← The Rundown
The Rundown
The Slow Burn of Tariffs & AI's Infiltration of Corporate America (ft. Ara Kharazian)
00:35:38
Loading summary
Transcript63 lines
- [00:00]
A
Welcome back to the Rundown, one of the top business podcasts in the world. Today we are talking to Ara Khorazian, an economist at Ramp. Ramp is a fintech company that provides businesses with corporate cards. And it's Ara's job to analyze the spending of thousands of businesses to generate some interesting observations. So in today's conversation, we discussed what the data is saying about money spent on AI, what it says about the AI bubble, how tariffs are impacting businesses, and, and why Ara is still a big fan of government data. This was such a fun and insightful conversation. I think you guys are going to really enjoy it. So let's get into it. All right, guys, Today we are talking to Ara Kharazian, an economist at Ramp. Ara, thanks so much for hopping on the show today.
- [00:52]
B
Thank you, Z, for having me. I'm a big fan of the show.
- [00:55]
A
I appreciate it, man. For people that aren't familiar with Ramp, can you just give the listeners a quick update on what Ramp does and your role as an economist at Ramp?
- [01:05]
B
Sure. So Ramp is a financial operations platform. It's used by businesses everywhere from like a startup to a sort of massive retail company or a big tech company. And they use it to pay their bills, manage credit cards, process their expenses, and automate a lot of the financial steps. I say all of that because that's the data set. I then have access to essentially anything that a business spends money on outside of pay. Payroll is what I work with. And my job is to do research on that data set, find out where businesses are spending, where they're pulling back, where they're investing, and ideally get a forward looking sense of where business is going and where the economy is headed.
- [01:43]
A
So that's really cool. So essentially you're analyzing where like thousands of businesses are spending their money and coming up with insights based on all that data.
- [01:52]
B
We have a Data set of 45,000 businesses across the US and it's pretty diverse. So we got everything from tech companies buying AI to construction firms sort of seeing the impact of tariffs and manufacturers. And so it's a really diverse sense of look at where business are spending, especially in a time where governments shut down. We can't use a lot of government indicators to figure out where the economy is headed. Private data sets are going to become increasingly important and there's generally not a lot of focus on business data sets anyway. A lot of the economic indicators we tend to follow in the economics world focus on consumer spend, things like inflation or anything like unemployment. Those are all important data sets. But business spend can often be a leading indicator of where the economy is headed and that data is just not available anywhere else.
- [02:42]
A
Well, you said a lot of the magic words that we're going to talk about today. AI tariffs, government data. Let's start with the AI question first, because that's what everyone wants to talk about. You wrote a great blog post recently asking if AI is a bubble and everyone wants to know if that's the case. And you tried to answer the question using data and not just, not just vibes. What did you find out from your research? And I guess, simply put, are we in an AI bubble right now?
- [03:08]
B
Well, you frame the question very well. Right. Everyone's talking about whether or not AI is a bubble. And for the most part those conversations focus on whether or not the companies building it are investing too much from sort of a capital expenditure standpoint. Are we building too many data centers? Are we building, Are we, are we buying up too much stock in large tech companies developing AI? And all those questions hinge on whether or not AI is going to be a helpful, productive investment for companies to then buy and work with in their actual operations. Right. The concern is that, oh, we're sort of over investing in this software and this technology that no one's actually using. The reason why I think those concerns exist is because there's not really a lot of data about whether or not companies are using AI. That data set is just not available. So what I wanted to do with my blog post was to provide some data on the market about what companies are actually buying in terms of AI, if they're investing in this and if the products are getting better and more useful for those companies. We're still very early in this AI ecosystem. And so I don't think anyone's expecting that it's going to be extremely transformative the next first couple of years, but we should at least have some sense of whether or not this is starting to generate some returns for the companies buying it and implementing it across the enterprise. And that is generally what we found. A couple of things I would say is AI products are getting stickier. So retention of AI products is now about 80%, up from about like 50% 2022. So what I mean by that is that in 2022, but this is pre chatgpt, you know, about a year in to using an AI product, only about 50% of companies were still subscribed to that product. That's not way higher, it's about 80%. So what we find is that if you are A company, you're buying an AI product or service, you are seeing benefits of it throughout the year such that you are extending your contract, keeping it on your books and continuing to maybe resubscribe to it the following year. This is second thing we're finding is that AI products themselves, contracts themselves are growing in size. About 20, 23, the average contract for an AI product if a company was buying was maybe about like 30, $40,000. That's generally what you'd hear from founders who are also selling their products to two companies. That's now changed completely. We're now seeing average product size of about $500,000 a year. And we estimate that the average AI contract will hit $1 million next year. So the contract sizes themselves are growing and that tells us that the investments in AIR are not just seen as small pilots. These are large scale investments by organizations that are seeing their impact and want to integrate AI throughout their larger organization.
