a16z Podcast - a16z Podcast: Making the Most of the Data That Matters

Episode Date: January 7, 2016

Every organization these days is clear about the need to get its data act together. But that doesn’t mean the path toward data bliss is clear. Data has gravity. It resides in different places at different organizations -- on premise, in the cloud, and flowing from external sources. And the rate of change within organizations is always different. So an approach towards handling data that works for one company may be the exact wrong thing for yours. Steven Sinofsky leads a conversation with three founders -- Prat Moghe, from Cazena; Gaurav Dhillon from SnapLogic, and Roman Stanek from GoodData – about the opportunity and variety of ways forward for companies looking to make the most of the data that matters. The views expressed here are those of the individual AH Capital Management, L.L.C. (“a16z”) personnel quoted and are not the views of a16z or its affiliates. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by a16z. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation. This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Andreessen Horowitz (excluding investments and certain publicly traded cryptocurrencies/ digital assets for which the issuer has not provided permission for a16z to disclose publicly) is available at https://a16z.com/investments/. Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Past performance is not indicative of future results. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://a16z.com/disclosures for additional important information.

The content here is for informational purposes only, should not be taken as legal business tax or investment advice or be used to evaluate any investment or security and is not directed at any investors or potential investors in any A16Z fund. For more details, please see A16Z.com slash disclosures. Welcome to the A16Z podcast. I'm Michael Copeland. Every organization these days is clear about the need to get its data act together. But that doesn't mean the path toward data bliss is clear. Data has gravity. It resides in different places at different organizations, on-premise, in the cloud, and flowing from external sources. And the rate of change within organizations is always different. So an approach towards handling data that works for one company

may be the exact wrong thing for yours. Stephen Sinovsky leads a conversation with three founders, Prat Moghay from Kizina, Gora of Dillon from SnapLogic, and Romand Stonic from good data about the opportunity and variety of ways forward for companies looking to make the most of the data that matters. Stephen Sinovsky kicks things off. This is going to be super fascinating because behind the scenes, all of you share a very similar set of problems and challenges and opportunities when it comes to dealing with data. what differentiates you from your competitor is how do you get the data and what do you do with that data and what decisions do you make based on that data and it's a world that's just

being completely inverted from what we used to think of data used to be the province of a very small number of people who would generate reports print them out move them up the chain and distill them down and as bened they talked about in PowerPoint slides and now we have the opportunity if you build out the right infrastructure to access that data and analyze it, look at it, and make choices all from a mobile device, all using the cloud. And so that is the centerpiece of this section. And what's really interesting is that we represented in you are CIOs and CMOs. And so we have a sort of a supplier-consumer relationship that we want to navigate.

The desire for ubiquitous access, the needs of security. the challenge of on-prem, cloud, hybrid cloud, private cloud, and then just the desire for faster, more, and better. And so to explore this topic, I'm super excited to bring up three great executives and founders of portfolio companies that will introduce themselves as they join us here. Hi, I'm Pratt Moggi, founder and CEO of Kizena. I'm Gorof Dillon. I'm founder and CEO of SnapLogic. I've been in the data business for 20 plus years.

formerly co-founded Informatica and as chief executive built that up into a decent-sized public company. SnapLogic is version 2.0 of my journey, and I'm pleased to be here, Steve. And Erman Stanek, founder, CEO of Good Data. So I'm trying to decide where to start with this. And I think I just want to start with a big question that I do want each of us to look at a little bit, which is just demystifying the big part of data. and help people to understand how, as building out a new company in the new sort of cloud, SaaS, you know, mobile world,

what is big about big data? If I could take that. Yeah, yeah, sure. Yeah, so it's interesting. You know, on one hand, I tend to cringe every time I hear the word big data. But on the other hand, you know, if you just look at the world, you know, I was before Kazina, I ran the product line at Netiza, which was a big data warehouse appliance.

And if I remember our customers back then, anybody who had three or four petabytes of data was considered a huge customer. Now when I talk to people, they talk about mobile data, social data, existing data, and so petabytes is no longer like being up there. On the other hand, we talk to many customers where it's not about volume.

