[00:00 - 00:01]
Aditya: Welcome everyone to the first episode of our brand new podcast from Reo.dev. This is the DevGTM craft, and we'll be talking to developer GTM practitioners, pulling back the curtain on the smartest playbooks and dig into insights that make the developer GTM just so very unique.
Today we have our first guest on the first season. We have Orlando Nieves. Orlando leads revenue operations for Unstructured.io and lives in Denver, Colorado. Orlando, welcome.
Orlando: Hey Aditya, thanks for having me.
Aditya: Great, so let's get right into it, right? Tell us a little bit about what Unstructured.io does. I know you do a little bit around helping large organizations adopt LLMs, but yeah, would love to learn a little more about what Unstructured does.
Orlando: Yeah, yeah, you can think of Unstructured kind of just as an ETL platform specifically for Unstructured data types. So not only for LLMs, but for all data that's going to be feeding any Gen-AI project.
[00:01 - 00:02]
Orlando: So, you know, data being machine readable it's going to be critical for any AI initiative so that, models can act on your company's knowledge. We refer to our solution as ETL+, because the Unstructured platform takes care of not only moving that data from source to destination, but it takes care of everything ranging from maintaining connectors to transforming all your file types, third party enrichments, security and compliance, model updates, all of it. So you'll notice I just named a ton of different components involved in our solution.
And that's because we have something that we call the rat's nest and, the rat's nest is when companies try to build this data pre-processing solution in-house. It's wild, it's expensive, because it takes just a huge amount of engineering hours to build and maintain. It has all these components tangled into each other. So the whole point of Unstructured platform is to take care of that for you in a nice clean package that just plugs into your tech stack
[00:02 - 00:03]
Orlando: so your engineers can focus on building the part of your product that your users are actually gonna interact with.
Really, so the reason I'm on this podcast is Unstructured has a robust open source package that is typically a part of that rat's nest I mentioned. And so I really focus on how are we going to convert those open source users into enterprise customers.
Aditya: Awesome. That's very interesting. And I would imagine like the use cases for your end customers to work with this, run the spectrum from customer success to sales, to marketing, to a variety of end functions are probably going to consume data that's kind of transformed ETL plus by Unstructured. So it's possible even that developers are working on different things in different parts of the organization and making multiple rat's nests in different places across the company.
Orlando: Yeah, exactly. And I'm sure you've seen it. It's so easy to make a rat's nest and all these different nests existing in different parts of the companies and one thing breaks and suddenly the whole system just comes crashing down. So we aim to address that and get rid of that possibility.
[00:03 - 00:04]
Aditya: Yeah, and I think as startups are especially good at making rat's nests even when there isn't a need to do that.
Orlando: Exactly right. They always say with startups, you know, create things that aren't scalable, which good. Yeah, you start doing what you need to, to get your startup off the ground. But eventually you're going to have to formalize it because you are going to want to scale.
Aditya: Absolutely. Yeah. And you yourself, you come from a slightly different background from, what we typically see in SaaS and developer tools, don't you?
Orlando: I do yeah, I was a, I was a military intelligence officer for seven years, directly before coming to Unstructured. So it is my first private sector company. But yeah, it's been, I've been here for well over a year and it's been a great learning experience for sure.
Aditya: That's amazing. And what brought you to Unstructured and revenue operations in particular out of the military?
[00:04 - 00:05]
Orlando: Yeah, so in the later years of my time in the military, I was running a software development data science team. And that made me realize that I do have that passion for data. And I do have that passion that would carry over into industry. While I was also doing my more pure intelligence responsibilities, I noticed that my favorite thing to do was to operationalize intelligence. So it was essentially it was enabling that execution phase and allowing the front line to get after a target. Now that's directly translated over to what I do now at Unstructured. While I was initially standing up our internal data team, I had a very similar feeling. In rev ops and go to market engineering, you have all these responsibilities to include strategic global reporting, which is very important.
