83 – Advanced Sentiment Analysis and Reporting
In this episode of the Suite Spot, we celebrate another exciting solution update launch with Host Ryan Embree and VP of Product & Technology, Jason Lee. Ryan and Jason roll out and explore the new advanced sentiment analysis and reporting for hotels.
The episode starts with Jason defining what review sentiment is and what was behind his team’s initiative to enhance Travel Media Group’s sentiment analytics and reporting. Ryan and Jason then walk through several use cases for leveraging sentiment analysis and how hoteliers could use this data to make better operational and capital decisions for their property. They wrap up the episode talking about how this new update fits in TMG’s suite of reputation solutions and how much it will cost hotel partners.
If you are interested in finding out more about Travel Media Group’s new advanced sentiment analytics or to submit a question for future episodes, call or text 407-984-7455.
Our podcast is produced as an audio resource. Transcripts are generated using speech recognition software and human editing and may contain errors. Before republishing quotes, we ask that you reference the audio.
Hello everyone and welcome to another episode of the Suite Spot. This is your host, Ryan Embree. Thank you for listening, wherever you are listening from out there. We’ve got another jam-packed, exciting episode for you today. With me today, is a very familiar guest, you’ve heard his voice on here and with this guest typically comes some very cool, exciting new news. Jason Lee, Vice President of Product Development and Technology at Travel Media Group. Jason, welcome back to the Suite Spot.
All right. Thank you, Ryan.
Yeah, and we’ve got some exciting updates to announce on this episode and it all starts with sentiment analysis and a complete revamp and upgrade and advanced reporting for the sentiment analysis for reviews. So Jason, before we jump into these updates, I want to just start at a basic level. We’ve done it before, but I think it’s best to reset and explain to some of our listeners, what exactly is sentiment analysis when we’re talking about hotel’s online reviews?
Okay, so when you think about sentiment, you think about what is maybe the mood of the person writing the review, right? That could be part of sentiment, like, is this person satisfied? Is this person unhappy/happy? Are they a mix of emotions, right? So there’s that part of sentiment, but then there’s also the other parts of it, which detail what caused that emotion. So we use something that a lot of people use, especially in machine learning or AI, which is natural language processing. So it basically takes all of this text. So you have tons and tons of text that guests have given you over time. And we take that text and then we’re able to break that down into various layers. So one layer would be the sentiment of that text, which would be whether someone, whether someone was maybe generally happy or unhappy, but then we could then break it into elements. And then where there, they might be happy, neutral, or unhappy with those elements. So, so it kind of starts with natural language processing and then determining sentiment. Then it turns into clustering aspects or clustering topics. And these would be things like this happened in the room that had happened in the lobby. Did it, you know, what, where, where are we talking about? And then it, then it goes more into detail into what exactly happened in that space. So if it was something with cleanliness, what was not clean, and maybe there was more than one thing, but, but basically taking all that text and then pushing it into these like more digestible pieces of data and then, and then displaying that. So, so you can see over time if, from cleanliness, if it’s, if it’s something that is specific, like, let’s say, it’s your floors. If floors are having a cleanliness issue, is this something that’s been happening over time or is this something that’s recent? And if it is recent, then could it be something that as you fix it, you could see it being fixed through the semantic data that’s being pushed through our system.
Yeah, it basically takes all of the noise surrounding reviews and really just gives you the meat and potatoes of what you need to make those operational decisions. You know, I think the best way to describe it is if you’re listening to this, and you are a hotelier that goes through and hopefully are reading your reviews, I want you to think of the longest review that you’ve read, and we’ve come across some novels. Travelers love to type kind of these long paragraphs about their experience. Now, imagine if that was almost taken as notes and just broken down into individual segments and parts of your hotel experience and given you an just an overall summary of what you can improve on that’s what sentiment analysis is, is it really keeps a hotelier focused on the things that, that people are loving about your hotel experience and maybe the things that, that you need to work on as well. Now, Jason, one of the things that I love talking to you and the product side about at Travel Media Group, with these updates, there always seems to be this story behind why this was a focus. Why now? So was there just something that sparked saying, Hey, we’ve got to really do some enhancements here, or what opportunity did you see to improve this sentiment analysis?