It's about having existing data, but just being able to get it together, to analyze it faster. So a lot of it is about agility of data. I sort of define big data as it's a mindset. It's about being really fast about using data to make decisions. So it's not just about parabytes of data. It's about how fast can you leverage data to create business outcomes.

And so that mindset is what is different now. So just to help with a little context, so for Kazina in particular, where is Casina on the stack of this problem? What is Casina? Yeah, but not too much of a problem. a pitch. No, no, it's not a pitch, but really. Yeah. So what... That's always a founder challenge. You know, you ask for the pitch and, you know, right away, you get it. Right. Yeah, we solve world's hunger problem. Now, you know,

what Casina is essentially is trying to move big enterprises to leverage the cloud for their big data processing. And so what we're seeing is large enterprises, CIOs, CMOs, they're at this crossroad where the stack is transforming and clouds coming along. So there's an opportunity for a new platform, but they're all wrestling. with figuring out how to use the cloud, how to make it easy, and that's what we address. Sure. So taking a slightly different perspective,

you're coming at it from above. In some ways. Yeah, like what's big about your big data? Well, you know, it's bigger, no. But actually, you know, the sort of breakthrough for me on all this is, if you think about the data warehousing industry, which is about a $10 million industry, some local successes here in Europe,

business subjects, phenomenal success in the 90s, various other things that have come from Europe, Click, etc. But basically, if you think of the 90s, what we essentially had was the industry around data warehousing and analytics that was fundamentally about the barcode

scam. Here's a technology that was invented to help you get out of a supermarket faster. You're standing in line, you can check out faster. That begat an industry of analytics. Nielsen and IRA would count how much beer, was it more local

brands, people drink more Stella or something else, you know? And now, we're going beyond that, sort of comparing this versus that versus geography, in big data to me, the fundamental breakthrough is providing information from multiple places

and producing insights where the data finds the data. So, for example, a consumer package goods company, traditionally, if they were looking at the sale of lipsticks, would be looking at in a classical business intelligence way, price, volume, geography, et cetera.

But when you bring in a social media stream and you see the discussion around that particular product, you find that out-of-stock is like a big deal. So what the big data provided is an insight that out-of-stock, because colors of lipstick come and go, being out-of-stock was a huge issue to people, and more importantly, that it was out of stock because people were ending the life of that product

based on volume. Whereas they should be, and now they are, looking at the lifetime value of the product, particularly lipstick shades that apply to minorities or someone who may be a lower volume purchaser, but once they find the right shade, they're going to buy it for life. So to me, data finding the data is the magic of big data.

That is the promise that is being fulfilled in a predictive way that we never could do in the 90s. It's not about volume. It's the context. It's not about volume. But with good data, part of it is actually putting that in front of

your typical member of the marketing team, the sales team, the field. How does that fit into the big side of big data? Absolutely. We believe that with all the investment in Hadoop and this and Hadoop that most companies are still data bankrupt. Hadoop or data warehouse or whatever is a place where data goes to die.

And our goal is to actually change it. And we see good data as the last mile of analytics. You know, that's the last mile that connects your users, your business partners, your business networks, internal and external audiences with data in Hadoop and with data in data warehouses and so on. And it's kind of non-trivial because we all kind of know how data works and so on. But our customers are people literally in the field and people who manage stores and people who manage, you know, sandwich shops and so on.

And we need to deliver data to them in a way they can actually understand. And there is a big kind of impedance mismatch between the way the data is in data warehouse and Hadoop, which is actually a huge advantage of Hadoop that it can be stored in so many ways. But it doesn't help somebody who manages a sandwich shop to actually understand that. And so our goal is to be that kind of last mile of analytics. And the way we actually do it is that, you know, we actually let other customers, big banks, big telcos, big, you know, insurance companies and so on, to white label good data and sell it under their name. So, you know, we have about half million users,

and very few of them actually know they use good data because they see somebody else's logo, but it's okay as long as they get access to the data, which is kind of the biggest problem today. So what I find fascinating about trying to navigate this space is that in most corporations, finding the answer to any question is often incredibly difficult. And yet, I want to know, like, anything. Is there a movie ticket available? How many cars are available to drive me somewhere?

