The difference decision intelligence makes

In this week’s episode of the Hotel Moment podcast, Bryson Koehler, Revinate CEO, guest hosts to interview Jeff Jonas, Founder, CEO, and Chief Scientist of Senzing. Jeff paints two pictures — one in which hoteliers practice responsible data management and another in which hoteliers fail to make intelligent decisions because of disorganized and inaccurate data. Identity resolution is one of the most important tools Jeff believes every tech stack should have to manage their data, understand guest profiles with accuracy, and execute data-powered marketing initiatives.

Tune in for a detailed conversation about data management and how it impacts your hotel’s data security and profitability in a world dominated by AI influences.

Meet your host

As Chief Marketing Officer at Revinate, Karen Stephens is focused on driving long-term growth by building Revinate’s brand equity, product marketing, and customer acquisition strategies. Her deep connections with hospitality industry leaders play a key role in crafting strategic partnerships.

Karen is also the host of The Hotel Moment Podcast, where she interviews top players in the hospitality industry. Karen has been with Revinate for over 11 years, leading Revinate’s global GTM teams. Her most recent transition was from Chief Revenue Officer, where she led the team in their highest booking quarter to date in Q4 2023.

Karen has more than 25 years of expertise in global hospitality technology and online distribution — including managing global accounts in travel and hospitality organizations such as Travelocity and lastminute.com

Connect with Karen

Transcript

Karen Stephens Intro – 00:00:04: Welcome to the Hotel Moment podcast presented by Revinate the podcast where we discuss how hotel technology shapes every moment of the hotelier’s experience. Tune in as we explore the cutting-edge technology transforming the hospitality industry and hear from experts and visionaries shaping the future of guest experiences. Whether you’re a hotelier or a tech enthusiast, you’re in the right place. Let’s dive in and discover how we can elevate the art of hospitality together.

Bryson Koehler – 00:00:36: Well, hello, everybody, and welcome to the Hotel Moment podcast. I’m Bryson Koehler, the CEO of Revinate. It is great to be with you guys again. Thanks for joining us, where we talk about a lot of things that are facing the hospitality industry. How do we continue to help asset owners and management teams continue to improve the experience that they give to their guests, the booking experience, the journey to the property, the on-stay and after-stay experience, and just continuously working to make hospitality great and profitable for asset owners and management teams all around the world? And as you know, if you’re a listener to the podcast regularly, at Revinate we believe data is key. We believe that data is foundational to really understanding the guest, making sure that you’re getting the right guest to the right hotel, into the right room, at the right rate, through the right channel, and repeating that at scale. We believe — start with data. The more data you have, the better you understand that data, the better you can activate it, the richer and more powerful the outcome can be. And so today, it’s really a great honor and privilege to have an old colleague and friend of mine, Jeff Jonas. He and I worked together at IBM back in the day. Jeff was, and still is an IBM fellow, and he ran the contextual computing work at IBM after IBM acquired a start-up that he had started prior to that. Since then, Jeff is the CEO and Founder of Senzing, which is an awesome entity resolution capability, and one of the leading entity resolution capabilities out in the market today. And Jeff is one of the smartest, most engaged people I know around data, entity resolution, AI, and the future of compute. So, Jeff, welcome. It’s great to have you here today.

Jeff Jonas – 00:02:34: That was very kind. I love the way you talk about data because it’s how I feel. It reminds me of an O’Reilly book on beautiful data.

Bryson Koehler – 00:03:01: Data to me is alive. It’s breathing. It has a heartbeat. And you can feel the data flowing through the veins of your company, across your pipeline. And it’s always living. And it’s an organism that you have to care for, and feed, and grow, and develop, and educate. It really is, to me, just a part of the soul of your business.

Jeff Jonas – 00:03:05: I think about data as an observation space. It’s like if you were to talk to an organization as a whole, and you want to know how smart it was, and what did it know, you’re really asking about what its observation space is. And that’s all that data that’s flowing through. Some of it throws in, shows up a little batchy. Some of it’s showing up in real-time streams. And the question at the end of the day is, with every piece of data that lands, you just learn something. Does it matter? And if so, to who? When you get into real-time data, which is one of my obsessions, you’re about making better decisions faster than your competition.

