Drill to Detail - Drill to Detail Ep.63 'Stitch, Talend and the Evolution of Data Pipelines' with Special Guest Jake Stein

Episode Date: April 15, 2019

Mark Rittman is joined by returning Special Guest Jake Stein, former co-founder and CEO of Stitch and now SVP of Stitch at Talend to talk about the evolution of the data pipeline-as-a-service, data catalogs and data governance and the vision behind Talend’s acquisition of Stitch.”The Vision behind Talend's acquisition of Stitch”“Stitch is joining Talend”“dbt: Analytics Engineering that Works”Talend homepage

Welcome back to another episode of the Jewelry Detail Podcast and I'm your host Mark Rittman. And for this episode I'm very pleased to be joined by a friend of the show, Jake Steen, SVP of Stitch. So welcome back Jake and it's great to have you on the show. Thank you, Mark. It's great to be back. And Jake, for anyone who doesn't know who you are, just maybe just explain, just tell us who you are and what it is that you do within the industry. For sure.

So previously, I was co-founder and CEO of Stitch. Stitch was acquired by Talend about four months ago. And Stitch is a data ingestion product. So we help people get data from lots of different data sources into their cloud data warehouses and data lakes. And prior to Stitch, I was actually the co-founder of a company called RJ Metrics, which is a full-st stack business intelligence and data analytics product that included data ingestion, transformation, warehousing, and visualization. And we sold the original RJ Metrics business where we extracted some of the data ingestion technology from that product and made it into a standalone service,

which then spun out as part of the original acquisition and became Stitch. So when they come to write the history of the modern analytics stack, probably the acquisition of RJ Metrics by Magento and the kind of sp-off of yourselves as Stitch and then Fishtown Analytics with DBT, probably one of the most significant dates in that history, really.

Yeah, it's been really fascinating to see, like, the parallel paths of Fishtown and Stitch and how that was really informed by our experience at RJ, where I think one of the things that we learned over time at RJ Metrics was that there was definitely a significant market for the kind of product that RJ Metrics was delivering, where it was an entire integrated package. And it really followed kind of the original model of ETL, where we were ingesting your data, we were cleaning it up and transforming it.

And then we were putting it into the data warehouse and kind of had the reporting on top of that. But after working with lots of different customers and seeing our own analytics, we saw the pull from the market and the opportunity candidly that was created by these cloud data warehouses you know first redshift and now things like snowflake and bigquery and azure sql data warehouse and it it really there there were two big insights i think that came out of that that that are really represented in in fishtown and stitch where one is that basically the reordering of ETL into ELT, where the extraction and loading potentially makes sense to do before the transformation. So get all the raw data into those cloud data warehouses and then do the transformation in the warehouse itself.

And that's really the reason it makes sense that Stitch exists is because that transformation can be done in the warehouse, where previously it wouldn't make sense to do that in your, you know, the Tessa or Teradata box. And then the other piece that I think is really core to Fishtown and DBT, the product that they maintain, is around analytics being a subfield of software engineering, and all the tools of software development like version control and code reuse and collaboration, all those things really being key to the analytics toolkit. And I think that's the core insight behind DBT and how Fishtown works. And we've worked together on a lot of different customers. So yeah, it's been a great partnership with them

and all informed by our shared history. So Stitch was fairly groundbreaking at the time in that you focused on creating this data pipeline as a service for data engineers that really did one job well, which was to transfer data from different kind of SaaS data sources, all different places into the data warehouse, ready to be transformed and do that in a reliable and a kind of SAS data sources, all different places into the data

warehouse, ready to be transformed and do that in a reliable and a kind of cost effective and a easy to use way. What gave you the inspiration for that? And what was the confidence really to launch it as a product? Yeah, the thing that was really helpful for us is that we got a lot of demand for that product from our customers at RJ, even when that was not our product itself. Where one of the other big trends that was happening in the industry is that after, you know, RJ Metrics had been operating for a number of years, there was a new generation of competitors that popped up. Companies like Looker and Periscope and Mode Analytics and Chartio, where there was a lot

