S02 E11: The Reverse ETL Revolution: Overcoming Challenges in Syncing Live Data to SaaS Tools with Tejas Manohar, Co-founder and Co-CEO at Hightouch

Hello everyone and welcome to the modern data show and we are back with another insightful episode on all things data. Today, we are thrilled to have Tejas Manohar, co founder and co-CEO at Hightouch, which is a data activation platform that syncs data from sources to business applications and developer tools. This frees up engineering time for the data team and delivers actionable data directly to the business team, ensuring data consistency across the organization. Prior to Hightouch, Tejas worked at Segment, where he held several positions. He is a skilled engineer, manager and an entrepreneur with deep understanding of data integration and analytics. So welcome to the show Tejas.

Thank you for having me.

So Tejas, let's start with a very quick background about yourself. Tell us the story. How did you move to the US, how did your journey happen at the segment and what led to the birth of Hightouch? Yeah,

for sure. I was actually born in the US, so my parents actually immigrated here from India while working as I. T. technology consultants initially did a bunch of government contracting for different state governments Florida, Michigan, eventually Tennessee, and decided to settle down there. So that's where I was born and raised. I got into programming. I was 1st, like, my parents tried to introduce me to programming and technology at a young age, but I didn't have much interest in it. I think they were very disappointed that it didn't click or I didn't get into it, but it wasn't until maybe fourth or fifth grade, actually, so pretty early still that a family friend of ours, whose parents were also in a similar situation, got really into programming games and developing games and little websites. And that's how I got into coding through a friend. Actually, it's much easier to get in when a friend introduces you than your parents. There's some rebellious nature when it comes to that. Yeah, I got into coding, like, lots of different side projects from a young age. Worked at some startups locally in Tennessee, as well as interned for some companies through my school years, like HubSpot right around the time of the IPO and other sort of marketing technology software. And eventually, when it came time to look for a job Segment was a company that was really on my radar. I had used it at a couple of jobs as well as internships. And I reached out to the founders to see if I could join cause I was just a huge fan of the product. So that's my kind of story up until the Segment.

Amazing. And tell us a little bit about what was your role at the Segment?

Yeah, for sure. So at Segment, I was an early engineer there. One of the first 10 engineers at the company. And it's quite interesting. So it leads into Hightouch slowly, but surely at Segment, my first project was working on this product that they called warehouses and cloud sources. So it was basically initially this was at the time that Amazon Redshift was gaining some steam in the market and every company, previously, it was very difficult to use data warehouses before Amazon Redshift was in the market. You had to manage your own servers. They were like expensive technologies out there. You'd probably have like a team of DBAs to manage this stuff. And then Redshift came about and was like, Oh, we can manage a data warehouse for you in the cloud. And now Redshift is seen as a technology that's not so great. But back then it was the hot new kid on the block. And Basically with Amazon Redshift, like, the question was, how do you get your data into this? Like, how do I get my data from Stripe? How do I get my event data from the website? How do I get my data into Redshift so that I can run queries and run analysis out of this fancy new data warehouse? And Segment was really specializing in the problem of helping companies collect their customer data. So we decided to put out a product in that space. And I think at some point Segment was the biggest writer to Amazon Redshift warehouses. We were writing to thousands and thousands of Amazon Redshift warehouses and you were really tight with AWS as a result that and over, over that time, like the first question was really, how do I get data into this data warehouse, but over the next four years that I was at Segment, I saw the way that companies looked at data warehouses just drastically changed. So first, the question was like, how do I get it into there? And how do I build a BI culture and different solutions, like, Looker and, Mode analytics, Periscope ,Sigma, all these different startups started rising up and bringing data to the people at companies. Like not just people who can write SQL, but a bunch of different roles. And by the time I left the Segment I started to see that, the data warehouse was initially just a solution that people would use for advanced analytics, like when Adobe Analytics, what's called Omniture, or Google Analytics, or Mixpanel, or Amplitude, couldn't answer their questions. They needed some extra power. They would reach out to Data Warehouse and BI. By the time I left Segment, I realized that the Data Warehouse was the source of truth. for everything around a company. Like everyone was looking at the data warehouse as their source of truth. And that's why I decided to found Hightouch actually.

Okay. So then comes the point, tell us in the simplest word, what does Hightouch do?