Well, we had some basic sentiment data and one of the things that’s unique about Travel Media Group is that we actually, in the reputation space anyway, we actually have people that are handling reviews. So, and by that meaning we do review response and we talk about that many times, but we also in the process of review response, we are hand tagging. So we actually have physical people going in and saying, you know, this was positive, this was negative. And kind of doing what you just said and those little, those little bite-size pieces of data to give a hotelier, a good peek into what’s going on. But what we found was that these were things that we defined. So we defined, you know, what, what they should be looking for. So we had, I think, 28 tags that they could choose from, and what we found was that there’s a lot of things that are in between those tags and cleanliness was probably the biggest one where it’s like something that’s clean or not clean. So we could give all this data like, Hey, there’s a cleanliness issue, but we couldn’t provide the subsequent data, the additional information that said, what was dirty, what was the problem, or what was clean? Right. So we couldn’t give that additional piece because we were, we were really just hitting, we reading the review and hitting the tag. What we were able to do is we were able to use this data to be able to create the first part of the machine learning model, because we have been, we tagged at this point, you know, over a million reviews. So we have, so we have all of this data that we inherently own. And so we were able to start to build this model, using that data and then using additional sentiment clustering to really build out a really robust model. And, and, and ultimately, you know, the, the drive behind wanting to do this was that we’re only supplying, basically, half of the story. And, and if we really want real reputation change, I think real service change, real actual operational change at, at any business, but at a hotel specifically, comes from having actionable information. And, and that’s what I wanted to give. And, and I think we have a really solid system that gives that information. Just, I think in a way that is, that is, not only is it actionable in, in aggregate. So taking multiple reviews and creating, creating this kind of aggregate information, but also defining a single review. So there’s just a lot of, a lot of good, in, in being able to supply this data that, that not only, like, hits the high level, but also gives, gives a hotelier real actionable items.
Yeah, absolutely. I mean, there’s a big difference between someone coming to your front desk at checkout and saying your hotel isn’t clean and someone coming up to your front desk and say your hotel isn’t clean because the lobby was dirty or the bathrooms were dirty. And when you start to tag those specific areas, now all of a sudden you can focus on it and create that experience, because that might be the one thing that’s holding them back from a four star to a five star, and improving your reputation. So this is a very, very powerful tool. And I think one of the best ways to kind of describe the enormity of this upgrade is, you know, walking through some of these use cases. You gave me an excellent one the other day about talking about your most recent reviews, your, your previous 10 and uncovering potentially, you know, something like bedbugs or hotel safety, COVID. You know, if these tag words are found within a review, those are occupancy killers and conversion killers. And all of a sudden until that is moved out of the view of the traveler, your conversion rate is all of a sudden going to sit very, very low. So I know you’ve already described a couple Jason, but wanted to give you a quick opportunity to maybe give some other use cases, maybe the good and the bad about how this could uncover some problems and potentially fix some problems for hotels.
So kind of in high level, there’s a bunch of different use cases for it. But I think in high level, we, bunch of bunch of those sentiment clusters together into categories. So you could look at like, let’s say rooms. So rooms is a category and this is, and then, and then it breaks down into an additional, I think, 18 different categories or different sentiment clusters. And then underneath that, there’s probably another 500 keywords that, that live underneath that. So a use case would be, I go in and I look at rooms and I just see overall, are people satisfied with my rooms or not? And this would be taking all of these like tons of elements into consideration when giving that, that diagnosis. And you can see that over a six month period inside of one of our dashboards. So you can look at it that way and be like, okay, yeah, this looks pretty good. Or you could look at, then, then you could break down further and say, I want to now go into a, to a sentiment cluster or into a tag, and I’m going to see cleanliness and that now I’m just going to look at cleanliness over time. But now I want to, in cleanliness, I want to actually see what’s not clean. So then I go and I look and I see, oh, it looks like I’ve got, it looks like one of the trending categories is bathroom. So then I click into bathroom and I see that I have a problem with towels, right? So that, it’s that kind of like that, that, that, that progression of being able to look at something at a high level and see like, okay, this, this is pretty, this looks good. Or maybe this is this, there’s a small issue here on the high level, and then I can break it all the way down. So that would just be one, that’s one use case. Another use case would be in competition. So it could be like, I can look, I can actually then stack, let’s say I take sentiment data for all hotels in the United States. So we take, we can take that same sentiment data that your competitor, that you have, and we can now have you compare yours against theirs. So you can see how are guests feeling about my rooms, compared to guests feeling about their rooms, right? So you can kind of get an idea about what’s going on with your, with your competition, other than just that high level. And a lot of times with competition, we look at rates. We look at, you know, maybe we drive by there and we see how many cars they have in the parking lot, but, but really understanding what a guest is experiencing is, is an important factor that is sort of that intangible. But you know, that the data that you’re mining that from also is the same data that is informing guests of whether to stay with them or not, or whether to stay with you or not. That’s like the power of this, all of this information. But competitors would be another major use case.