of innovation happening on just the front-end visualization layer. Products like Tableau had already existed for a while, but there was this new generation of visualization companies. And we were somewhat competitive with them, but we had a lot of customers who were interested in the data ingestion piece of RJ Metrics and wanted to use that in combination with someone like Looker, for example. And for a while, we viewed this as a little bit of a competitive threat. We wanted that customer entirely for ourselves. And after really going deep with those customers and thinking about what parts of our day we're adding value and how we could work with them, we realized that maybe the most valuable piece of what we were doing was this data ingestion. And the other

interesting thing is that that problem was actually becoming much more widespread and much more challenging to solve because you've probably seen the, you know, those infographics about, you know, the number of different SaaS tools that companies use today. And, you know, it's just exploding. You know, every year there's a doubling or tripling in the number of options out there. And, you know, Stitch even when we were independent before the acquisition, we were a 33-person company, and we had more than 40 different SaaS tools that we used to run the business. And so to get data from all those different things, and you can imagine how that problem explodes at a larger organization, there's more and more different data sources you need to integrate with and all of them have varying different kinds of apis different uh reliability of those apis and and so more and more companies had more and more data that they needed to consolidate and then the other piece of it is that just the the explosive growth of cloud data warehouses like you know if you go

back a decade or two, it was a very small percentage of businesses that had a data warehouse. So the fact that there are tens of thousands of companies using Redshift and that, you know, Snowflake is tripling year over year in terms of, you know, thousands of users, that just like explodes the number of people who need help getting all those different data sources

into those data warehouses and data lakes. So who typically are the people that buy Stitch and use it? Who's the typical customer persona that you target the products at? Yeah, so the most common persona of the user of Stitch is actually an engineer. And at some organizations, they have the title of data engineer. In some cases, they're just engineers who are responsible for provisioning data. And the next most common user is the actual consumer of that data. So that's, you know, the analysts or the data scientists. And they, you know, Stitch, you can go deep and, you know, some of the engineers are, you know, contributing to our open source project or are doing, you know, controlling Stitch via our API, but you really

can, without writing code, just sign up and get started in a couple minutes. And then the last use case, excuse me, is folks in kind of like a project manager or a product capacity where they're monitoring the success or monitoring the results of something. And this is just feeding into their dashboard. So it's really in some ways they're acting as an analyst in that case, just with a different job title. So as well as yourselves, there's obviously in the market, there's Fivetran, there's Supermetrics, for who are more i suppose they're marketers um you know data sourcing company um but what about companies like uh or products like say segment how does uh how does stitch compare to to segment and what they try to do uh yes so it was i think we we started with a very different use case and end goal in mind than someone like segment uh and you Segment has since added some features

to their product so that they definitely have some aspects of their product are competitive with Stitch. But I would say the majority of how folks use Segment is more complimentary. So if you look at the evolution of their product, they started solving the problem of, if you look at the number of JavaScript tags on any given website or SDK is loaded into some mobile app there's a whole bunch of different things and really what you want is to take one event

and send it to a bunch of different services which I think is incredibly valuable and the use case we started with is that companies are using lots of different products to run their business. Some of it's event data, but a lot of it is data that's totally different than what you'd use segment for. Things like accounting information or inventory or the actual ad spend data. And so that's stuff that's totally separate from what you would use a segment for. And so

the use case we're trying to solve for is get the data from all the different business systems you use, some of which are internal to the company, some of which are external and hosted in various places on the cloud, and then get that into one consolidated data warehouse or data lake so that you can use it to power analytics and data science. And someone like Segment has added a feature where you can take the data that you're running through Segment and also send it to Redshift or another data warehouse. And for that piece, they're competitive with us. But we're again, they kind of backed into that where this is our entire focus from day one. So last time we spoke, you told us about the Singer project that you're sponsoring, the open sourcing of the connectivity that Stitch uses.

How's that going now? It's been really fantastic. Yeah, I've been pleasantly surprised. Well, I guess maybe pleasantly affirmed is the better thing to say, just because I think it was probably the biggest bet we've made in the lifetime of stitch where this you know uh so for anyone who doesn't know singer is an open source project where um we basically take the the portion of stitch that actually interfaces with the the external services so it's made up of

two components there's, which pull data from data sources and targets, which send data to destinations. And we have basically been in the process of open sourcing all of our integrations, as well as allowing community members to customize some of the existing integrations that exist or build brand new ones. And some of those we include in Stitch. All of them you can use totally separate. These are all self-contained programs that you can run on your own or you can run on the Stitch platform.