- [05:55]
A
Okay, got it. So people, so companies are buying more AI, they're sticking with their AI softwares that they end up signing up for at a higher rate than they were previously. So, so what that indicates is that all this investment that these AI companies are making is leading to meaningful adoption of these AI tools.
- [06:17]
B
That, and I would say some of the biggest contracts we see aren't necessarily for the large model companies. It's not just OpenAI and Anthropic. This product ecosystem has matured a lot in just the last year, such that the largest contracts we often see are for things like companies building a sort of dedicated AI customer agent software. That's one of the main places we've seen AI integrated into enterprise is companies using AI to automate the tasks of many customer service agents and allow their sort of human customer service agents to focus on tasks that only humans can do. Therefore, so we are seeing these productivity gains. We are seeing companies using AI in ways that is both helping consumers and then also saving them some money. But that data has not previously been available or easy to find. Yeah.
- [07:09]
A
Which I think is very interesting here is like you're actually looking at real spend and it's not just like I said, it's not based on vibes where everyone's like, well I don't know, I feel like, no, I feel like people aren't using chat GPT as much as they used to. But you're seeing that like companies specifically are spending money on AI. Going back to the question about the spend on OpenAI. Anthropic, is OpenAI still dominating right now though? If you Compare it to Anthropic. And like even Gemini, are they still the top dog?
- [07:37]
B
OpenAI is by far at more companies than any other AI model company. They lead the adoption rate. So more companies subscribe to OpenAI than subscribe to any other AI model service. I think there are about 30 or 40% of companies in our data set right now. Anthropic is a much smaller share, about 10 to 20% of companies, depending on the sector. But it does have stronger market share in certain areas. So Anthropic is very competitive with tech companies, for example, and the tech companies that use it tend to use it for very large investments. API spend specifically. But OpenAI is definitely the leader in business adoption.
- [08:18]
A
Gotcha. I guess that's not really surprising because they were the first ones. They kind of have the name brand. So if you're a company and you're selling CTO approaches, you like, oh, let's go with the OpenAI seems like the safe option to start off with it.
- [08:30]
B
Is in many ways starting to sound more and more like its lead in the consumer space is also driving gains in the business space.
- [08:40]
A
But how does that compare to, like, these hyperscalers? I'm thinking like Microsoft here, right? I know I live in Houston. A lot of my friends work in, you know, just normal industries, oil and gas, construction, things like that. And a lot of them are hooked up to Microsoft services. I think Microsoft offers like Copilot and things like that. Are you taking that into account? Are you able to tell from the data that a business has, like a Copilot add on that that counts? Does that count as AI spend or can you not? Do you not have granular detail like that?