I mean, so incredibly powerful. Again, right now with where the industry is, we have staffing shortages out there. We think about investment. Obviously capital investments are one, but also time investments. Hoteliers need to know where they need to be spending their time, where their staff needs to be focused on. So if you have your staff at, you know, and really cleaning rooms and, and focusing on that, and they’re still not getting the job done, you know, that’s a huge indicator there. Or if all of a sudden, you know, breakfast keeps coming up as a negative tag, maybe moving staff to, to, to help people during breakfast to help that be a smoother experience for travelers. So staffing is huge right now and, and using that time wisely and having some data behind that is, is just going to give you as a hotel GM or owner peace of mind about decisions you’re making. And then the whole capital side, right? We are in, we are in the midst of industry recovery. We know that Delta variant’s out there. We cannot afford to just kind of put a blanket out there of capital investment and kind of cross our fingers on. Hopefully our guests are gonna like this and it’s going to improve the experience here. No, we need to think methodically about these purchasing decisions that we’re making. And it could be something as simple as changing out the towels because GM’s and owners are making very tough decisions right now. They don’t have the capital that they did to sit back and say, okay, well, should I replace my towels or should I change the bedding? It’s either one or the other. So having this data in your back pocket will, again, just give you that peace of mind, very, very powerful stuff. And then moving onto your competition. When we think bigger picture here, not only are we finding our competitive advantage, but we need to go one step further, and we talked about this on this podcast, about utilizing your competitive advantage on platforms like social media. If you’re seeing that your competition is, one of their biggest complaints is the pool area or breakfast, then your pool and breakfast area need to be superior. They need to be put up on social media. They need to be marketing and advertising that because that’s what travelers want in your area. So, you know, we, we say it all the time, knowledge is power, but this data that you’re providing hoteliers really, really is powerful and can have some revenue impacts to a hotel.
Yeah. You know, I think sometimes I think about like, you know, when, when I ran hotels and what my mindset was in my day to day, you know, and, and a lot of it was putting out fires around things like you’re talking about around staffing, around guest issues, around bigger issues that would, would happen at the hotel. But, but really truly understanding a guest’s experience at your hotel is, is something that, that was sort of elusive to me. It was something I had a feeling about. Right. And I think what happens a lot of times is that we do have those, there’s a lot of reputation data. Reputation data has been, we’ve been collecting it for our hotels for a long time. And if you’re a hotel that’s been around for awhile, there was, there is, you know, thousands of reviews. Thousands of guests have, have left their, their, input on their stay. But, but really understanding that, and, and, and sometimes it’s almost like, you know, going to the doctor and the doctor’s like, yeah, you do have high blood pressure. You know, you maybe have been feeling kind of weird. You do have high blood pressure. You definitely need to take some medicine to, to get this under control. And I feel the same way about this data. It’s things that we might be feeling, but it’s now where you can actually get some confirmation of the issues that guests are having. And, and what’s really fascinating to me about this data is that when you see issues, the, like the bigger issues, you know, the service issues, the, I guess, room condition issues, those things are consistent. Guests, a guest that has a problem with that, other guests are having that same problem. You see, you’ll see this happening consistently. And in some cases, this could be years. Years of guests having these same experiences that are negative. And it’s not until you maybe take a, take a second and really look at this and get honest with yourself about like, this is what these guests are saying. It’s not one guest that was complaining and one person trying to get a free room. This is a multitude of people that have had a very similar experience that was negative at my hotel. And so taking that information and really saying, okay, I’m going to change this. And whether that is then diving in with your housekeeping and getting cleanliness right or getting with your staff and making sure that guest or service recovery is happening on a regular basis. Those are the things I feel like this, this data does unlock if you can be open to it. And I think that’s, that’s sometimes that’s the hard part is that, you know, anytime change happens, anytime there’s, any time you have to actually make true change in anything- in your life, in business, it requires some sacrifice. It requires, it requires effort. But I do believe that this data really does put forth a really comprehensive way of understanding a guest’s experience over time. Yeah.