And we were hoping that we get a lot of community engagement and hoping that a lot of people would get involved and use it and contribute. And I'm looking at this spreadsheet right now where we keep a little bit of an inventory and there are 92 community built integrations in Singer. And those are things that, you know, some of which our team contributed to, some of which we've offered feedback on, but this is all stuff built by people who are not employees of Stitch or Talon now. And that's been really great because,

you know, the business problem we have or that we had is that there's always like, you know, lots of companies use MySQL, lots of companies use Salesforce and Google Analytics. But there's very often, you know, a handful or just one data source that's really specific to this particular business. And it's not going to be on our roadmap. It's probably not going to be on the roadmap of any ETL vendor. And so this enables our customers to build new things themselves, or it also gives an opportunity for our system integrators and consulting partners to build out new things and allows like one of my favorite examples is

we actually have a customer that's a chain of go-kart courses. And there's a go-kart course management software called ClubSpeed. And we did not have a ClubSpeed integration, but they worked with one of our partners. They built out a ClubSpeed integration. And now I can confidently say we are the preeminent go-kart course partner for anybody who needs data integration. And just things like that that are unlocked by enabling anybody to build these integrations and it's also been really cool that you know a lot of this gets used on stitch but other companies have also gotten involved so

gitlab is a big contributor to um to singer and they use it in some of their open source projects something called multano, Data.World, also built like a desktop singer runner for folks to just run it from your desktop and make that easier. So it's been really interesting to see people doing it for their own use case, people using it to improve their products, and people just building it

because it's a cool open-source project. So yeah, I've been thrilled with it. So the last time we actually met in person and talked in person was the Looker Join Conference at San Francisco last year towards the end of the year and then of course the big news since then was um that you were acquired by talent so you know what was it that was what got you interested in talent and what made you you know pick up the phone when they rang and um you know express some interest i think the i mean the

thing that first got me interested in talking to them is that I had met Mike Tuchin, the CEO of Talon, years ago. And he reached out to me and, you know, I thought he was just a super smart guy. So I'm definitely always interested to talk with him. But the thing that got me actually, you know, interested in the conversation and continuing it was that there's a real, like, alignment of vision in terms of where we see the world going. And we each had something that the other one was, could really gain from. And that, you know, the whole would be much greater than the sum of the parts. And so for, from the

Stitch perspective, we had always, you know our strategy was land and expand, where we really wanted to provide a very fast, frictionless experience, very fast time to value, and let people get up and running in minutes and get value. And then if they like it, they can do a free trial and become a paying customer. And we had that going well. And we have thousands of customers using Stitch. And one of our big strategic goals was, okay, how do we figure out more products and services? How do we deliver more value to these companies? So our selfish reason for doing that is we want to generate more revenue. And so we were in the process of doing that is we want to generate more revenue. And so we were in the process of doing that and

we were staffing up and we were selling some larger deals and expanding the use case. And that was something that was already part of our vision and a big priority for us. And we saw that the companies we were working with, this was obviously not the only data need they had. And so we just wanted to make sure we want to solve a bigger and bigger part of their problem. And Talon, on the other hand, I mean, you're absolutely right. The company's been around for a while and their legacy is in, you know, they started as an on-prem ETL company. And so they have been undergoing this really fantastic transformation, however, where they, you know, several years ago, they launched the cloud version of Talent. And it's by far their fastest growing product. It's growing at a huge,

huge rate year over year. And they kind of see the future and they want to accelerate that transformation. And they also are looking for basically the way to land new customers in this frictionless way. So when Mike first approached me about the acquisition and he said, this is where we're going. This is the transformation we're on. We want to accelerate that. And we love your go-to-market model and we want to meld it in with ours. And the other piece is that all those different things that companies need to do once we ingest their data, it's transformation, it's data quality, it's governance, it's all those different pieces.