- [09:11]
B
That's a good question. So, and it's one of the things that's been very hard to measure is that a lot of the development in this market, particularly from the larger companies like Google and Microsoft, their product strategy has been to offer AI services for free. So Google started offering Gemini for free in all workspace plans. It was around that time that we started seeing Gemini adoption fall in our dataset. Because we're only looking at paid adoption. Now, our instinct is, I mean, I think we can reasonably assume that people are using Gemini because it's integrated for free in so many of Google's tools. Same goes for Microsoft's tooling. And we're able to see some of that because we do see line item spend on certain kinds of plans. Like, that's one of the things we get from data, is that we get to see everything on a receipt. But I still do think it's the case that companies that are investing in large scale investments, particularly throughout a large organization, they're not just looking at chatbots and they're not just looking at the kinds of tools that are integrated directly in Microsoft Office Suite or Google Workspace. One of the interesting things I think about this market is that there is so much attention placed on the large model of companies. And I think some of that makes sense. But it then tends to overlook a lot of the investments happening in enterprise scale tools. They are very popular but don't get covered a lot in the news. Again, some of the largest contracts that we see aren't for the model companies. They are for observability platforms so that companies can monitor the performance of their AIs. They are for trust platforms so that companies can make sure that their AIs are actually working as intended. And they're for customer service platforms. They're for coding tools that work on the model companies technology but do something a little bit different, geared towards software engineers.
- [11:07]
A
Yeah, I feel like a lot of these coding tools, aren't they just using the latest GPT model or CLAUDE or whatever in the back end to like power their services?
- [11:18]
B
They're often built on them, but they are making vertical specific investments that make these products work in the enterprise. I mean, one of the main holdups to adoption that we've heard from large businesses, from enterprise, for example, is that they don't want to integrate AI fully across the organization because they can't trust if it will work effectively. They can't trust the output. They can't trust if they'll be right. They can't trust that a chatbot will say something weird to a customer talking to it. And products to monitor that kind of behavior are being built in the AI product ecosystem and then they are being bought by customers. But again, it's happening outside of the main model companies. Now those developments are still happening. The main model companies, they're doing their evaluations, they're trying to improve these models for enterprise. But a lot of the platform investment that's happening is happening outside of the sort of big three.
- [12:16]
A
Just zooming out a bit though. You guys have businesses in all kinds of sectors outside of tech. Are you seeing a higher adoption rate in certain sectors compared to others? I'm just kind of curious to see that kind of breakdown.
- [12:28]
B
So tech and finance lead, but the fastest growth is coming from two sectors I think are actually unexpected. Healthcare and then construction, manufacturing. So we are seeing that these Kinds of sectors that are typically seen as late adopters to new technology are adopting AI faster now than we expected. Now I want to caveat that and say that, you know, the investments being made by construction manufacturing, they're not necessarily being made in sort of robotics technology en masse. You know, we're seeing AI being used for planning purposes, back office automation, certainly in the design function in the healthcare space. AI is being integrated more so throughout. You know, heard many stories of doctors using AI to automate their note taking. There's some productivity benefits to that, but we're still a long ways away from these sort of AI automation necessary to really launch us into the next era of productivity with things like construction manufacturing, that we're going to need some investment to robotics first.
- [13:35]
A
Okay, so we're still early there, but it's good to know that like you wouldn't think that health care and construction would be investing in AI right now, but the fact that they are is.
- [13:44]
B
A good insight appetite for more investment from these industries. And there is this increasing attitude from the operators in these industries that they need to be on this if they want to be competitive. And that's not always been the case with new technologies that are coming to these sectors.
- [14:04]
A
Yeah, absolutely. That's good to know. Have you seen, with all this increase in AI spending, have you seen this come at the expense of companies cutting back on other categories? Is it actually leading to productivity gains and potential increase in margins or have you not seen that yet?
- [14:23]
B
So there is some early research in this area. It's a little too soon to say, particularly it's labor market impact. And so results are mixed there. So I'll, I'll let the research just speculate.
- [14:39]
A
Go ahead, let it fly.
- [14:41]
B
But I do think that at least in the specific places that it has been adopted, in the highest numbers in customer service and software engineering, there are clear productivity benefits. I mean, we can say, we talk to CEOs particularly in the tech sector, who say that there has been this decoupling from company growth to headcount growth. So if in the past to grow a tech company, it was expected that you grow your headcount. We are seeing companies able to grow their revenues without dramatically increasing their headcount, specifically at tech companies. So again, anecdotal evidence, but that is the sort of attitude We've heard from CEOs and CFOs recently.