Yeah. So what Hightouch does is we help companies take data from all the systems where they have it. So these data warehouse systems like Snowflake or Google BigQuery or Amazon Redshift and bring it to the systems where business teams live so if you have a sales team working out of salesforce and you know They can currently only see when's the last time they reached out to a contact or what's the email they sent or? What sales stage are they in and you want to equip that sales team with information like- hey When's the last time your customer, you know, logged into the product? How many users did they invite to their workspace? How much billing consumption do they use in your app? In the last 30 days. Hightouch makes it really easy to take that data from the warehouse and just ship it over to these downstream tools whether it's something like Salesforce for sales teams Gainsight for customer success teams, Marketo, Braids, Salesforce, Marketing Cloud, Facebook ads. We plug into like 200 plus systems and just make it really easy to get data from the warehouse into these systems.

And how did companies used to do these things before Hightouch or still if companies are not using Hightouch, how are they doing it right now?

Yeah, for sure. So basically what I'd say is like a lot of companies we see are something we call data rich, but access poor. So tons of data in the company. That's not the problem. They just don't have really good ways to access it and ship it to the right places. Before Hightouch, we see a number of things. One, a lot of times people are truly just shipping CSVs around, like they're just, sending a CSV of the, users that match this criteria or or refresh of your full customer base or something like that to a sales team or marketing team so that they can upload it to their tools on a daily or weekly cadence. A lot of other times, teams get tired of those CSV uploads and need them to happen faster and end up tasking an engineering team to code some scripts or, nightly or hourly data pipelines that just automate these API calls. But before Hightouch, there wasn't really a solution that made it so easy to take data from your existing systems and ship it to all the places you need.

Is this what is called reverse ETL?

Exactly. So it's a funny term. We coined it in the early days of Hightouch. Everyone's familiar with ETL, putting data into the warehouse. And we do the opposite. So reversing that process, taking data from the warehouse and bringing it to all the systems, all the systems where people live. And it's quite funny, in the early days people were really opposed to the term they'd be like what is reverse ETL? It makes no sense. ETL can be in both directions or why would I move data out of the warehouse? I just put all this effort to get it in there. It's super funny how the technical community responded, but over time it really picked up because I think. It just became this obvious missing thing in the market.

Yeah. And I think so, I can dive deeper into, and I would love to dive deeper in terms of how do you guys maintain those connectors and all those complexities around that. But before I get into that, let me take a little bit of a step back. First question that I have is, when you started Hightouch, right? You were probably one of the first people to do this, right? As you said, you were the one of the people who actually coined this term, and popularized it. How did you validate as a founder? How did you validate and what was your first moment where you thought? No shit. This is something that people really need and this is not just something that is there in my mind.

Yeah No, that's a great question. Let me think back for a second So I think we, when we were working at, as with me and my co founders, so there's three of us it took it's a funny story, before we actually worked on Hightouch, we were working on some other startups, actually in the travel industry, COVID took our whole startup journey by a turn and, made us pivot and find something else. From the beginning, we were good for a good reason. Yeah, I would definitely not want to be working in the travel industry regardless of COVID. But it's interesting. We took about six months to say, to hit what I would call product market fit. But we really knew the pain was strong from the beginning. Like, no matter if a company was using a solution, like Segment was super advanced, was super primitive, like. They were having trouble getting the right data to the right business teams at the end of the day. And the solution for what we were doing changed throughout the course of those six months, but the problem was the same. It's quite interesting. We got a few signals from customers while building early versions of the Hightouch products. Saying, I just have data in my warehouse and, or I just have data in Looker or Tableau. I want to get into Salesforce and we initially, we didn't think that was like a big enough problem to build a company around. I think especially coming from Segment, a lot of people think, oh, it must have been so obvious coming from Segment. It's actually the opposite. I thought, I thought maybe this is too small of a problem. Maybe we couldn't really focus on building a company around this because it seems like something someone else would pursue. But actually what I found was that. Yeah. There was a whole like persona that was being really underserved in the market and was becoming the center of attention and the source of truth when it came to data. So Segment and all the previous sort of solutions that helped you get data to your tools, really focused on different personas like software engineers or the iPass space focused on like IT and how they can plug SaaS tools together. But really what I found is that there's whole communities of data teams, data engineers, data analysts, analytics engineers, data scientists who are holding the golden key to the castle and the full knowledge of the company and these data warehouse data lake type systems. And that's becoming the centre of the company as a result, it makes sense to build something that they can use on top of the platforms that are becoming the source of truth in the company. So once I realised how different of a persona that was, and they were just using totally different tools. Looking at the problem in different ways. And would never really use a solution like a Segment that you have to instrument on your website or your mobile app and do all this like SDK stuff. I realized, there's actually an opportunity to build an independent company here because there's just like a massive wave happening in the industry.