They've already got all these products for that. And so the more I talked to them, the more I got excited about basically doing the things that we wanted to do, just accelerating the timetable and doing it on a much bigger stage at a larger organization with bigger budgets so yeah it was something that's i'm still really excited and when i heard about the acquisition um i thought the timing was was good because i was increasingly hearing from customers from the industry from you know people speaking to people that i suppose people's needs and requirements now around data integration and uh i suppose the curation and management of data are getting more and more kind of demanding. There's more to it now than just kind of landing it and transforming it.

Oh, absolutely. Yeah, this is and it's so interesting because, you know, the conversations I've been able to have with customers before and after the acquisition are so much candidly better now. And this just happened where I was at a Snowflake event a couple of weeks ago. And the first half of my presentation was, here's Stitch. Here's how we can help you get your data into Snowflake. This will unlock your ability to do a POC. It's super fast time to value. It's going to be great. And then, you know, three hands go up and say, this is all great, but I'm at this, you know, multi-billion dollar multinational company. I've got, you know, 10,000 different tables in my database. I'm covered by GDPR and the California equivalent. I forget the acronym now. And like, I really need compliance with all these things. And now I'm

able to say, yes, we've got a data catalog that can help you do that, that can both empower your internal analysts and can help you comply with GDPR, right to be forgotten notices, and all those different things. And it's like, customers have had these problems since we started the company. Before, we were just very careful to say, this isn't a problem that we can solve for you today, but we'd love to hear about it. And maybe we'd refer you to a partner. But now, this whole ingest and then integrate and then manage and then govern, we see that as an integrated customer journey. And it doesn't necessarily always happen in that order order it's kind of a circle where all the pieces lead to all the others um but but yeah this is um all these different things that we were starting to think about in

our exploratory product development um are now something we can just you know go move headway faster on in the last episode i of the show that i recorded with Mike Ferguson, we talked about enterprise analytics, enterprise data warehousing, and then we got on to, in particular, data governance and master data management. Are they, again, areas that you think maybe are interesting to invest in and maybe work on together with, you know, talent? So, yeah, short answer is yes to both. There were definitely folks who would ask us for that sort of thing when we were independent. And we could kind of tell a story about that, you know,

and, you know, either refer to someone like talent or, you know, let's combine Stitch with DBT with a consultant and, you know, we would clean up your data. But I think there's some fundamentally different challenges, which I think, you know, Tristan and the folks at Fishtown would be the first to say dbt is amazing at some things and it's not a solution for something else. Whereas when you have a product that's like built from the ground up for data quality and, you know, the whole, some of the inherent value prop is that we're going to score the quality

of your different data sets and also do the transformations that will clean up the quality and you can track the quality improvement over time. Because, you know, there's one thing of having like a quote unquote golden record. But, you know, odds are that when you implement it, it won't be perfect you're going

to learn here's this other rule that we should set up or here's this this better deduping or whatever you do so tracking the actual improvement on that over time i think is is really key and so i think with with something that's purpose-built for that you can really um move into a much more nuanced sophisticated view of how these different pieces fit together but i suppose the other side of the coin is that you know master data management and enterprise etl suites is you know it's been done before the other companies out there doing it how do you see stitch and uh and talent really you know differentiating themselves in this market so uh there's i think the the question behind the question, if I can guess, is that,

you know, like, how do we, how do we take some of the stitch DNA and infuse that into talent products and services? And I, I, I think that there's definitely some concerted effort right now going on to, to see how we can do that in the right ways, but also not, not in every way, because I think there are some products which, just for example, I think just to make up a product customer pairing, I doubt that General Electric is going to purchase their data catalog in the same way that a Stitch customer tries our Salesforce to Redshift integration. They're not going to go to the website, download it, do a 14-day trial, and then swipe it on their credit card. Some things, in the same way,

they're not going to buy their new accounting system in a totally frictionless way. So I think there are some products that really make sense as part of that expand motion, and some products that make more sense as part of that land motion. And once we have a relationship with a business where they're using Stitch to load the data, maybe using... But I think one area where we are really focusing on getting some of the best parts of Stitch and infusing it into some of the talent products is with a new product that's coming out soon. It's in the early adopter program right now called Pipeline Designer, which it's basically a totally cloud native experience. And it'll have the easy to try, easy to buy, very great, candidly, a fast time to value. And that's another thing where we can take the best parts of the Stitch experience