Can I get a plane ticket, a hotel? Like, as a consumer, I have, like, this immense access to data. And so I think, what is it, like, how do we break down that barrier? think representing the CMOs of the audience, like they know that all the barcodes are being scanned. They know that you're using a great reporting. They know it's there, but there's some impedance mismatch. What is, like, you want to go? It's good. I think I ask a good question if they're fighting over answering it. I think, you know, you can give your perspective, but I think that I sort of take a contrarian point of view to, I think it's not about

It's not about technology, first off. And so it's not fundamentally about saying, I want to ask any question I want. Because the moment you take that approach, it then becomes, like you were saying, there's a Hadoop store, you know, can I ask any question? As opposed to sort of saying,

what are you really trying to get done? What's the business outcome? And so what we've seen, then you've looked at many big data projects, the ones that fail are ones where people have taken this approach of saying, I want to collect all the data, and then I want to figure out what questions I can. ask. I want to look for hidden patterns. As opposed to people who sort of looked at it and say,

I got a marketing problem. I don't know how to track my existing customer so that I can upsell. I want to convert an existing customer much better. I want to give a better experience. And so whenever they've approached it with a business problem and then to say, what data do I need to bring together to answer that question, you a lot better in terms of formulating a narrower scope of those projects and asking those questions. Any time it becomes like, just collect the data and figure out what it is, then you start having those issues. So it's sort of the top

down approach versus the bottom of. I have a slightly different vantage point, you know. You disagree? No, not yet. Okay, okay. But I'll wait until you say something. We have a five-pound veteran who says, you disagree first. He won, see, he said that. So, you know, the vantage point we have

is, and what I've come to believe in, again, looking at the 90s and looking at this decade and this century is that it is not that we a priori know what the problems facing us are. We don't, because there's too much going on. There's too much change, you know, all the way from world economy, recession, entrant of competitors. It's just the intensity of change is too great. And I think sort of the hangover that the data industries had is this whole data warehousing batch hangover. And the truth is, we're living in a world of streams of information

and consumers, particularly, you know, people who came into the workforce in this century, millennials, have an expectation of wanting stuff now, right? So the vantage point that we see is combining streams and, in a sense, to use an overused word, mashing things up and providing new insights, is hugely important, and that is dynamic, and it is done in an interactive way. Broadly speaking, you'll have some kind of thought, like are we trying to sell more widgets, or are we trying to kill the competition or whatever. But you don't know exactly how until you engage.

Well, let me ask you, because I think part of it is how do we get from a model of like every Friday I'm supposed to show up and find the products that aren't selling well or find, you know, where do I need to stock something? Where does exploration? How do you enable that? That's where I was actually going. I actually believe that the biggest parameter of data, not big data, small data, any data, is the rate of change of business.

You know, you don't want to be doing the same search every Friday. And in a current setup, IT is supposed to govern and curate data and business is supposed to do the exploration. And it doesn't work, you know, for the IT to really kind of be, you know, in charge for the tools and the timeframes and so on and turnarounds, they are way too long for business to actually kind of depend on it. And so business then goes to Excel and other products that you might be familiar with.

I'm going to return to that one in a minute. So that's the biggest problem. And I've been in so many meetings with a CMO and CIO where there's like a zero cognitive overlap. They have not the same kind of shared interest and so on. And so I actually believe that your example, that I is a consumer, I have so much information and so on, That's actually not a good example because the information I get is some sort of a curated by, you know, by Intuit and curated by Google and so on.

And business people don't have that kind of experience in any company. And that's why they go to Tableau and Excel and ClickTech and so on because they essentially kind of resigned on IT getting them kind of that information. And it is in Hadoop and it is in data warehouse and so on. But there is this kind of, you know, again, it's the last mile. And I'm not saying that we are kind of able to solve it, you know, generically at good data, but we are at least solving it for certain types of problems, you know, getting data to business networks, getting data over the boundaries. So there will never be one solution for everything because the biggest problem in the business,

the rate of change is way too high. But I feel like there is there's... If I could just take... No, go ahead. So you talked about this exploration idea, right? And then coming back to Romans example. So we have, we're working with, speaking about how data changes business, business models, there is a fast-growing restaurant in the U.S.