- [15:24]
A
It kind of sort of checks out because I think one of the concerns that a lot of people have is with the further adoption of AI, it's going to lead to a cutback in hiring people, specifically entry level people. And I was just curious to see if there was anything concrete in the data. You're saying it's still early, but anecdotally it seems like CEOs are kind of saying that, yeah, like we don't have to grow headcount in order to grow revenue.
- [15:50]
B
Well, but that doesn't necessarily mean that we're going to see job losses again, because I, I am of the school of thought that AI is going to follow the similar path of every other technology that is automated human work, in that we will find other things to do and invest in. And I don't think that AI is going to be able to automate all sorts of human work in many ways. I think the cost savings to companies will also lead to more investment, which I do think will create more jobs and particularly more innovation in places that AI just cannot perform yet. And that's the path that we saw with the Internet, that's the path that we saw with cars, that's the path that we saw with all of the early technologies of our time, is that we seem to keep finding other stuff to do. And I'm not a doomer on this stuff.
- [16:40]
A
I will say I'm totally with you as well. I mean, we saw it with literally every piece. I mean, just take spreadsheets, for example, right? Like spreadsheets came in and like we have more work than ever before with any piece of software. It's been that way. Let's hope that it stays that way for AI. I choose to be optimistic as well. Shifting gears a little bit though, from AI to something that is more concrete. Well, maybe not tariffs. You wrote a great blog post about the economic impact of tariffs and why it's so confusing to track. And I found that to be interesting. Can you talk more about what you found when trying to analyze the economic impact of tariffs?
- [17:18]
B
Well, tariffs, I think are so, from an economist standpoint, required a lot of us to adjust our priors a bit. I mean, when tariffs were announced, the market tanked and not for no reason. The tariffs that were announced were significant and large and the largest that we've had in over a century or so. And so you would imagine that would cause a lot of turmoil in markets, particularly markets that had just come out of supply chain disruptions from the pandemic era. And we're still struggling to. And we're still, we're still sort of working toward, toward resolving all of the supply chain issues. We're still dealing with some bout of inflation as well. And then it's now been six months, more than six months since Liberation Day, quote, unquote. And we haven't seen that much of a price impact like here and there we have, but broadly we haven't seen the amount of inflation that we were worried about. Economy has not fallen off a cliff. Most of our fears of a recession have subsided. And so people are showing. So we're looking back at the data and trying to figure, out, okay, why it's not that tariffs haven't had any impact. A couple of things are going on. One, there's been a lot of renegotiations. The actual tariff as implemented have been much lower than what was initially announced. Second, there are impacts, they just tend to be pretty localized to certain sectors. So manufacturing, retail. And then third, tariffs haven't actually been fully enforced. You can find this in our data set, but also in public data sets about tariff revenues that the actual, you know, the actual sort of share of transactions showing tariffs has doubled. We're seeing only about like 1.4% of business spend from manufacturing and retail show tariffs. Now it's about 3%, but it's much slower and more gradual than you'd expect. It's not like, oh, the tariffs were announced and then immediately they started getting collected at the ports. We're seeing similar stories from businesses who are paying those tariffs are that there are a lot of frictions in actually implementing these kinds of policies. First, we need the enforcement to actually figure out, wait, what actually gets tariffs and how much. I think even people follow tariffs very closely. Could not tell you off the top of their head what the current China tariff rate is. And that's not because they're not doing a good job. It's just because it is genuinely really hard to follow even if you're. Yeah, and I'm a fairly well informed person and I'm praying to God you don't ask me that what the current average tariff is. And second, there are people involved in this process, right? There's someone at the port who is assessing tariffs and inspecting boxes. Again, there are impacts and there are frictions. And we have talked to business who have told us that tariffs are having impact on their business, that they are having to switch suppliers and having delayed shipments affect their operations. All of that is happening, but it's also then dampening the originally expected impact of tariffs themselves. We thought they were going to be applied much more dramatically and then for a few reasons, mainly because of the renegotiations. And then secondarily, because it's just taking a long time for them to actually be enforced. For that reason, we are not seeing the price impacts that we would otherwise expect.