and marry that with some of the transformation and quality aspects of some of the other talent products. Whereas when you're using things like that, then you can ladder up to something like the data catalog, which is going to be a more enterprise-wide, more highly considered type of thing that someone uses. So I suppose now you're getting involved in enterprise sales and enterprise customers with talent. The kind of sources and systems you need to integrate with now are going to, I suppose, be changing. You must be encountering things like Kafka now, for example,

on the list of technologies that customers are asking you to bring in as data sources. Two different parts of that, I would say. One is, you know, for these organizations, do they need different like data sources and destinations? Like, you know, the talent customer, you know, needs SAP integration and Stitch doesn't offer SAP integration today. So that's something that, you know, or Kafka integration, I think that's a great point. And I think as we do more cross-pollination across Stitch and the talent customer base, I think you're 100% right that one of the things we're thinking about is which integrations were not on our roadmap, at least on the near-term roadmap, that we should add because

of, you know, our updated view of who we should be serving. The other piece of it is, you know, what are they potentially doing instead? And I think that is actually pretty similar, where folks are, you know, there's the same number of people who are, there's always people who are trying to roll their own. And there's a, you know, somewhat different competitive set. But I think that element of it is less changed than the making sure we can meet those customers where they are with the the tools and products that they need to integrate with so now that stitch a part of talent would you ever use or would you continue to use dbt in the future for anything oh sure and one thing i'll say is that you know we at stitch still use dbt a ton

um for our internal analytics. And I think it is a lot of the things that I think are great about Stitch are also great about DBT in that it's really focused on, you know, a particular problem that a lot of people have. And I think it's great at solving that problem. And it's also great that it doesn't try to boil the ocean. I think that's in some ways DBT strength. And one of the things my colleague, Sean, came up with this tagline, and I love repeating it any chance I can,

where we used to talk about ELT as extract, load, and transform. And now we've kind of broadened our horizon to say ELT is extract, load, and then talent. And so when you need all those things as part of an integrated suite, like transformation and data management and data governance and data quality, I think dbt is not trying to solve all those problems. And so if you were to try to use it to solve all those problems, you would be very disappointed. But if you're very interested in command line, SQL-based pushdown transformation, you know, DBT is a

phenomenal tool for that. So I think it's, I totally expect, and I would be shocked if it wasn't the case today, that there are people, you know, organizations that are using talent, that, you know, there are people that are also using dbt on different parts of the org and so i think it's really based on the persona of the user based on the job they're trying to do um and i do think that there's probably opportunities to make those things work better together and then you know uh in the same environment to make sure they're orchestrated better um but yeah i do think there's a place for both well to my mind really it's just about understanding the job you're trying to do you know is this about is this about

you know an analytic workflow a data modeling to support some analysis and reporting or is it more than that and that's when you start thinking beyond um i suppose um you know transformation frameworks and uh and think about etl tools and data integration suites. Yeah, I think that's a great point. And it was interesting. I know you were at Looker Join this year. They had a session about, you know, when should you do your data modeling and transformation purely in LookML and when should you move it into your ETL tool?

And I thought I actually didn't, I wasn't able to attend the session, but just the fact that they are very upfront about the fact that, you know, LookML is a great product and it enables their users to do a lot of amazing things. And it is not the be all end all solution for, you know, data transformation.

And I think, you know, I have a very similar mentality about dbt where there's a whole category of uses that it's great for, but there are certain things that should be done with a different framework. And you're right, where there's some cases that the data needs to be transformed prior to loading. And just to give you an example, a highly nested data coming out of MongoDB getting loaded into Redshift if you were going to de-nest that data

Redshift has a table and column limit that you can potentially hit and so you need to transform the data prior to loading it that's not every case but there's some cases where Stitch would not be a good fit for those customers so I think there's lots of different opportunities where, to your point,

you start maybe with one tool and you realize a different framework is necessary to solve your evolving problem. So looking to the future and going forward, what's your vision around this kind of area? What problems do you still see that need to be solved from customers? And what's Stitch's roadmap really going forward in the kind of near future? I think about the goals in a couple of big categories. There's, you know, just growing Stitch.