You just mentioned a very interesting thing is, data warehouse becoming like a heart of the data teams, and that wasn't the case up until few years back, data warehouse where still treated as a kind of a second class citizens, and this is something that probably changed in past couple of years, with the evolution of the modern data stack and kind of a bunch of companies agreeing to the fact that data warehouse is indeed the heart of the whole data ecosystem and data engineering. Tell me, how did companies like dbt or companies that truly profess the data warehouse being the centrepiece of this entire universe help you? Is that something that you would attribute to the success that you have seen with Hightouch?

Certainly I think there's a lot of market tailwinds that really benefited Hightouch and gave us early growth and attention in the market, early on we found customers that were rallying around this idea of a modern data stack, or we're adopting Fievtran, dbt, and we're bonding this vision of the data warehouse is the centre of the company that's, started using Hightouch and really found the idea fascinating. A lot of those slack communities like. The dbt community is locally optimistic. We're pretty pivotal to our early growth and maybe our first 20 or 30 customers. Now, like zooming fast forward, it's a few years. I don't think it's as big of a movement that we're riding or following, we're really more just following companies that are investing a lot in Snowflake and a lot of our largest customers don't really use other tools in the modern data stack other than Hightouch. Funny enough. But I would say more so than the modern data stack, what really, like, made Hightouch positioned to be so timely was this idea of actually, like, yeah, ELT in particular. So I think the Cloud Data Warehouse, made it a lot easier to just say, okay, let's dump a bunch of data into the warehouse and then let's transform it and build these golden models in the data warehouse itself. So that actually made a unique set of data available in the data warehouse that wasn't anywhere else in the company. People started using the data warehouse to build formulas like lifetime value, churn risk models, all sorts of different things that just didn't exist anywhere else in the company. It wasn't just a place for ad hoc analysis. It was really a place to build these source of truth definitions about your customers, your company, etc. And dbt only catalyzed this and made it a lot easier. But that trend of ELT is actually what I think really made Hightouch the right solution at the right time because there's all these companies where the data warehouse was not just becoming the most comprehensive source of data. It's actually becoming a data silo, funny enough because it had this unique information about your customers, about your product, about your business, about your financials, that just didn't exist anywhere else and obviously needed to be sent everywhere else. So I would say that is the trend that we really rode at. At Hightouch more than anything else.

Got it. And I think so and this is something that I've also seen from my personal experience even before Hightouch people were sending data back to these CRMs and other systems, but they would probably do it. And, I probably did it using those scripts. I would just create a script that would execute API and publish it to those operational systems. One important thing that we have seen in the ETL industry is the maintainability of the connectors. And I think that's the biggest challenge, even if you look at the history of ETL as a sector you had those singer tabs that would allow you to color, collect the data, but they were not well maintained and that led to the birth of air bite and so on. And we know the history, how are you guys taking care of. Connecting, maintaining those 200 odd connectors that you just mentioned.

Yeah. So you hit it on the, you hit it on the nail there, right? I think building connectors, building a one off script, not that hard. Scaling those with new business requirements, maintaining them as the environment changes, the API changes, the data changes, as the schema changes, that's where the challenges really exist. And, yeah, internally, we've built a lot of functionality, like, almost like a framework to make it easier to build these connectors, and not just to build them, but to version them to test them for things like speed as well as throughput to test how we handle different air conditions or weird datasets that we might operate on top of across all destinations. So we basically built a really interesting framework around this both for I would say the ancillary concerns of building connectors. So testing it and measuring it, et cetera as well as the operational concerns. So things like monitoring the connectors, figuring out when there may have been an issue introduced. Or that comes up to you to an API change or something like that, as well as even developing and shipping the connectors. Things like, audit, like having really good libraries to do things like retry or ding. So something really important in Hightouch is that a lot of people write connectors. It's like a simple loop over the data and the data warehouse you mentioned. Send each row or send a batch of rows to a downstream system, like a Salesforce or Adobe or whatever it is. And we actually do it a bit differently. We actually have like automatic change data capture off the data warehouse. So only the new rows, change rows, removed rows would be sent to downstream systems. They'd be automatically batched up. If one row fails, we'll pull it out of the batch and put it in a dead letter queue after retrying it a bunch. We have, like, a bunch of mechanics, that also make it a lot simpler. So you don't have to implement these in one off, one off scripts. But right when we write a new connector you get all these things, like, out of the box.