But my favorite part, the part that I find the most exciting is working with data and operationalizing it for the go-to-market team
[00:05 - 00:06]
Orlando: to get after what we would consider their targets, right? So marketing and sales, for example, enabling them to know who they need to go after to convert prospects into deals and when they need to do it.And the most important part of that is having actionable data to provide.
Aditya: That's amazing. That's an amazing insight that for revenue operations, I think we very often lose sight of the forest with the trees, working on connecting this tool to that platform, moving the data from here to there, when the reality is more about how we're creating usable, actionable data points and insights for the GTM team. That's amazing. Yeah, that's a very wonderful insight over there. And I think it's a good segue as well to talk about SaaS GTM team's favorite topic, which is the tech stack. So yeah, shed a little light on what's the set of tools and platforms that you're using across the Unstructured GTM process.
Orlando: Yeah, talking about tech stack and the startup with Go To Market that always leads to
[00:06 - 00:07]
Orlando: our rat's nest as well, so we had... We had our own for a little bit, but we've cleaned it up the past few months. Yeah, so we use HubSpot as our CRM. We have a few other tools integrated with that. So we do scoring with Keyplay. We use Sixth Sense for automated and manual prospecting and enrichment. And then we use ReoDev for tracking developer intent signals on our open source package, our GitHub, our docs.
Orlando: So once we have all this data that we have access to, we pipe it, including that Reo data, into our internal data warehouse, which feeds internal company level dashboards. And then we also pipe that data to our CRM to trigger the outbound email sequences and motions that marketing and sales engages with. Additionally with that, because we do get some contacts, whether from Reo or whether from Sixth Sense, we get some that only have a LinkedIn profile
[00:07 - 00:08]
Orlando: and we use Meet Alfred for automated outbound for those contacts specifically, so that we could have just automated outbound on LinkedIn.
Aditya: That's very interesting. So how many reps do you have enabled by this entire system?
Orlando: Yeah, right now we have... four total enterprise AEs that are enabled by the system and then a full marketing team as well.
Aditya: Oh, that's very nice. Right. So you mentioned Reo, right? And that a lot of the data that is enabling your entire GTM motion is coming out of the intent signals and developer activity being tracked by Reo. So let's talk about that a bit, right? So what is the input to Reo? What are you tracking? And where does that go? And what's the output from that, all those intent signals?
Orlando: Yeah, so we track our open source repos from GitHub. That tracks both full installations and the follow on telemetry pings. And then it tracks our one time Docker installs.
[00:08 - 00:09]
Orlando: In addition to that, it tracks our docs so... we have the ability to see not only what doc pages are getting visited, but also, for example, what commands are being copied from those docs. Additionally, we could see who's visiting our blog posts, who's visiting our website. We are able to pull in community Slack interactions, GitHub stars, forks, pull requests. We could see LinkedIn activity, all of that's getting compiled. And then we push it into our CRM in our data warehouse so it can be actioned on by... by our go-to-market team.
We primarily focus on the developers that are in that developer funnel of being around building and deployed, with at least a dozen separate developers, a strong customer fit score. And then on top of that, in our CRM, we use Keyplay to run scoring on those accounts as well so that we're able to be very targeted with our efforts.
Aditya: Got it. So Keyplay also does like a opportunity sizing
[00:09 - 00:10]
Aditya: and readiness to go kind of scoring on top of the intent signals that you're getting from Reo.
Orlando: Yes, yeah, and it's also able to create lookalikes for that as well.
Aditya: Okay, that's very interesting. So yeah, that's creating a good amount of value add on top of the intent signals. And you are probably, when you're looking at the sum total of signals from Reo, you're looking both at the total sum of signals, that is people who are working on your open source project, as well as looking at the docs, but you're also probably looking at increasing trends of activity, that is companies say that are... very quickly ramping up in their activity probably gets more importance than someone that's kind of in a steady state.