You know, we've got some aggressive, you know, customer acquisition goals that we're focused on in 2019 and beyond. So there's a lot of things that we're going to do to make the product better and attract more folks. There's also things around the joint user experience. So making sure that it feels natural to a customer that they would start with Stitch and add on additional talent products. And thinking about that in a way where we really don't want to compromise the things that made Stitch great, but I want to think about when is the right time and when is the right point in the customer's journey to offer them the opportunity to do something more. So we don't want to get in their way when they're starting, but once they've had success with replicating data source X into their destination, they're probably the right time to think about, should you be transforming it? Should you be, you know, adding this to your GDPR compliance process? And so there's the, how do you be thoughtful around,

you know, helping people solve more problems? And then also, how do you just make sure the user experience is cohesive? So it's things like, you know, single sign on and having a similar look and feel and enabling people to, you know, pay for multiple products, one bill, like that's all stuff that, that we want to tackle. That's, that's kind of just a table. What about another, another related area, data quality? What's your thoughts on, on that and any ideas you've got around products and again, you know, problems that you'd like to be able to solve? Yeah. So I, there's, there's two categories of things where I think we can make a really big difference. I mean, it's a very big problem. But I'll mention

I'll mention two right now. One is just taking, you know, the vast majority of data sets, I don't think there is actually any data quality consideration. You know, people, there's some stats from IDC that it was, it was something like 55% of data within organizations is unaccessible. And of the 45% that is accessible, 47% of it is unreliable. So it's, you know, less than 25% that overall is, you know, available and trustworthy. And so making sure that people can access it quickly, you know, with speed, and then they can trust it. When when they've got it, that's a big, you know, theme for talent, speed and trust together. So the, I think just surfacing the fact that these different data sets exist, and that they need to have some thought and grading and,

you know, work done on their quality, I think is step number one, just making sure and including that in the catalog itself, because that metadata about whether or not, you know, that the data exists and that the data has been cleaned up, that it has been governed, that's a really core piece. And then, so one piece is kind of like that rising tide that should lift all the boats. And then there's a lot of work that can go on in these different areas that are very domain specific. And so you could think about a simple example that's really common is addresses. Lots of different companies have address data. It's a perennial problem that I live in Pennsylvania. Someone could spell it PA,

it could be lowercase a, it could be pen, all these different things. And so helping people in that, there are several town products that like ship with, this is how you clean up addresses. And so there's a whole bunch of domains where you can take advantage of both things that are built into the product and commercial databases of enrichment sources, as well as open source, uh, libraries for, for cleaning up different datasets. So I think it's, it's kind of this continuous process of servicing quality as a concern across more and more data sets, and then giving people the tools to get a lot of the value with a minimal amount of effort. And then at some point for everything, you need to go very specific

to each business and use case, and ultimately someone needs to be writing code. But I think you can get a lot of the value by giving the right prompt to the right people at the right time and surfacing the right tools that'll help them, you know, get a lot of the way there. Okay. Okay. So tell us a bit about talent. I mean, probably actually a lot of people, a lot of the, most of the audience probably be less than familiar with talent. Tell us a bit about who they are and the the philosophy behind what they do and and how they work sure yeah

talent um started uh in france so you're right about that and they still have a very large presence uh in france they're now headquartered uh in redwood shores california and and they're really like a worldwide organization there There's folks in California and France and Germany and China, lots of different places all over the world. And it's been really fantastic for me, getting to know the folks. I went to the Talon sales kickoff in January and had lunch with a couple of guys from the Japan team and got drinks later with some folks from the Australia team. So it's a very international business. And one of the things that I've been really fascinated by

is that they are able to keep, I think, a really great consistent culture around all those different offices and around all those geographies. And I think part of the way they've been able to do it is that the company was international from day one. And so they are great about things like working remotely. They're great about kind of taking the interesting parts from all these different cultures that

they have. And they've also got customers in all over the world. So when we're thinking about, okay, let's do the customer advisory board, we have all these different diverse viewpoints and backgrounds represented, which I think makes it just a more interesting place to be. And also helps us get to better answers faster, which is a lot of fun yeah so that um that international aspect must be interesting and quite refreshing um i also understand they do quite a lot bit with um with open source as well there's quite a lot of open source sponsorship and open sourcing of the products that they uh that they offer so yeah this is another reason why i was