Bryson Koehler – 00:03:31: Yeah. And look, I mean, data isn’t free, especially in today’s world where you’ve got to secure it, and maintain it, and replicate it, and all of the things that you have to do as a good custodian of data. And if your data is just sitting there collecting digital dust, and you’re not able to activate an action off of it, what are you doing with it? So you’ve been working in and around data for your career. How have you seen data, the access to data, the amount of data, the relevancy of data, interesting new data sets — how has that been changing? And have you been seeing it really accelerate over the last couple of years? Where are we on this data spectrum?

Jeff Jonas – 00:04:12: I’ve been wrong about this for decades. I have kept predicting that data will consolidate. You’ll get more unified views. There’ll be fewer piles. You’ll have more keys to understand who’s who. But what’s really happening is that thermodynamics law called, “entropy”, is winning. The universe is trying to break up into small pieces and spread out. And some big projects we did in the 90s for in hospitality would have maybe five big data sets. But that’s changing. There were more data sets ,and more sources, and more third-party data that you might want to combine to add some new columns to understand, like a latitude, longitude, or a buying preference. Or are there… Are they an influencer? And so it’s growing. There’s a greater number of piles. And the way I view data is, it’s like puzzle pieces. Like the blue puzzle pieces over there, the yellow puzzle pieces over there. There’s magenta. And if you can’t bring the puzzle pieces together — that’s context, by the way, context computing — if you can’t bring the puzzle pieces together, you really don’t understand who’s who.

Bryson, what’s the most number of data sets you’ve seen in one hospitality company that would be data you’d want to combine?

Bryson Koehler – 00:05:21: Some of our most complex customers are approaching 100 different data systems across. Because you think about a very complex five-star resort that has 15 different food and beverage outlets, and a spa, and it has golf, and it has all of these other types of activities like the kayak rental shop down at the beach. And, like, it adds up. And then all of the normal things that we see — Wi-Fi systems and where you’re logging into that, which is a different data store, and then your key card system. And just. It’s crazy. But yeah, we’ve got customers today that have upwards of almost 100 different systems trying to connect together and operate that property.

Jeff Jonas – 00:06:05: Entropy is winning. It just keeps up.

Bryson Koehler – 00:06:08: Well, you know, I guess in the day of more vertical compute needs, you really had this need to pull it all together because to reach out over the network was really expensive latency-wise, cost wise, whatever. But in today’s horizontal scaled computing, maybe it doesn’t matter as much as long as you can find ways to key, and link it ,and bring it together, and identify how you attach those attributes back to an entity. Maybe it doesn’t matter that that entropy is winning.

Jeff Jonas – 00:06:39: The record link – kehyingand linking, which is kind of in the same category of what we would call entity resolution, is determining when two parties, two people or two companies or vessels, whatever, are really the same, even though they were described differently. When they made their reservation, they used this information. When they signed up in the loyalty club program, they used that information. And you don’t have three customers. It’s really one. And then when you do that, your opportunity to also figure out who’s related to households, families that share email addresses, even though they don’t share the same address. And that becomes really powerful as well for understanding value and risk. We’ve done a bit of work in hospitality for risk, especially in Vegas.

Bryson Koehler – 00:07:17: Yeah, I think — look, if you have multiple views of the same carbon-based life form, then you’re going to end up displeasing that person, whether you market to them too frequently, or to your point, risk. You’ve done a lot of great projects across a multitude of industries. How are you seeing your customers think about the ROI of entity resolution and getting that right? And maybe you can talk broadly, and then maybe give a couple of hospitality examples to bring it home for this podcast.

Jeff Jonas – 00:07:46: If you think you have 20% more customers than you really have, then you make a whole bunch of downstream decisions poorly, which would include machine learning decisions, would include what kinds of marketing to send to who. You just get it wrong. And it can damage your brand. A great example from casinos is if you have what’s called a “whale”, which is somebody that’s maybe dropping a million dollars a visit, maybe more. When you do that, if you don’t recognize the whale as the whale — and If you show up on Thursday, we’ll give you $2 off the buffet They’re just like, Hey, you do not know who I am. I mean, you’re taken away from your brand. It’s really a data quality thing because if you can trust who’s who, and who’s related to who, you can make exquisite decisions and offers.

Bryson Koehler – 00:08:30: And how have you seen the accuracy of that improve over time? And kind of where are we against what would be the ideal state where we have 100% confidence that if we’ve got these two duplicated profiles, this is actually the same person? We’re going to merge it together. We have 100% confidence. Where are we on that continuum of confidence on data sets and entity resolution?