You just have to think, you know, that that guest that made it to TripAdvisor or made it to Google or, or gave you that sort of feedback, you have to imagine how many more guests were probably feeling that way, that just didn’t make it to the site. Right? So it’s, you’re right. It’s not just that one guest, it’s a multitude of them. And the scary part to think about it, especially if it’s negative sentiment, is how many of them stayed silent, had this issue and now have either, you know, said some word of mouth or never plan on returning to the property again. So it’s uncovering these and prescribing the solution that’s key and you’re right. There is so much data out there and that is what makes the Travel Media Group reputation experience unique. We are not only providing the data and more of it, but we’re also providing solutions. Our reputation influence helping increase the amount of reviews and positive reviews that you’re getting, as well as the review response partnership, which has been huge with the staffing shortage. Hoteliers are spread very thin right now. So having a review response partner, that’s always going to have your back like TMG, all apart of that reputation suite of solutions, not just the data, but the solutions, the backup that we provide. So I think that’s a great segue into, is first of all, congratulating you about this new launch and these, these analytics, but I’m sure a lot of listeners are wondering, okay, how much is this going to cost me? So, Jason, do you want to break them the news?
It is not going to cost you anything right now. So if you are, so if you’re a current review response customer of Travel Media Group, you’re just going to get all this additional data. You’re already getting a version of it, but you’re gonna get all this additional data. If you’re an existing reputation customer, you’re going to get all of this data really for the first time. So you’re going to see a lot of this stuff for the first time. And this is going to include, initially it’s going to include two really comprehensive reporting suites that have to do with review sentiment over time, and then an advanced sentiment report. And then it’s also going to have a sentiment breakdown by review in each individual tile. So each of those, so you would see a TripAdvisor review, you’d click on show sentiment, and it actually shows all of the sentiment pieces to that review, but then also highlights the sentences that are positive and negative. So it gives a lot of information. Helps you kind of like weed through, you know, a review pretty quickly where you, you know. Sometimes you go, I go to a review, oh, this was the review that talked about cleanliness, but it might be a paragraph long. And you don’t, you got to read through the whole thing to find the sentence, well we highlight that sentence for you, so you can kind of pick it out quickly. So yeah, so it’s just going to be part of the suite and cause really what this is about, it’s, it’s ultimately about how, how, how can we help, you know? And like Ryan was saying with respond and resolve, we can help with, with, you know, responding to reviews. So, so when you have, when you’re short on staff, short on time, we have, we have you covered on that. On our influence side, it helps you get a better quantity and quality of review. You know, and then, and then the analytics really are supposed to be kind of showing, you know, how, how you’re performing and helping you understand where you sit in terms of, being able to capture guests and, and convert them into bookings. But in this case, this, this really is this other part of it, which is, you know, how can we help you better understand the guest experience and, and then, and then really dig in and make changes. But one thing I was going to say too, Ryan, is that like a lot of the examples I gave were negative, but sentiment also has positive aspects to it. It tells you what you’re doing right and it gives you great opportunities to reward your staff, reward great staff members that are standing out. Maybe a maintenance guy that, that went the extra mile and, and, and did some, did some stuff for, for guests that they, that they really praise them for. So all of these, those pieces do get highlighted positively as well. So there were some, there’s some other really great opportunities other than just self-help, there’s some, there’s, there’s some good, like self-congratulation stuff in there too.