You mentioned something very interesting here is you have built a kind of a change data capture from these data warehouses that allows you to read those change events and then execute on the top of that. Does that mean you? Do not create a workload on the data warehouse itself. That means there is no additional cost of operating Hightouch on the top of a data warehouse. No complex queries or those workloads are because all of these data warehouses are computed based, right? You pay per comput. So does this mean Hightouch is not increasing the compute cost of my data warehouse?

Yeah, we do. So we do drive compute on the data warehouse and simply running a SQL query, whether it's on a stream table from a snowflake that shows the change data capture or whether we're building our own change data capture inside of your data warehouse, with temp tables that we maintain, we do this automatically and have a bunch of different modes. It does run compute in one way, shape or form on your data warehouse. But compared to all the other processes, it's not like Hightouch is driving up the compute, like crazy in the data warehouse. And we've definitely put in a lot of work to make that process efficient. So it doesn't blow up our customer's bills.

And you just mentioned that, you've also developed a framework internally to be able to build and maintain these connectors. Do you also plan to introduce or release this framework to the community so that other people can also contribute to the long tail of connectors that you probably aren't getting to right now?

Yeah, It's a great question. It's a great question. So we do have a public framework for building connectors. So we have a whole custom destination framework online. It's not quite as sophisticated as what we have internally, but it does buy you a lot of the benefits that I mentioned, like. Automatically doing change data capture, automatically doing things like batching, giving you frameworks and methods to send rows that you need to be retried in future syncs. So it does have those sorts of functionality built into it. Internally, given that we all code in the same programming language and stuff like that, we've built more stuff to make the integrations easier to build, but we do have a framework out there and We are always looking for ways to make that better and better because it's something that's used quite heavily from our large clients.

Amazing. And you guys started off with reverse ETL, but I think so now you have two more suites of products. One is composable CDP and the other one is the personalization API, right? Tell us the story behind these two products and does that mean are you guys now heads on competing with Segment on this as well?

Yeah, great question. Great question. It's a loaded question. So I think first and foremost, one of the most exciting parts about working on the reverse ETL product and working on a product that's so horizontal, right? We integrate with 200 systems and we integrate with about a dozen different data warehouses as well as source systems. And companies are using Hightouch for all sorts of different purposes, right? Getting data to sales systems, getting data to marketing tools. Running targeted advertisements off their data, powering ERP systems and like commission calculations and financial systems and all sorts of things. And what's really cool is that we get to see how people use their data to power their business operations. And we always try to learn more about not just what are people doing in the Hightouch product, like, what features they're using to sync data, but what are you doing before Hightouch? And what are you doing after you put data into one of those systems, what is the overall business problem you're trying to solve with our software? And that's been a big focus from the start. Just when we were three founders and as we scale the product team and the marketing team here, we've always laid an emphasis on really understanding our customers at that level. So when we did look into our customers at that level, we actually learned that sometimes people are doing things inside of reverse ETL where we can actually then go to take their specific use case and build an application on top that streamlines that whole business problem of activating your data. So the customer studio is one example of this on top of reverse ETL, where you can place a SQL query in or dbt model and sync fields from those into systems like Salesforce or Marketo or Braze. We've actually built this thing called Customer Studio. It was initially called Audiences that's like a full featured marketing app. Where marketers can actually come into our app and do some reverse ETL, they don't call it that, themselves. And what that looks like, it's like, they can build their own models by, or audiences in particular, by filtering down the tables. If you have a users table, we built a UI where marketers can come in and say, I want to find users who added something to their cart. But I didn't check it out within a day. And I want to send those emails, send emails to those customers and ads to those customers. And not right there is a data activation problem, but what we saw is that, companies were having marketers ask data people to create one model on Monday, create another model on Tuesday, create another model on Wednesday, all slight variations of the previous ones. And just like have this back and forth process in the company. And when we saw that happening so much for marketing and advertising tools, we thought, man, like, marketers want to get their hands on this data directly. We've talked to them, we know that's a big driver. Why don't we just build part of the, like, an additional layer in our platform where marketers can start using reverse ETL and data teams don't have to support them in other systems, they can support them through the data warehouse and through the source of truth at the company. So that's how we came up with with the customer studio. When it comes to personalization API, it's interesting. It was the same process. We saw companies syncing data to databases like Redis or DynamoDB or Postgres or MySQL from from Hightouch. So they were like, I want to take this data point like lifetime value or propensity to buy and sync it to one of my more transactional production databases. So we started reaching out to those customers and saying, what are you doing? Like, what are you doing with getting this data into a production database? And we learned a lot of them were trying to personalize their actual product or application experience. And that there were even additional problems we could solve on top of them trying to personalize their application experience. Because if you just put data in database that's one thing. But then you have to build an API. Now what happens if you want that API to be called from different regions you're a media site and some people are reaching out to it from India, some people are reaching out to it from Japan, some people are reaching out to it from the US. And you want it to return at really low latency? So we, once we actually started talking to customers and figuring out what they're trying to do, we built a product around it called personalization API, which gives you an API off your data warehouse that can do just that. Instantly.