confident that talon was a really good fit for Stitch. So the Talon, literally from day one, the core of their product has been open source. So you can go download Talon Open Studio today, and a tremendous number of people do that every month. And there are commercial modules and add-ons that you can use and they sell, and that's their business, which is very large. But the fact that we had committed ourselves to open source and that it's a really core part of our strategy, when I thought about the idea of potentially selling Stitch, I was concerned that that element of, that, that element of it, that, that they get it and that they value it and not, not, I didn't want it to be something that they just like tolerated. I wanted it to be something that they were like, cool, we see the business value of this and we're excited about continuing it because this is in their DNA and they've done this from day one, I think has been great. And they also, we've been able to learn a lot from them in terms of how you think about that as part of the strategy, how you can potentially increase engagement. They have this amazing thriving community where just a huge number of people are asking questions, answering each other's questions.

And it's a great feeder for the business. And it's also just an awesome thing for the world that all these people get to use this product, many of which, you know, might not be customers anyway. So yeah, it's a really, really big part of their business. So what does Stitch being acquired by Talent mean for Stitch customers? Yeah, so in the very short term, nothing has to change for Stitch users, you know, they continue to use the product. The things they liked about Stitch will remain the things they liked about Stitch. I'm really proud of the fact that 100% of our team got offers in the acquisition

and 100% of our team has stayed through. So the same folks that they liked working with on our support team are still there. And the thing that they should be excited about is that we're growing the team much faster than before. And so that means, you know, more engineers making the product great, more support folks, um, helping them out. Uh, and so there's, uh, the, the pace of investment in Stitch is, uh, is increasing, which I think will show up in more benefits for them over time. And the piece that I mentioned before around giving them the opportunity to solve more aspects of their data challenges with Stitch is something that I'm really excited

about and I think they will be too. So one last thing actually while you're here. I noticed that there was some changes on your website about pricing for Stitch and to do with overages and making it more, I suppose, predictable pricing. And what's that all about really? Sure. And this was in the works from well before

we were thinking about the Talon deal. So the way that Stitch pricing used to work was that there were two components to it. We had these subscription tiers that you would sign on to that gave you different things at different levels. But one of the biggest levers was data volume. And once if you went over the limits of your tier, we would basically charge you these incremental overage fees. And when we originally designed our pricing, the rationale behind those overages was that there are pretty big gaps in between these different tiers. So let's give you the ability to pay incremental amounts in between

those tiers. It did solve that problem, but it created a lot of uncertainty for our customers. And we got a lot of good feedback, I thought, from those customers that this was just hard for them to figure out. It was hard to budget for. It was hard to forecast. And it made it almost like Stitch solved all these problems for them, but he created this other job. And so we heard a lot of that feedback. We thought about it. And we think a better way to solve that smoothing between usage levels is just to create some more fine-grained plans in between those two. So what we've done is we rolled out new pricing that if you add up at various price points what you were paying with the overages plus the subscriptions before, it's about the same as the new subscriptions now. It's just there's a whole

lot more levels. So the average customer's bill is going to stay about the same. It's really just about having more predictability and having smaller jumps in between those plans and also just having a little more peace of mind that you know that you signed up for this and then your price is going to be exactly this. Okay, fantastic, fantastic. So how would somebody find out more about Stitch? Yeah, so the easiest thing to do is just go to stitchdata.com. There's a big button to sign up so you can just try out the product for no cost at all in an unlimited two-week trial. And Talon.com is their website, and we'll have more cohesion between those two things over time. But those are the two main websites.

And I also highly recommend following us on Twitter. It's Stitch underscore Data. And my coworker, Mark, is extremely snarky and sarcastic, and I think everybody will enjoy it. And on your blog recently, you wrote about your vision going forward and how you saw things in the industry. So we'll link to that on the show notes as well. Yeah, absolutely, yeah.

So stitchdata.com slash blog and I appreciate you linking to that specific post. And that's where we do a lot of our thinking out loud about it and you can also follow me on Twitter. I'm just at Jake Stein. Fantastic. Well, Jake, it's been excellent speaking to you thank you very much and uh yeah it's been great having you back on the show and best of luck with uh with the acquisition and working within the sort of talent world and uh yeah it's been great to catch up with you likewise mark always a pleasure i appreciate it Thank you.