Jeff Jonas – 00:08:57: There’s been some really nice breakthroughs. One we would call “entity-centric learning”, and that is when you can learn over time all the different addresses they’ve used. If you’ve seen two dates of birth, you’re keeping them both. You’re not throwing one of them away. And if you’re assembling what you know about somebody and the growing number of features, your ability to find that other record because they changed their name, their last name, because they’re using their nickname over here. They’re using “Rudy” instead of “Rhodey”. And so when you’re building a union of the different names and ways they’ve expressed themselves, the different dates of birth. By the way, it’s a subtle story, but it’s a fun one, is this is a bad daddy story. I have a son with two dates of birth. It’s quite tragic. And here’s how this happens. This kid gets born on this day in September, but I have convinced everybody, the mom and the grandparents, that it’s a different day and we celebrate his birthday on the wrong day for five years. His birthday is wrong in every system, every healthcare, Riky Rick magazine, everywhere. I order his birth certificate, take him to Mexico, and to my surprise, I would have been wrong. Now, imagine that. Now, there’s a second date of birth that emerges. Any good data quality system would notice there’s 200 of them that are right, and now there’s one that’s different. You would suppress it. You would not let dissent fester, but it turns out you have to be able to let dissent fester. Otherwise, it can’t slowly amass. So when you think about data and entities like that, where you’re letting the features fester, the quality gets better. And in fact, you could say that bad data is good in this regard. It’s really weird. When you search Google and it says, Did you mean this? It didn’t have a dictionary. It remembered everybody’s errors. My brother’s name is Rhodey, but now and then, people spell it Rudy or say Rode, R-O-D-E, to spell it the wrong way. But if you don’t learn that over time, it’s not as smart. So anyway, that’s one of the things that’s happening in the identity resolution space that’s riding the accuracy up where you can make important decisions with more confidence.

Bryson Koehler – 00:10:52: Yeah, I think back to my weather days and the implementation where we started to bring in personal weather station data. And we had meteorologists that were aghast that we would want to bring in this dirty data that —- you can’t bring in the personal weather station data that’s sitting on top of somebody’s garage. We know it’s wrong. We know it’s not right. But you actually do want to bring it in because, you can actually, over time, start to understand its comparison against what would be known good data. And then as you start to really get more specific on your last long example, you can start to look at the deltas. And the delta itself is way more valuable than the data entity or numeric value that it is feeding. And so dirty data can be just as important as clean data.

Jeff Jonas – 00:11:41: That’s a great example.

Bryson Koehler – 00:11:43: And when you think about fraud, you think about, where bad actors are trying to manipulate data. And sometimes they’re doing that. And I guess we can extend that a little bit into data poisoning, where we’re purposefully manipulating data in ways to gain outcomes. But sometimes it’s not that sophisticated, and people are just trying to manipulate systems or to cause fraud against a company. How do you see people that actually do the work to get the entity resolution right, get their data clean, you know, brought together and these profiles created? Does that give people a leg up in your experience when they’re trying to surface bad actors and fraud?

Jeff Jonas – 00:12:23: We’ve done a lot of work with bad actors over the years, or we’re helping companies deal with bad actors, I should say. And the number one tradecraft for bad guys, and this goes from terrorism to drug cartels, is this single tradecraft called “channel separation.” They do not want you to realize it’s the same person that transferred this money, that did this. And to do that, you use identity obfuscation. Now, you and I do channel separation too. If I send you an encrypted spreadsheet and then call you with the password, I have separated the channels. But bad guys do this with their identities. They don’t use the same name, address, and phone on every record. So how would you ever find them? Interestingly, that’s where we first developed entity-centric learning to learn. And we saw this in Vegas. You have people that have 32 different names they use, eight social security names, five dates of birth, and they use different combinaton you don`t have a system that can support those disagreements, five different dates of birth, eight social security numbers. But when you have systems that do that, your ability to find that Ken and Mark are the same people is because there’s other records in between that create the glue. And that leads to this technology for Vegas we built was called “Nora”, Non-Obvious Relationship Awareness. It’s ultimately what we sold to IBM in 05. And when you do that real well, it allows you to take just traditional messy data. And a messy data in hospitality is just a hotel reservation data. It could just be anything, really.