This is kind of like the next Chipotle. And the guys there, they grew up in Chipotle, which again is a pretty hot chain, but they basically decided to build ground up completely differently, right? And the way these guys are thinking about data and people driving in to, you know, these guys serve fresh salads, but they look at you and say, this is Stephen, Stephen likes eggplant, or Pratt's vegetarian, or, you know, and so their whole idea is that if I could profile people coming in at noon, as they come in, I'll basically figure out how to build the right product for you, right?

So it's fresh, but it's customized for you, but you still want to do it at scale. So there's a whole new breed of, we heard this this morning, like the full stack, you know, sort of the full stack app, like verticalized, experience is everything that matters. I think that's where it's going. I think where it's going is all that data gets surfaced. It's in a product. It's in some form where if it

can be surfaced to the right people, you get magic. So let me ask you then. Is this what you describe, like so that's not like a person doing a report. So help me to understand, is there there's some elements

of a whole new style of analysis based on machine learning, based on incorporating other data sources, where, how does this, because I think that this leap is super, super critical to understanding that, that, where, why the new tools have to be cloud-based, why they're, they do what they do. Right. So, so it's, it's not about what's not selling on Friday, but why. It's more about what could we do based on the data that we have that would help us

be more successful without doing a traditional business intelligence, and it doesn't matter where they do it on-premise, in Excel, in the cloud. The traditional rear-view, mirror view of business intelligence has some element of return, but it's also at some point being well done. There are ways to improve that, but we're getting to a point of diminishing marginal returns on that. Where the returns are is the wealthy people in technology are doing predictive analytics and trying to figure out what their data is telling them.

And in that, they're using machine learning algorithms and certain kinds of open-source algorithms many of which come from Berkeley's Amplab, and they do categorization, next corner, graph. Just for a University of California, Berkeley, has a whole variety of some of the leading technologies that, like, Spark has come out of there. You too can download them and use them, right? But you need the people behind it.

So what we have now is a new population of user who is using the data from a Hadoop or something, and that person is a data scientist. This is now widespread outside of financial services. You know, Goldman Sachs, Morgan Sachs, always had quant jocks. Now, everybody has quant jocks. And how can we obliterate the barrier of enabling that person with the information in near real time to do better job of predicting their company's future?

I think it has shifted. The battle has shifted to that. So. I just agree. Okay. Well, now I can't even talk. I don't believe it. It's about data scientists.

I absolutely believe in what you said. This is kind of a, you know, machine learning and so on. unfortunately, some, you know, most companies are not big enough to have big enough sample for machine learning. You know, big banks, what makes Google, Google, what makes Amazon, Amazon, is that the data sample is so big that you can actually really learn from it. Typical company would look at their, you know, thousand invoices, and there is nothing to learn from.

So I actually believe that that's why analytics is done in the cloud as actually a lot of value, because we see data across tens of thousands of companies, and we can actually do machine learning from, you know, massive data sets that in the cloud. individually don't actually mean anything. But Roman, that's not analytic. That's reports. No, no, no. It's not predictive analytics.

All right, well, we're going to take it out back. So let me, let me, let me, let me, let me, let me, let me, um, let me turn it around and, and ask it, ask it a little bit different way. Because I think, so there were two people mentioned Excel, so that was, that was my, that was my, that was my cue to talk about Excel. Yeah. So, so my, what, what I think is so interesting is that if I were to, to query the room, huh, a pun, and we would find out that most people, that most people, that most people. find Excel the most valuable analytical tool that they're using and and there are maybe two reasons let's just touch on you know one of them is just that it's the one that they can use that does what they want but another one is that the

the gap between the CMO and and the IT organization is often there's data missing and that there is some source like it could be geographic data it could be like wow this report doesn't even list all the stores we have or all the outlets or it doesn't have our web logs or there's just another part of the data that isn't yet in Hadoop, isn't in some lake. And so so much of the job is just bringing together and then applying that knowledge

because I do think that that's what differentiates. You know, like the difference between Minneapolis and Bentonville and the U.S. is not necessarily the product they sell. But I started with this one, and it's going to be short. I always believe that there are two types of people in the world, people who can use Excel Paveau tables

and people who cannot use Excel Paveau tables. you know, and we all was belonging in one of the categories. And believe me, the people who can... We worked really hard to make pivot tables, you know, easy to use. I know, I know, I know. It's not about a tool. Actually, ironically, later in the afternoon...

