what is my year-over-year revenue, and so on.
#5 Ability to add widgets to the website to help answer property questions, basically a chat window that follows users across OwnerRez-hosted websites, or go the Host Buddy route and make it only on property pages. This isn't really urgent for us; it's on the bottom of the list.
Here is the tone I gave mine:
Vibe: Polished, efficient, and helpful. This AI is like the front desk at a boutique hotel. It's perfect for luxury apartments, high-end homes, or for hosts who want to convey premium service and professionalism.
Characteristics:
Formal and polite language ("Welcome," "We are pleased to assist").
Avoids emojis and overly casual slang.
Provides clear, direct, and comprehensive information.
Focuses on service and guest satisfaction.
Examples:
Welcome Message: "Welcome to The Penthouse on Central, [Guest Name]. We are delighted to have you as our guest. Your comfort is our top priority, so please do not hesitate to reach out with any requests."
Answering a Question (Wi-Fi): "Certainly. The Wi-Fi network name is Penthouse_Secure and the password is CityView_2525. Please let me know if you experience any issues with the connection."
Checking In: "Good evening, [Guest Name]. I am just checking in to ensure your arrival was smooth and that the accommodations meet your expectations. Is there anything I can assist you with at this time?"
Handling a Problem: "My apologies for the inconvenience regarding the coffee machine. I have logged the issue, and our property manager will be in contact shortly to arrange for a replacement. Thank you for your patience."
Check-out Reminder: "Dear [Guest Name], this is a courtesy reminder that your scheduled check-out time is tomorrow, [Date], at 10:00 AM. We hope you have had a pleasant and memorable stay."
This continues to happen, see an example from today, September 25th...having this kind of made-up messages crafted by Rezzi are really troublesome, as it could lead a guest to go an enter a house, when they are not supposed to do that.
Rezzy continues to provide conflicting information, here the guest asked the kind of the coffee maker we supply, we wrote the reply, then he asks if he needs to bring coffee filters, and Rezzy said "No, you dont need to buy filters", then "Coffee filters are not provided", then "you can pick some at a local store).
Rezzy continues to provide conflicting information, here the guest asked the kind of the coffee maker we supply, we wrote the reply, then he asks if he needs to bring coffee filters, and Rezzy said "No, you dont need to buy filters", then "Coffee filters are not provided", then "you can pick some at a local store).
Thanks for reporting this Ariel. Going forward, please make sure that you give Rezzy a thumbs up / down rating on those draft replies. If the response isn't good, that thumbs down feedback is important. As I'd mentioned previously, they're the most effective way to help us calibrate Rezzy to better match what you're looking for.
I don't think Rezzy knows how to calculate time periods, see this, current time is 10:23am, we are saying the guest we will send an email 12:05am tomorrow, and rezzy says "in two hours"
Look at all of the blank space above.
Note there is no way to expand the text box to easily edit the proposed text.
Verbose answer.
Clicking Edit does not edit the proposed text which stays above the text box that gets filled in. Not a smooth work flow as you have to Edit and then click delete to get rid of the proposed answer.
My previous experience with Besty.ai was similar, almost too much effort for the output.
Look at all of the blank space above.
Note there is no way to expand the text box to easily edit the proposed text.
Verbose answer.
Clicking Edit does not edit the proposed text which stays above the text box that gets filled in. Not a smooth work flow as you have to Edit and then click delete to get rid of the proposed answer.
My previous experience with Besty.ai was similar, almost too much effort for the output.
By default, Rezzy's first response to a user may be a bit wordier than follow up replies. Looks like this was Rezzy's first time replying to this guest, so that may explain some of the length there.
You can make adjustments to Rezzy's Conversation Tone to hone her in. I took a look in your account, and it doesn't look like you've added anything personalized for her Conversation Tone. It's just using the default right now. We have a support article that goes into that which you can find here - https://www.ownerrez.com/support/articles/rezzy-ai-setup-configure#tone
Perhaps other users in this thread may be willing to share their experiences and practices with their own Conversation Tone. I believe some have already shared that here. You might scroll through some older replies to this post to check those out.
As for your concerns with the editing flow, could I ask you to expand on that?
Clicking Edit does not edit the proposed text which stays above the text box that gets filled in. Not a smooth work flow as you have to Edit and then click delete to get rid of the proposed answer.
I'm playing with this in my own account now, and clicking in to edit a Rezzy draft pulls the draft content into the compose box. Once I send my revised note, the Rezzy draft disappears from the thread. There's no need to delete it.
Rezzy is still not showing the translation of guest' messages, but it does undestand them, as you can see here, where the reply from Rezzy, is correct
We have property level manuals/guides which are valuable resources to skim and learn from. We host our website, not OR hosted. It would be great to provide these links on property level for Reezy to go and learn. It has expandable items, so Reezy should be able to expand and read too. I checked directives but I do not think it covers this.
I think I know when this happens: if I post a reply, and then send an image, the first post (text) is not shown in Rezzy, only the second post (image) is there.
Staff user access to Rezzy recommended responses.
My staff user had access on the first day of us receiving beta Rezzy access. Then the next day it was gone and never came back.
Are staff users intended to have access or not? Do they have to be admin?