Bryson Koehler – 00:13:49: Well, you’ve got an opportunity if you’re doing a voice call and a voice channel reservation for somebody to fat finger something, you’re coming in through a third party channel who doesn’t share all of the information with the hotel. You arrive at the hotel, and there’s missing info on your folio. And so you hand your license over to the front desk, and they’re typing that in. There’s so much room for just human error that you’ve got to get really good at understanding that, Wait a minute. No, no, no. These are the same. And we’ve all experienced this where you show up at the hotel, and they say, Hey, Jeff. Welcome. Is this your first time? And you’re like, No, I’ve been here 24 times prior. And it’s not because the hotel doesn’t want to welcome you back, it’s that the system, and their data is failing them to really deliver and delight. And I guess that’s the same risk that happens on fraud. If you can’t pull that together, you’re not going to realize that this one actor is defrauding a points program, or maybe they were banned from a casino, and they’re still showing up, but they’re showing up under a different name. The use cases all come down to — you have to get your data organized.

Jeff Jonas – 00:14:59: Yeah. You’ve reminded me, I showed up at one of the biggest brands in the US. I’m not going to name them because it’s just embarrassing. Okay. I show up. I – travel agent did not put in my loyalty card. They go, Yeah. Are you a loyalist club member? I go, Yeah. So they searched me up and they go, Well, which of these Jeff Jonas’s are you? And I’m literally all three. I go, I’m all three. I go, Well, put them together. They go, Oh yeah, we really can’t do that. There’s some kind of phone number. We really can’t know. So I have to just go, Okay, well, which one’s got the most points so I can just?. I pile onto the winner.

Bryson Koehler – 00:15:29: And that happens all the time, and the systems, and who really has the system of record. And obviously, that’s the work we do at Revinate, the work you do at Senzing, and just is — we’ve got to get to a place where we continue to increase the confidence level, where we are merging those duplicates, where we are finding errors and filling in the gaps and the holes. And it’s been great to see — I think you’ve been working in and around hospitality for quite some time. If you think back to maybe some of your projects, probably dating back into the nineties on this, you saw great results even back then, I could imagine. Maybe talk about what a project might’ve looked like back in the nineties, and where we are today with the capabilities we have with data machine learning AI. This has been quite a journey.

Jeff Jonas – 00:16:16: This has been quite a journey. Your question takes me back to this. One of my favorite projects ever was for Cendant. Some may not know the name. It was originally an HF, hospitality franchise. It becomes Cendant — 5,000 data sources. They own Ramada, Howard Johnson, Super 8, DayZ, NightZen, Travelodge, Wingate. Also by the way, they own Avis and Budget Rent-A-Car, Century 21, ERA, Coldwell Banker, Home Welcome Wagon, Jackson Hewitt. The list goes on. 5,000 data sources. We had to learn — this is mid nineties — gave us a whole bunch of really important architectural things about how do you even do that? How do you build a system that can take a billion records, about a hundred million people, and create that? There was like a Gen-two engine for us, but build that in 18 months. So many things were learned, but here’s what the system does. When you can combine all that data and have confidence in it, the quality of predictions, and this is the story from back then is, you can literally name who’s going to go to Orlando in the second week of May. And you could also know because of lead time to booking. That’s the attribute here. If you can see enough lead time to booking information, you know, whether that person is easier to influence about when you market to them two days before that trip, two weeks before that trip, two months before that trip. And a decade or two after that, I reached out to the CIO of that group just to catch up because it was a great project. And I said, how was that project? And he goes, It was the, maybe the greatest project I ever worked on. And they use that marketing intelligence to drive enormous by far paid for that project. When you have those kinds of insights. Yeah.

Bryson Koehler – 00:17:48: Well, when you get confident enough that you’re willing and able to market and take action in a different way, you can actually have a targeted message versus like, Well, I’m not really sure. So I’m going to water down my message. And I got to create this segment that’s 80%, because I don’t really know. Versus I’m going to create a segment that’s 10%, and I’m a sniper now versus kind of a shotgun. And I’m really able to get, hone in on it. If you think about the evolution of that since the mid-nineties to where we are today, it seems like we’ve come a long way, and the other side of it, it seems like we’re just at the start. You think about where we are on our AI journey. AI today, obviously is a big piece of what you guys do, a big piece of what Revinate does. Maybe talk a little bit about in your product today, how do you guys use artificial intelligence, machine learning to get the results that you have?