Wow. That's really interesting. And, as a founder. This would have been a very challenging thing for you because you shifted your go to market strategy from focusing on developers and data engineers to now marketers. And now you're talking about software developers. How did you manage this internally?

It's a great question. It's a great question. So first and foremost, we focus on. I actually still the data team, so we try to get into companies mostly via the data teams that are powering these processes or working with other teams on things like personalization or marketing or sales operations or stuff like that. We focus on that data team persona because, you can't go to market with 3, 4, 5 personas, but we had to train our internal teams, sales teams to be able to speak to multiple stakeholders to the company, not just to be able to speak to data people, but to be able to speak to marketers, marketing technologists, or, performance teams, or, product engineering teams, because while data teams are involved in the process, and can you bring us into it saying, Hey, this is a really novel solution to what we're trying to solve at our company. They're not the only person you need to talk to get the deal done and to explain to the company as a whole, especially in larger companies, like Warner Music or some of the Fortune 500 companies we work with. You have to talk to multiple stakeholders to actually get the deal done. So yeah it's definitely been a challenge. It really comes down to investing a lot in enablement and, teaching the team the right messaging for talking to these people. But our focus on data teams has actually been a key to our success. As a company. And we're just, you were making a big bet that more people at the company are becoming data literate as well as data teams themselves are interacting with more and more stakeholders and more and more processes the company as a data warehouse becomes, the best source of data in a company.

And with that , where do you see the future of Riverse Etl? And, we have recently seen, segment coming into this, they'll, they launched their own kinda reverse ETL platform within the Segment platform. So where do you see River Reverse ETL going on from there? And what's your take on Segment doin this?

For sure. So fundamentally I'll answer the question with the segment first. Fundamentally, there are systems that really use the data warehouse as a source of truth and there are systems that don't so like Segment while and many companies in the sort of martech space while they're adding And these connectors or integrations with things like the data warehouse, there are two things. One, you're creating a redundant copy of your data, not just in the data warehouse, but in Segment. And then two, you're leaving a lot of the power of the data warehouse behind. So if you look at how we've built our product both normal Hightouch as well as Customer Studio. We really understand that every business is different. And, this idea of a customer 360 or, whatever people are talking about, single source of truth. It can't just be boiled down into, like, users and event data, which is really what customer data platforms like Segment focus on. They really just focus on, tell us about your users, tell us about actions they're taking. But you know, every business is different. Ticketing sites have artists and tickets. Retail companies have, products and inventory. One of our customers, PetSmart, has an idea of pets and it cut the customer's actual pets that they need to track in their loyalty programs or a bank has the idea of a bank account. So there's all these related objects that really are important to build out this idea of like a customer 360 and Hightouch has the ability to tap into all of those in your source systems and to write to any objects in the downstream systems to just making it the most complete platform and then being built natively on the data warehouse as well. You can use it instantly. You don't have to have like a huge, million dollar contract to get started with something like Hightouch. We allow you to connect your data warehouse, connect it to downstream systems, and just start chipping away at solving use cases at your business. Which is really powerful. Plus the costs are way cheaper, given that we're not running two data warehouses, one for ourselves and one for you, we're really leveraging the system you're already, you've already invested in as a company.

And your thoughts on the future of Riverse ETL, where do you think this segment would go from here?