Orlando: Exactly. We focus on the companies that have layered signals. So like not necessarily one type of signal, but different signals from different individuals layered on top of each other. We also don't want to be too late
[00:10 - 00:11]
Orlando: when we reach out to these companies because we don't want them to already be committed to that rat's nest or to have the mentality of, we built it, right? We're proud of it. And we want to keep it.
Orlando: We want to get to that sweet spot of reaching out at the right time so it's pointing out their issue but they're not so invested in their homegrown solution that they don't want to come away from it.
Aditya: That's such a big insight, right? Like if you catch them too early, they're probably not going to be fully aware of the problem. And they're like, this doesn't sound like such a big thing that I need a separate solution for it. And then if you catch them too late, like we built it, now what? I'm not going to unwind the rat's nest just because you came up with a new solution. Yeah, that's pretty interesting. Yeah, so there's going to be a lot of then rapid decision-making and real-time analysis. Are you deploying any AI or any AI agents on top of this?
[00:11 - 00:12]
Orlando: Yeah, so actually the way that we measure when we would want to reach out to a customer, we actually use some AI agents to continuously analyze our current pipeline so... essentially what we're looking at is we're looking at the companies that we're converting, but we're specifically analyzing where they were in their developer funnel. What signals were they exhibiting when they converted? So then we try to replicate that with the timing of when we reach out to companies that look like them.
Orlando: So essentially we're using that historic and current pipeline data to drive our future decisions on what the threshold is. But also when, right? The when is extremely important with that And I will also say like this, type of data that we're getting is also how you just discover the ICPs for your paid product, right? You really start to see the delineation and trends between who would be best postured for an enterprise solution and who's not, right? Just because you have a company using your open source
[00:12 - 00:13]
Orlando: doesn't make them worth pursuit and that's completely fine, right? You have to be targeted with who you're going after.
Aditya: That's true. Someone may just be happy making a very small rat's nest and that's enough and that's all they need. Right? Does that extrapolate well though, especially when you're talking with large companies and kind of like an enterprise GTM that because everyone says that every enterprise is going to be different and it's a whole separate process every time. So do you find that tracking from the timing that you had with previous closed one that works well applying it to new potential deals and saying, okay, this worked. This timing worked for this company in healthcare, we can apply it to another company in healthcare and that tracks well. Does that track well?
Orlando: Yeah, I'd say track well, I mean, I think the... the use cases that we come across, especially in the AI space, because it's growing so fast. And you have literally every single industry out there, right? Like every fortune 1000 is trying to build their own AI solution.
[00:13 - 00:14]
Orlando: All the use cases are just so unique and different from each other. Um, but you have those, similarities, right? So I think going based off the timing of closed ones is good enough to be able to get you in the door. And then it's at that point, you just have to conduct discovery. and figure out, how am gonna actually tailor this solution to feed this specific use case.
Aditya: That's amazing, absolutely. And is there any particular use case that kind of predominates for you in terms of like say, handling the rate of customer success or manufacturing or anything that is a dominant use case for you guys?
Orlando: Yeah, I mean, we see some industries that are more common, like the finance industry is a big one. But at the same time, I think it's also just as diverse as... just the total addressable market that's out there conducting their AI initiatives.
[00:14 - 00:15]
Orlando: Because you have some companies that have been around for over a hundred years, right. And they have all this logs data and it's extremely complex and they want to have like this internal chat bot but they also want to be able to make new product recommendations based off of a generative AI model that they have running in-house. So it's really is just so diverse. But fortunately for Unstructured, regardless of what the use case is, you need that data that you have just historically sitting around in your company in all your different locations. You need that in a format that is machine readable. And then once it's machine readable, you can feed that really into whichever project that you need it to be.