Absolutely, super important to mention as well. And again, that’s giving you a competitive advantage and showing you what you’re best at, at your hotel. And at the end of the day, travelers want a level of hotel experience and have expectations of that level of experience. So if we can show, if we can show and share where we’re exceeding those expectations and focus on that, I think that’s just as important as fixing, you know, what, what they’re finding to be subpar. So, Jason, absolutely unbelievable episode love to hear all these launches and innovations that your team is doing on the product side. So congratulations again. You know, I just would challenge our listeners, whether you’re working with us or not, you know, 2022 is going to be a huge year for the industry. The reputation game a decade ago was almost a sit back and react. And, and now we have hotels that are attacking their reputation. They are proactively going out and they’re using tools like this, to, to essentially manage it the same as you would your, your rate system, because it, it it’s that powerful and has the impact of conversion is just, is, it’s just incredible that we’re finding on our side. So, you know, before we close out today, I’ll just, you know, ask for any final thoughts from you, Jason, and then we’ll obviously give a number if they’re interested in kind of learning more about this advanced sentiment analysis and reporting. So Jason, any final thoughts?
Yeah, one final thing I was going to say, and I, I kind of like glossed over it when we went in the middle of this, but we actually are pulling sentiment data. So we’re running this, this advanced machine learning model on all reviews for all hotels in the United States. So if you’re listening to this and you know, you’re, you’re interested in, in talking with us, we could actually have a conversation with you and actually show you the stuff that’s going on at your hotel right now. So, so that part of it is, is really exciting, but what it also gives us or what it’s also going to give us in the future is that we’re what we’re where we’re at right now with, with these reports and, and tile action, that’s just the tip of the iceberg. Where we’re headed with all of this data, so once we have all of this data, we’re going to be doing advanced analysis on this. And, and so this is the stuff that’s going to be really exciting coming. And if you’ve been a customer with Travel Media Group for awhile, you know that we keep adding stuff to our suite, and we’re going to continue on that path of just giving more and more relevant information to hoteliers to help them better serve their guests.
You’re absolutely right. It’s a super, super exciting. That means we’ll definitely have to have you on again to announce all of these innovations. But you’re right, you know, I would say if you’re a hotelier listening out there, give us a call. If you haven’t before, reach out to us. You can reach out to us at (407) 984-7455 or visit our website at travelmediagroup.com. We talked to a lot of hoteliers that think they have a good understanding of the places that they need to fix and the gaps in their hotel experience. But once we pull this data, you know, we might give you some insights that you didn’t have before. So no risk in reaching out. You know, we’d love to talk to you. That’s what we’re here for. That’s why we started this podcast to help hoteliers like yourself. So, Jason, I’ll go ahead and thank you again for being on the Suite Spot. I want to thank all of our listeners out there and we’ll talk to you next time on the Suite Spot. To join our loyalty program. Be sure to subscribe and give us a five-star rating on iTunes. Suite Spot is produced by Travel Media Group with cover art by Bary Gordon. I’m your host, Ryan Embree, and we hope you enjoyed your stay.
Hotel Marketing Resources
113 – TMG Hospitality Trailblazers: Alexi Khajavi
Join Suite Spot podcast host, Ryan Embree as he welcomes the President of Questex, Alexi Khajavi, as an official TMG Hospitality Trailblazer! These two hospitality-focused individuals discuss some incredible travel and tourism subjects in this episode, giving viewers...
112 – Responding to Guest Satisfaction Surveys
In this episode of the Suite Spot Podcast, host and TMG Marketing Director, Ryan Embree and Chief Technology Officer, Jason Lee, tackle how guest satisfaction surveys impact a hotel’s operations and influence overall bookings. From topics such as what questions to ask...
111 – TMG Hospitality Trailblazers: Patrick Pahlke
Join Suite Spot podcast host and TMG Marketing Director Ryan Embree as he sits down and welcomes a very special guest, the EVP & Chief Commercial Officer of Sage Hospitality Group, Patrick Pahlke, in the latest episode of the TMG Hospitality Trailblazers series! ...