- [20:54]
A
So do you think that the slow rollout is impacting obviously the price impact, but are companies just eating the tariffs for now until they kind of figure out where everything settles out?
- [21:07]
B
We've heard a couple things from businesses on our platform. We have heard that they are. There are many businesses that are changing their suppliers. They're not necessarily moving manufacturing to the US in many cases they can't. Right. Like you could imagine, we talked to a business that produces a cosmetic that uses coconut oil. And he. We don't really grow coconuts in most places in the United States. So he has to figure out some other suppliers they are maybe moving. Many of them are moving supply chains to countries that are relatively low tariff. Right. The tariffs that were announced were not across the board, the same across every country. So they're looking at that. And then second, I think they're also waiting to see where policy does end up. You know, it's only been six months. This rollout has been slow, so it's been laggy in that way. And that is at least providing some time for businesses to figure out how they want to change their operations. And also, you know, think about price increases.
- [22:03]
A
According to your data, as Of September of 2025, 3% of bills and invoices are showing a tariff charge. Where do you see that ends up at in 2026?
- [22:16]
B
It's hard to say. Look again, it's going to depend a lot on the renegotiations that happen. And just showing a tariff charge doesn't fully capture the impact of tariffs themselves. Tariff charge does not actually capture the actual size of the tariff. But what is notable about the tariff policy as implemented is that it is applied across pretty much every country that we trade with. That's not generally been the case. Most tariffs have been applied either on a very small set of countries or on a very small set of imports. I mean, if we follow the trend, we're going to see it stay in the low single digits maybe, I mean, or potentially hit the. I mean, low double digits like 10, 11. That would be on the high end of what I'd expect. But it is really hard to say given how laggy the policy is.
- [23:03]
A
Got it. Yeah, it's. I think everyone's trying to figure out what's going to happen. So the next question I wanted to talk to, shifting gears a little bit from tariffs, is government data. I Think the people are starting to lose confidence in like the data that we're getting from the government. A lot of scrutiny around that, especially when Trump fired the head of the bls. Where do you see? I'm trying to rephrase this. Are you concerned about the reliability of the government data, especially with all the revisions that we're having? And are you concerned about the way that government data is collected?
- [23:44]
B
Here's what I'll say. There are reasonable concerns about any statistical measurement. And we should always try to make sure that we are measuring things in the way we intend. We should always try to make sure that our methodologies aren't drifting away from reality. I do think there has been a lot more criticism of government data than is fair. And I will say that as much of it is coming from people in my role, right? Like I work in this. Actually less so from people in my role, but people sort of from the private sector who see the strength of private sector data sets and think, well, why do we need the government data sets? Private sector data sets will never replace government data sets. And in the work that I do, government data sets are both important as a benchmarking tool and in helping me triangulate my results. I could not do my work without strong government data sets.
- [24:42]
A
Can you talk more about that though? I'm really curious because from what I've read, there seems to be like a bit of a flaw when it comes to the government collecting data in the first place. Because they're doing surveys, right? They're doing phone surveys, they're calling people, they're sending out letters in the mail to try to get information. Whereas you guys like a ramp, you have like real time data on, like where businesses are spending money. Why do you say that government data is so critical when it, like the way they collect it is almost flawed?
- [25:12]
B
Well, in some ways it sounds flawed, but in some ways it's actually. Look, I don't think surveys are perfect. In many ways, surveys are less helpful data sets than what we are able to produce in the government in the private sector. But the government is able to collect data on a broad swath of Americans, particularly Americans that are very hard to reach, and particularly Americans that are not often served by the private sector. If you really want to build a data set that is comprehensive and measures every part of the American economy, you do need government investment and you need a government operation that does that. Now you can work with private sector sources and the government does the government ingests. I don't think people realize this the government ingests partners with private sector operators all the time to improve and benchmark its own data sets. So the government's doing all this work that many sort of detractors from the private sector say that it's not. And I think private sector data sets are still important and in many ways they can offer a lens that the government cannot. I've tried to do this with our work on Ramp AI Index. We report on AI adoption at US Businesses using actual spend data. The US Government also has a census survey that goes out every two weeks to businesses that asks them a similar ask them a question like, do you use AI? And I have been on the record criticizing this collection method because for this specific purpose, yes, I do think we should be using spend data. And I think the way that the government wrote the survey question doesn't really make sense and doesn't really capture the breadth of how AI can be used. And I think the question is also, this is a little bit more technical, but is written in a way that most people don't really understand. It's kind of used, written in this econ speak way.