Jeff Jonas – 00:18:45: First, maybe for the audience here, I will separate out AI from ML.

Bryson Koehler – 00:18:47: That’s why I said it is two things.

Jeff Jonas – 00:18:49: Yeah. But it’s not obvious to everybody because sometimes it gets bundled up in the one thing. And I learned this phraseology while at IBM, and I’ve loved it since. AI are systems that act “human smart.” Machine learning are systems that learn to experience. Many AIs use ML. They don’t have to, you can have a system that’s “human-smart.” And so that’s the way when I’m using the terms now, I’ll be using it in those. On the AI part, our system, it figures out who’s who. And to be clear, we sell transmissions, not cars. Meaning, people who build systems for the market need matching. We have this matching thing that they plug in that makes that easy.

Bryson Koeheler – 00:19:28: Yeah, Revinate being a system, the car, and you being a transmission component in that car.

Jeff Jonas – 00:19:35: Yeah, so for people that build software, we don’t have user interfaces. People that build software, they can either go try to build high-quality matching for $30 million and find it hard. We spent $50 million building this Gen 6. But our AI, the part that’s human smart is, we beat humans consistently in accuracy. Oh, a human will find a few things we missed, but we will find more things the human missed. The machine learning is, we’ve done things like, we’re using our name comparator, for example. How you compare names has been trained on 850 million names. The way you compare a name that might be Spanish with five parts versus Arabic. If it’s an Arabic name with Mohammed in it, and Din and Haj, Bin and Haj aren’t part of the name. If you don’t first understand the culture, you can’t match the name right. Addresses are so problematic. The machine learning work that we’ve done on parsing global messy addresses to figure out which pieces of the addresses really is a unit number versus the street number versus the name versus a region. If it’s in India.

Bryson Koehler – 00:20:33: India is still the hardest, right? Like, that’s my recollection.

Bryson Koehler – 00:20:37: It is definitely one of the hardest. Well, there was one other country that was actually harder, but it’s small. So we use traditional types of machine learning for doing out-of-the-box name and address matching. And then we’ve invented a real-time learning. Get this. You’ve ingested a billion records. You get a record billion and one. So your observation space is increased by one record. You have to say, Now that I know that, where does it fit? Is it somebody I know? Is it somebody new? Is it somebody related? But now that I know that, over the billion decisions I made, should I have made any differently? Did you just learn something where you now realize a junior and a senior you conflated, you should take apart? And I’ll give you a very interesting, another value prop, which I’ve only stumbled into recently because one of our customers has shared this with us, is, when you conflate records, you’re creating a data breach. Because when somebody logs in to look at their record, and they see somebody else’s, you have now exposed PII or financial data or health data to the wrong person.

Bryson Koehler – 00:21:35: I don’t want your folio of your stay. In my loyalty program history.

Jeff Jonas – 00:21:41: Oh, yes, you do. I’m kidding.

Bryson Koehler – 00:21:44: Maybe I want the points, but my wife might ask me like, Wait a minute. I didn’t think you were in Hollywood last week? So, those are great examples. And I love your way of thinking about artificial intelligence and machine learning. Again, we’re still in the infancy of after we kind of hit that economic inflection point where it began to make more commercial sense to leverage this more broadly the use cases are just going to continue to explode. And you’re a researcher at heart. And that is who Jeff is. So with your research hat on, where are we going? And where is it, large language models, small language models, the whole generative capability set that we’re kind of having hit that inflection point of the cost of compute, and all of this get to a place? And where do you see us going? And maybe specifically, where does that go? Going around consumer and consumer behavior.