Later in the afternoon, we're going to hear from the actual person who made them easy to use in 1989. And so your example actually assumes that people actually know how to use Excel table tables, and that's why some of the most frequently used kind of data analytics tools

are extremely basic, because they actually look and feel like sheet of paper like two-dimensional sheet of paper and you know so that's that's the problem with analytics that on one hand we have you know very complex systems with you know Spark and Hadoop and so on and yet you know most of the people actually using those tools you know don't like the abstractions they like sheet of paper rows and columns and and so it's very difficult to bridge that you know that's that's why you know you actually see a lot of kind of analytics that's successful, it's either embedding.

So it actually already is kind of pretty vertical and industry specific and so on, or people are using, you know, I believe that all of us compete with Excel and bodies. Somebody looks at it, put it Excel, and sends it over email.

We don't. For the record, we don't. Microsoft's a good partner and investor in our company. We don't. I have a very different point of view on this, which is just a show of hands in this room. don't feel afraid. How many of you use Tableau in your shop

are just two hands? I find that hard to believe. How about Excel? Yeah. So my view is that tools are really hard to change because tools usually embody a business process. They embody... So my view is that large companies,

particularly the ones where they're analytically driven, I think the way this is... for them to really leverage fast-changing data, the fast-changing world, you've got to sort of figure out how it gets there, like you were asking this question, right? So the legacy data flow is probably going to stay on-premise for a while. Now the question becomes,

how do you leverage these new technologies, how do you leverage the cloud? And so it's going to be an augmentation strategy where there is this concept coming up, which is called the pipeline, right? And the pipeline idea is that data, like a river, it flows. And so maybe some part of this data,

whether it's external or internal, will flow into the cloud. You know, certain kinds of processing will happen. And then over time, what will happen is you'll take that data and then maybe land it in certain place where data scientists can analyze it. Some data will continue to go to Excel. Some data will continue to go

to Tableau or the BI analysts. So it's not going to be a world where you know, everything just disrupts right? Overnight. People who do things will continue to do it. Sure. But, Well, let me, yeah, so let me, let me ask. Go ahead, go ahead.

Every tool, but the point is, every technology is good for doing something. It doesn't subsume. Spark doesn't subsume data warenousing. Hadoop doesn't subsume, you know, streaming. So they're just like different technologies for different jobs. So speaking of subsuming, which is a great way to ask this, because I do want to recognize that the CIOs in the room are dealing with a very, you know, real challenge and real opportunity, which is, I'm guessing, for most all the people in this room,

their system of record is an on-prem, structured, SQL, Oracle-based system. And for all the CMOs, that's the starting point for most of the data that they need to get to. How do, what message, how do we help the customers in the room bridge that reality that they deal with? Or said another way, like, where is the opportunity, how do they start a new project, what do they do in order to, So, maybe you think...

So, you know, we get asked this all the time. Sometimes I've been in there. Those people use my former products and so on. So look, here's what we recommend. First of all, I believe the CIO and CMO have kissed and made up in a big way. In a big way. Well, they're all here. Here they are.

Together, sometimes on the same table. Nobody's hit somebody on the head with anything yet. But beyond that, what you have is this concept of a pipeline. We used to call it an information factory. in the 90s. It's a pipeline now because it has real-time streaming attributes. But people should be thinking about being able to use the new price performance of Hadoop Spark to obliterate their traditional data warehousing appliance. Not unplug it, but the rising tide of the data lake, we think, will drown out the data warehouse in the fullness of time.