Aditya: That's really cool, right? I think you're sitting in a beautiful product market fit at the intersection of historically really messy data that no one tried to clean up and standardize. And the need for the next big shiny thing,
[00:15 - 00:16]
Aditya: Which is AI and LLMs and just working with that data and making it readable. That's amazing.
So just coming back to the entire GTM process, right? We talked about how you're looking at intent signals from Reo, you're piping this into your CRM and your data warehouse. You're doing a bunch of outreach. Can you shine a light on how exactly you're doing that outreach? Having looked at the intent signals, found the right time and you're going after, since you mentioned finance, like let's say JP Morgan, right? A company like that. Company like that, How do you actually do the GTM process from there?
Orlando: Yeah, our play really is to get our data in front of the prospect, right? For some reason, I've noticed people are afraid to place the data in front of the user, but they shouldn't be, right? It's very powerful. It informs buyers who may not even be aware that their engineers are using your open source or the work that goes into trying to make an enterprise solution and informs them of what's happening.
[00:16 - 00:17]
Orlando: There's primarily two approaches that we take when we outbound these prospects. We've noticed that LinkedIn right now is more effective for being able to get on the phone and get people's attentions and emails. And so when we're outbounding individuals on LinkedIn, we're sending them a microsite landing page that's custom made for them. We're saying, "Hey, thank you for using Unstructured open source... ...You're a very valued customer... ...Here's your landing page. ...That's going to show you your usage." And just like a once over on what's going on with your... with the way that you're using our product. Right. It's that hook, right? Cause it grabs that interest of what are you talking about? I use this product?
Orlando: And so while we're nurturing the devs with a useful content based off of their activity, we're going to the buyers on LinkedIn. And we're pointing out to them with these landing pages that their teams are spending all these hours to try to build that homegrown solution. We've noticed that it's extremely effective to communicate it in estimated money spent, right?
[00:17 - 00:18]
Orlando: So, "Hey, based on your signals, here's how many hours we assess your engineers are putting in to try to build your own rat's nest. Therefore we've estimated that you're spending this much money in engineering hours." And they come back and they say, "I didn't even know that this was a problem to begin with. Let's chat." Right? So that gets them on that initial call.
The second approach that we have is once we get them on a call for that initial discovery, we present a dashboard that digs further into their usage patterns to further communicate how much time their team is spending into building their homegrown system. So it's really that two-pronged approach that we've noticed is effective to communicating to buyers.
Orlando: And again, a lot of the reaction that we get is just they didn't realize that it was issue to begin with, right? And so we point out the issue and then we present the solution effectively.
Aditya: Wow, that's amazing. You're actually showing them the same Intel, the actionable Intel that you use to choose them for outbound to say this is why you need a solution.
[00:18 - 00:19]
Aditya: Doesn't it get creepy sometimes for your buyers that another company is able to see exactly what their devs are doing and how many documentation views they're making and how many Docker installs they've done Does that ever come up or are they just more appreciative of just the scale of the problem that's been brought to light.
Orlando: Yeah, doesn't really honestly, it doesn't really come up, which is why I really just encourage people to just go for it and present that data. I will say that too, though, there is due diligence that we do on our part to prevent any sort of issues arising, right? Like we put in our documentation, we put in our GitHub saying, "Hey, this is the telemetry, the way that we're going to measure your usage through some sort of telemetry tracking. In addition to that, we make it very transparent and how you can opt out. So that if you don't want your usage tracked then you're welcome to opt out of it.
[00:19 - 00:20]
Aditya: Yeah, because yeah, that's I think a very typical concern that all of us GTM people have in the back of our minds that we know we're getting a lot of signals, we've seen that, but when we talk to buyers, we don't want to pull aside the curtain and break that suspension of disbelief on exactly how we understand a lot of how their developer tools work and how their developers are using the tools, right? And I guess, yeah, especially on your telemetry, I doubt a lot of people are going to opt out because it is a commercial open source in that sense. There's an open source and there's a paid product and back of the minds of all the devs know that one day they may need to leverage the paid solution. That's fantastic. So you have a custom landing page and a custom dashboard showing your buyers exactly how big the problem is, right? That's fantastic.