So basically two things. I think one reverse ETL it is a platform for the data team in the end of the day. And we're actually causing this change in the way data teams work in the business. Data teams previously used to answer questions, build reports, stuff like that. Now we're actually allowing the data team to build real operational, automatic, live business processes that, that power the business. So things like. Powering targeted advertising off the data in your data warehouse work. Routing sale, your whole team of salespeople to the right customers and the right leads and helping them prioritize their day based on the end of the data warehouse. And when you do that you need to make sure these pipelines are extremely reliable and are things that you would depend on as a business. And when you think about software engineers, when they're building product and production applications they have a bunch of concerns when it comes to building applications that are super reliable and you would rely on for your entire business. They do things like version control their applications, do things like have really strong observability and alerting when things are going wrong. When things do go wrong, they can see exactly what's going wrong with debug logs and metrics and traces. And there's just basically all these concerns that CI is another one that software engineers have that data teams often don't have. So when it comes to, like, making the Reverse ETL platform better not only are we making it more scalable, more integrations, easier to use, but we're bringing all these software engineering concerns to the data world and making those really easy so that data teams can make their pipelines extremely reliable and avoid mistakes or mishaps or catch problems as soon as they happen in these data pipelines as they're powering more critical business processes. Now, in addition to just making the Reverse ETL product that exists today really good. I think what's interesting in the space is making the idea of activating your data, A, more self service and B, easier for certain popular use cases. As I mentioned earlier, in addition to Reverse ETL, we've actually shipped a product called Customer Studio, which allows marketers to come in and build their own audiences, sync them to different places like ad networks and marketing tools. Run analytics, run A B tests, all on top of the data in the data warehouse. And that idea is that, data activation shouldn't just be for the data team. The data team should be able to define the data that matters, and other business teams should be able to come in and get their jobs done as well. Personalization API is different, but a similar concept. We saw so many companies building their own APIs off the data warehouse, and we just made that process a lot easier. And handle the nth concerns that you have as you're scaling those APIs and making them more powerful. So I think what's really exciting is building additional functionality on top of reverse ETL. That makes it easier to satisfy certain use cases and easier for more people around the company, whether they're in RevOps or finance or marketing or advertising to come in. And leverage all the wealth of data you have in the data warehouse to get their jobs done.

Basically dive, going deeper into the value chain, right? That's what

exactly. Yeah, exactly. Got it.

As we inch closer to the end of the episode they just I have one question as a founder to another founder is we have seen the Cambrian explosion of tools that are coming out of modern data stack. And, in the past year or so, we have seen a lot of corrections. We have seen a lot of consolidation. We have seen a lot of hype tools fading away. And it's a natural correction that is happening in the industry, in the market itself. What would be your advice to another founder who is building something in the data space? In terms of how they ensure that they don't have a technology that is looking for a solution versus a solution that can be solved using technology? What would be your advice for that? That's a

Fantastic question. I think a couple of things, I think a couple of things. One really understand the business problem and make sure this is something that everyone, deeply cares about and really deeply provides value. There are things you could probably sell to data teams, but you need to make sure they're not just exciting for data teams and little tools or widgets is something that solves a first class problem in the business. And as I mentioned, like, we really had this mindset from the beginning of Hightouch. We didn't just see it as sinking fields from the warehouse into different systems. We saw it as. Sales teams are working all day trying to make revenue, but they're not routing their time properly because they're not empowered with all the data and intelligence you have in these data warehouses. Marketing teams are spending millions and tens of millions of dollars on things and on advertisements or email campaigns. And that money and that effort is not being, is not sending the best emails, the best communications to customers. Driving the most efficiency and revenue ultimately because they're not using data effectively enough. So first, just make sure you understand the real business problem that's being solved versus just building data tools for the sake of data tools is something I would really emphasise and then to, make sure that problem is big enough, right? Again, like, it took a while for us to find the right solution at Hightouch. Finding that product market fit means you have the right product solution that the market wants as well. And to eventually land on reverse ETL. That being said, from the beginning, we knew that the problem of activating data was rampant in businesses, and that was a problem that was going to be disrupted and solved over the next 10 years, whether it was Hightouch or another company. So I would say for, my advice to you to all founders is to. A, like focus more on the actual problem you're solving in businesses, not just the data tool problem. And then, you can market the data tool, but just make sure you have an understanding of where it actually fits in and then B, make sure that business problem is frankly, large enough and a first class priority, especially. In an environment like this where people are not looking to buy tools left and right. That was

An amazing advice. So thank you so much for your time Tejas today. That was an amazing episode. We learned a lot of things and we're looking forward to a lot of the next set of amazing things that are coming out of HiTouch. So thank you so much for your time.

Thank you. I really appreciate it.