So that is, like, that's an interesting exponential change. That's probably a huge opportunity that might be worth thinking about for a second, which is even if you have a petabut. in your structured stores today, if you turn on the right sources within the company, you'll very quickly have lots more than that to potentially work with that might end up being even more valuable. Indeed.

And, you know, I actually want to answer your original question, like how do we actually, you know, how do we deal with the fact that most data is on premise anyways? You know, cloud-based company, I need to grow my business. So our focus is actually on data that goes across the firewall anyway, you know. So we really kind of help companies to monetize the data and deliver analytics to the business networks.

Our biggest customer is one of the large credit card issuers, and they have merchants, they have, you know, issuing banks, they have acquires. So they have, most of their audience for data actually sits outside of the firewall. So instead of emailing data in CSV files or excels and so on, we actually help them to really kind of build kind of analytics distribution platform.

And I believe that that's, you know, we don't have time to wait for companies to move the data, the primary data to the cloud. That may not even happen. But more and more kind of data sets are being used in these kind of across the firewall, in mobile scenarios and so on. And that's where we see kind of 90% of our focus.

Is it really the case that, I mean, I think one of the things that's so interesting is that in general there's just more data outside of your organization than there is inside. A key part of the tools is just how you connect those? I think there's a few things going on.

There's definitely a shift towards the cloud. It all depends on the verticals. Some verticals are I'd say more skeptical. There's regulations that basically say certain data cannot physically leave, but in most companies, even with financial services, we've noticed that

they're very eager to explore the cloud, either for external data or to even look at all your internal data, and not all data is equal some data is PII some data is not and and there are newer technologies that allow you to encrypt data in motion at rest the clouds matured this very sophisticated security control so we're seeing you know it's not as religious as it used to be before and that's one of the reasons why you're seeing CMOs and CIOs come together because there are platforms that where the CIOs can

basically now stand-up projects to move data provision things in AWS, on Redshift, Azure, and run these projects so that they feel like they're part of a private cloud, right? While they're really running on public cloud. And so there's a... It's changing very quickly. It's changing. It's a dynamic. Well, it's always one of the challenges in these panels.

We want to address the breadth of needs and realize that people are always going to be at a different place in making this change. At the same time, you know, we are in Europe, and what makes this kind of, you know, slow down is this kind of local. regulations and so on. Instead of putting one data center for, you know, my all audience, you know, user base, we need to build multiple data centers and that kind of fragmentation and balkanization of data will continue. And that's going to be more and more difficult for cloud vendors to manage that and manage, understand all the regulations and so on. So I think, you know, look, the fact is data has gravity and you have to respect that. If your data's on premise, you should probably put your Hadoop or other kinds of analytics on premise.

If your data's in the cloud, you're using a website hosted in Amazon or a Deutsche Telecom or Swisscom in Switzerland, then obviously it makes more sense to have analytics located there. But I think in all cases, what is blindingly clear from looking at the various people we talk to is that you have to have a predictive analytics capability. However you do it, on-premise or in the cloud, you're dead without it. One, two, certain things make sense to migrate to that and certain things not. You know, if you're a credit card issuer, conflict resolution, how does a customer return a product,

there's no money in trying to put that into a modern system. You've somehow got it going, you worked out the disagreements, leave it where it is. But trying to understand how a social media stream interacts with their product, how people are complaining to Twitter rather than to your call center, is of great importance to you, right? So we think that that is probably what people should do is get a predictive analytic strategy in place

and start to bring, based on data gravity, the right sort of technologies to solve that issue. Yeah, feel free. I think you've talked about technologies and tools. The other thing we've noticed, it's all about people. So the shift, it's a transformation, and transformation always begins with leaders. And we've noticed that it's CIOs, CMOs, forward-thinking guys.

Sometimes it's CTOs that help push that. Sometimes it's CEDOs, chief data officers or chief digital officers. There's one great guy in the audience, Osama fired from Barclays. I don't know where you are sitting, Osama, but he's had those experiences before, and so when he's in a financial services company, he's thinking about how to bring those experiences to.

So I think it's all about getting those people and then bridging basically the gap. Awesome. Well, thanks everybody for a lively discussion and appreciate the insights on data. Thank you very much, everybody. Thank you.