Jeff Jonas – 00:22:39: I’ll start with what I see in these large language models, which is really, as everybody, I think, realizes an amazing breakthrough. I do believe it’s being overestimated by, let’s say, 50%. But its utility is still enormous. My litmus test about, Are you using it the right way? I’ve reduced to this Reader’s Digest version. If it gives you a different answer tomorrow, is that going to be a problem? If that’s true, you’re using it the wrong way. For example, I need to talk to my friend about this complicated topic, and what’s the best way to broach it? I haven’t talked to them in two years. So what’s the best way to get in there, and have this conversation? You’ll get a different answer every day for that question. But basically all of those will work. Or I want a recipe for a gluten-free menu for next week. You’ll get a different answer every time. But when you start asking about making decisions about people’s privileges, which door you’re going to knock down, who’s not going to get credit, and you get a different answer every time, that’s a problem. So I’ll call it flowering up, but adding language around things. And that’s why this thing called RAG, Retrieval Augmented Generation, for those that don’t know what that is, is when you’re having a conversation with a large language model, that’s going to — you ask it a question, it gives you back a whole page. RAG is like an antipsychotic. You inject in, in the case of let’s say hospitality — your customers, you’re injecting in who’s related to what products they’ve had, when they’ve stayed. It’s all assembled. It’s called a blob. It’s a grounded set of facts. And then it adds fancy words around it with a low chance of editing that. It can make you get an antipsychotic out of that. And I think that that’s going to show a lot of utility. And there’s a lot of folks working on that in R&D as we are.

Bryson Koehler – 00:24:20: And a lot of that R&D continues to be on how do we continue to make the differences of those answers that you might get day to day become more specific to me versus you? Learning that the way you ask something, and what you prefer, and what I ask and what I prefer, were different. That might continue to evolve from a capability set on how economic it is for us to build a small enough kind of language model that works for you and me differently. Even if the underlying kind of foundational data and model is the same.

Jeff Jonas – 00:24:55: Entity resolved knowledge graphs. There’s this whole thing about connected data and knowledge graphs. It’s about the entities, and the things you know about them are the graph of who’s who, and who’s related to like the scaffolding. And then on that, you’re addressing it with their events and their transactions. And do they use the golf course? Did they ever use those kayaks? And that’s your knowledge graph. But when that is entity resolved, you know who’s who. And you’re injecting that into the LLM, the communications you’re having with them, whether it’s in email, whether it’s in a call center where they’re calling you, you see that 360-degree view. The quality of those interactions goes way up.

Bryson Koehler – 00:25:30: What are you looking at in terms of investing? What’s next for Senzing, and your product roadmap? And what are you really excited about? Because you guys have had a great success story, market leadership. Where are you taking the business next? And what do you see over the horizon?

Jeff Jonas – 00:25:44: Well, this new engine we have, we have spent $50 million building. People don’t realize how hard it is to do your own matching. And we are polishing this now to support different languages. So people have a global business. So you can do business names in Mandarin, for example. We are constantly working on improving quality because there’s always a little bit more, but lowering the cost on cloud compute in the last 18 months. People that are running us, and these are billion recon system using, lets say AWS, we’ve taken 20% of their cost of goods to go out to operate their entity resolution in just the last 18 months. So taking cost out. So it`s more affordable accuracy, and our obsession with real-time, and low latency continues — you`re trying to make answers in the moment. I did a commercial when we were both at IBM on the round smarter plan. And i was standing at the side of the road, and I go, Imagine trying to cross the road when you can only see how the road looked five minutes ago. Businesses make decisions like that every day. It was a great commercial that was played at Wimbledon. But that’s the truth is, if they were just had a bad experience on the website and now they’re ringing the call center for help, you Should know that when your call center person talks to them. And that’s a story of making better decisions in real-time.

Bryson Koehler – 00:26:57: Yeah, I love that. And I remember that one, and I love it. And the rich guest profiles that we create, and a continued investigation around, How do we continue to make this better and better and better for our use case?. which is hospitality specific. My big focus there is around the omnichannel of communication that is out there. And if you call the call center versus you call the front desk versus you chat versus you email, you shouldn’t get a different answer. You shouldn’t get a different story. You shouldn’t get a different rate. You should get a consistent experience. And so helping our customers talk to guests, regardless of the channel, but have a common voice across that channel is a really important piece of what we are working on. And that really starts with data, getting your data organized, making sure that you’ve got the right entity resolution, that you’ve got the right way to synthesize those profiles, whether it’s a carbon-based life form or a silicon-based life form or a concrete-based, whatever. These are your entities. You’ve got to be really good at that. It starts with data. It starts with the plumbing of your data. And then I love your real-time push. I mean, I think that’s one of the most underserved pieces of this market is people need to recognize that in an online transactional runtime. You have milliseconds to make decisions, and things that are taking, if it’s over a couple thousand milliseconds, it’s not useful. And to your point, businesses making decisions off of yesterday’s information, governments making decisions off of last month’s information. You think about how flawed that decision-making is. It’s great to see you guys embedding your SDK in so many different use cases. It seems to me that that’s a key part to making decision intelligence in the world a better place.