- [27:04]
A
Well, who's answering it too? Like, is the CTO answering it or is it just some person at the receptionist answering?
- [27:10]
B
One person is answering. And companies are large organizations. Like these are all valid criticisms to have. And also, the government does a better job of measuring inflation than any private sector organization would. The government does a better job measuring unemployment than any private sector organization would. And from my perspective as an economist who works on private data set, my main goal is can I produce work that is helpful and additive to the public discourse that does something that the government, for whatever reason cannot adequately measure or something that another private sector organization can't adequately measure. What is my data set particularly well suited to do now for business spend? The government has a hard time calculating, estimating business spend because it's much harder to reach businesses than it is to reach consumers. There are just way more consumers to reach out to. So if you're doing some estimate of unemployment, the best way to do it is to call a lot of people. You know, there's a lot of sort of attention on like the private sector data. So why don't we just use adp? Well, ADP is used to enterprise. Almost half of Americans are paid hourly. They don't work at places that use adp. So that is, that is a noisy data set inherently. And no, no shade to my friends at ADP that you do great work. It's very helpful work. But they would say as well, wait. We need government data sets to get a really accurate measurement of our economy, particularly parts of our economy that are not captured by the private sector. And so I'm a passionate defender of our government data sets for this reason. And I say this because I am someone who produces work and data from the private sector that we need these kinds of resources.
- [28:56]
A
That's good. I'm happy that you're defending it because, yeah, everything that I'm reading, there's always criticism around it. I think some of it is fair, but it's good to know, like, as someone who produces insights from private sector data on the importance of government data. And I think you mentioned, I think the key point to me was like, consumers. I mean, a lot of the times the government is a lot better at reaching consumers who aren't tied into, like, the private sector. So that was. That's very helpful. And we'll see what, what happens with the government data. Right now it's a shutdown. So, like, we're not getting much, but hopefully that, that clears up soon and we start getting more government data. I want to end with some lightning round questions. This is something that was really curious about. I mean, you have access to all this. So much interesting data. What's like, the weirdest thing that you've seen? Expensed, maybe. I don't know if that's the right word. What's the weirdest thing you've seen? Expensed or charges on a corporate card, like strip clubs, whatever. That's not even like that interesting anymore. But is there anything that's like, really interesting that you've seen? You're like, oh, wow, why is, why are we seeing a spike in this spend?
- [29:57]
B
It's. Look, you can find weird things in transactions here and there. And you'd be surprised. Like, businesses often have a real business purpose for expensing something like Halloween costumes or like concert tickets for a client. Right, Something like that.
- [30:12]
A
Coldplay concert tickets. Got it.
- [30:14]
B
The funnier thing that comes up is how people. And again, everything I look at is anonymized. Like, I'm not looking at anyone's actual spend. I'm not seeing anyone's name. I. I don't even see the business name really. But the funnier thing I see is that when people use ramp to do their expenses, one of the primary use cases of ramp, you write a memo to your boss about why I needed this thing. And sometimes people are unusually apologetic for why they bought something like, oh, they had to Uber from an event back home. And it was like a $200 Uber and someone is extremely apologetic. I really looked through all of my available options and this is the only thing I can do that is interesting. And you see this window into how people, into people's, both their personalities, but in terms of how they operate, this unusual interaction of hey, I'm using my corporate card to pay for something. Is it okay, question mark, question mark, question mark. So I like seeing how people try to navigate this genuinely unusual social business interaction.
- [31:18]
A
Yeah, it's like, I'm sorry that I had to spend $500 on a steak dinner. There was, there was no other option. Everything else was close.
- [31:24]
B
I would say if you're in New York, we put out this thing about like the top restaurants in New York based on where people expense. And the thing I got a lot of heat for, which I shouldn't have gotten heat for because it was just, it wasn't my opinion, it was just where people go is that one of the top restaurants in New York was the Smith. And if you know anything about the Smith in New York is that it's kind of basic, like it's not that interesting of a restaurant. It's actually quite nice. I've been to the Smith. It's like one of those places you could bring anyone and people could just get anything they want. They always have a table for you. But it's one of those restaurants as well that, well, wait, you have your corporate card, you can go anywhere. Why are you going to the Smith?
- [31:57]
A
Gotcha.
- [31:58]
B
And, and I had to defend the Smith extensively in comments when people said, why would you put the Smith on this list? And I said, well, you're just, you're.
- [32:07]
A
Just going off of the data which, you know, it's so funny you say that. One of my questions that I had was were you able to identify up and coming restaurants based on the corporate spend? Because like you can see where people are spending the most amount of money. Maybe, maybe that's a ramp like side hustle right there where they, where they become like a foodie account where they're just identifying the up and coming restaurants before it becomes popular.
- [32:27]
B
It is by far my most recognized work has been my work on restaurants. It's the posts that I've done that have been most popular have been the posts that rank restaurants in each city based on where people spend and when I go to events. And if there are tech people or people who follow me at those events, almost always, if they are to recognize me, it's because they say, I saw your Post on restaurants, and what I realized is no one cares about economics. They just. They just want to know what restaurant to go to.
- [32:56]
A
Yeah, they want to. They want to go. They want to impress a date on a nice restaurant. I'll end with this. Based on the data that you're looking at, which city balls out harder when it comes to parties, holiday parties, corporate retreats? Is it New York? San Francisco? Is it Houston? What can you tell?
- [33:12]
B
New York.
- [33:13]
A
Okay.
- [33:13]
B
Quantifiably New York. We similarly put out research about which cities spend the most on alcohol and where it's growing. It's New York. It's New York. And also it's New Orleans.
- [33:26]
A
That kind of makes sense.
- [33:28]
B
The off sites that might happen are like, trips to New Orleans. Yeah. And just their history and culture. But places like San Francisco. Most cities actually in the US Are drinking less. You see that in consumer data, but you also see that in business data. New York is one of the only cities where people are drinking more in the last couple of years. And that share just continues to grow up.
- [33:48]
A
That is awesome. Like I said, your job is so cool. We could do this every month. You could probably have amazing insights for us. Ara, I really appreciate you hopping on today, and I'm gonna probably email you later to get the hottest restaurants. I'll be in New York in a couple weeks. You know, I'm trying to hit up some good restaurants or. Yeah, let's go to the Smith, and we'll just, you know, charge it to the corporate card. I appreciate your time. For people that want to nerd out on all your research, I believe you have a substack if you want to plug some of the stuff that you're doing.
- [34:18]
B
I do. I post everything on substackeconlab.substack.com that's RAMP Economics Lab. And I'm on Twitter. You can follow me there and ramp.com data if you're interested in anything Ramp.
- [34:31]
A
Related and future foodie influencer.
- [34:34]
B
Right.
- [34:34]
A
I'm telling you, that's a good side hustle. These foodie guys make a lot of money. It's a good side hustle to have once, you know, a little outside of work activities.
- [34:42]
B
Thank you so much for having me.
- [34:43]
A
I appreciate it, man. All right, guys, hope you enjoyed that fantastic conversation with Ara Kharazian. I highly recommend you guys check out Ara's writing. He has some great economic insights. I recently subscribed to his newsletter on Substack. It's really good stuff, and hopefully we'll have them back on soon by the way, if you guys enjoyed today's episode and have like five extra seconds, consider giving us a five star rating on Apple, Spotify, YouTube, wherever you listen to your podcast. And if you are listening on Spotify or YouTube, let me know in the comments what you thought about the episode. Your favorite part, if there's any questions you'd want me to ask Ara on a future episode. Thank you guys so much for listening, watching and commenting. Shout out to Mike and Connor for all the work behind the scenes and see you guys back here on Monday.