Jeff Jonas – 00:29:04: Indeed. Going back to the omni-channel, if organizations feel overwhelmed with their omni-channel entropy, if you think you’ve got omni-channel today, wait till tomorrow, it’s going to be omni-squared.

Bryson Koehler – 00:29:16: And is that because you see these kind of chat-type interfaces now springing up, and they’re going to be running off of different data sets? And that’s where you see that happening? Or what am I missing?

Jeff Jonas – 00:29:27: There’s so many new products that you can plug in and add into your ecosystem that create experiences for the customer. And each of those, create are curating their own new pile of puzzle pieces. Now you got light magenta versus dark magenta puzzle pieces. And this really comes back to when you can match and link your data and you resolve it, you got to make sure onboarding new data sets is easy. Because if every new data set you have to onboard is hard, then it’s going to become very expensive to live in this world where there’s lots of piles of puzzle pieces.

Bryson Koehler – 00:30:02: I think your point, Jeff, about we have to commoditize the working with data. This has to be a solved problem. And it’s great to see companies like Revinate and companies like Senzing who are really doing the hard work — your investment of $50 million. We just spent two years rebuilding our CRM into a CDP with an underlying knowledge graph. These are hard problems. These are really sophisticated. But the modern technology allows us to, in your case, have an SDK that you embed. How would you suggest people find out more about guys do and how you’re helping us improve decision intelligence across many different industries?

Jeff Jonas – 00:30:45: We’re Senzing, S-E-N-Z-I-N-G, senzing.com or Google Entity Resolution will be somewhere on that first page. Our purpose in the world is to make it easy for developers. And we’re the first to reduce this very complicated problem down to a single library. It’s as simple as this, almost as simple as this. No one’s got engineers that are still building spellcheck and grammar checkers. Like, do you have a team of people working on spelling grammar checkers? No. You’d plug in a library. That’s what we’ve taken over six generations of code and $50 million of investment. And my average engineer has been helping me on this for 20 years. And we’ve made it easy for people, small little projects on laptops to tens of billions of records for interesting government problems, voter registration and banks and healthcare, hospitality, bad guy hunting, customer 360. We’re trying to make that easy for everybody. And we’re having fun doing it.

Bryson Koehler – 00:31:34: That’s awesome. Well, thank you so much. I really appreciate your time today. Great conversation. And I think for our audience, continue to reach out to Revinate. If you’ve got a hospitality use case around how you bring data together, and allow us to help you on your entity resolution, profile management, and then activation across your channels. And if you’re an engineer working to solve problems, check out Jeff’s capabilities. They’re really top shelf and really will help you in all of this decision intelligence, make smarter, more confident decisions every day. Thanks for joining.

Jeff Jonas – 00:32:10: Awesome. Good chatting.

Karen Stephens Outro – 00:32:16: Thank you for joining us on this episode of Hotel Moment by Revinate. Our community of hoteliers is growing every week, and each guest we speak to is tackling industry challenges with the innovation and flexibility that our industry demands. If you enjoyed today’s episode, don’t forget to subscribe, rate, and leave a review. And if you’re listening on YouTube, please like the video and subscribe for more content. For more information, head to Revinate forward slash hotel moment podcast. Until next time, keep innovating.

About Revinate

Revinate is a direct booking platform that leads the hospitality industry in driving direct revenue and increased profitability.

Our products and our people combine to give hoteliers the superpowers they need to crush their goals. With Revinate, hoteliers shift share away from OTAs and drive tangible results across an individual property or a portfolio. Our industry-leading, AI-powered, customer data platform collects, unifies and, synthesizes data giving hoteliers a foundational advantage.

Hoteliers gain critical intelligence – guest lifetime spend, stay preferences, ancillary revenue, and more. With our Rich Guest Profiles database, hoteliers don't need to guess who their most profitable guests are, or how to drive conversions across email, voice, messaging, and digital channels.

Revinate's direct booking platform and omnichannel communication technology powers 900+ million Rich Guest Profiles across 12,500+ hotels to drive over $17 billion in direct revenue.

View source