And that basically completes the loop. If we're going from tracking docs and the GitHub, pulling it into your CRM and your dashboards, making the choices, especially on both the right company and the right time
[00:20 - 00:21]
Aditya: based on your past closed-won deals, and then using that data to make the problem really tangible for the buyers. And yeah, getting deals closed-won that way.
Aditya: Any particular standout deals that even if you don't want to mention the customer, but... that was unusual, looked like a long shot and you made it through and was something that was a proud win for the team.
Orlando: Honestly, I mean, I think we're landing pretty well with our intended use cases within the companies that we're trying to get after. I think it's just naturally being a startup, right, and... presenting solutions to enterprises. There is that skepticism from the enterprise on, "Hey are you ready to be able to deliver everything that you need to fit our standard? Right." As you know, the enterprise deal pipeline is extremely slow.
[00:21 - 00:22]
Orlando: You have to be very patient. It's very long. But yeah, I think because of the fact that we're just very diligent on also providing professional services and integrating our solutions architects with the enterprises that we are able to close those deals and to service these large companies.
Orlando: So I wouldn't say necessarily a surprise in who we've closed one. I think just in general, the surprise was looking at the open source usage so like... 73 % of the Fortune 1000 have touched our open source, right? And so... Just seeing like that wide, huge audience that interacts with it and that's trying these initiatives, that was the initial surprise. Now it's just a ton of different marathons going on in different directions at the same time in our company to try to reach out to those companies and to be able to service them.
[00:22 - 00:23]
Aditya: Yeah, that's a real big myth buster, I think for all of us in developer marketing, right? Like I think we've always thought of like open source as a separate tribe or the separate universe that exists completely separately from commercial software, especially enterprise grade commercial software. But what you're telling me when like 70+ percent of Fortune 1000 are using open source, that's not the case. It is very pervasive across the entire developer universe in that sense.
Orlando: Yeah. And I think another part too is this is AI, right? Has not been around for a long time. All these companies are trying these new initiatives.
Orlando: So initially when unstructured started, it was ETL for LLMs, right? And then it, as we started conducting these user interviews and started understanding the use cases from all these different companies, we started to see these use cases that we just had not thought of that were not just LLMs. And so we're learning alongside the customers as they learn about what they need from the AI, from their AI initiatives.
[00:23 - 00:24]
Aditya: That's amazing. Yeah. I think I come across on my ecosystem and communities, everyone's asking for the same thing that, I have this huge mass of data and how do I deal with it and how do I plug it into an AI? Should I just notebook LM the entire thing, but... I think I should just keep dropping Unstructured.io's URL into all these chats and all these Slack communities whenever I get the chance. Wonderful. I think that wraps up what we want to talk about from your GTM motion. We're going to do something fun, something interesting and what I hope will be our favorite part of the DevGTM craft, we're going to play a bit of a game. And I let you choose this beforehand. We had two options. We had "Explain like I'm five", and we had "Whose pitch is it anyway." So we're to do explain like I'm five.
[00:24 - 00:25]
Aditya: And what we're going to do for everyone who's going to be watching this is Orlando is going to be given a marketing or GTM concept, right? Like GTM engineering, for example. And he's going to have to explain this like he was talking to a five-year-old, right? We GTM folks can barely explain it to oursleves. So it's going be very interesting to see how we do it when talking to an imaginary five-year-old, right? So Orlando, I have a few options for you. I'll give you two at a time and if neither of them make sense, we can bounce to the next two, right? So the first two I have, the first is A-B testing and I have, the second one I have is user generated content, right? I have a hint for both, but tell me if either of these work and I can give you the hint as well.
Orlando: Okay, I think A-B testing I'd be able to do. Can you give me a hint for the user generated content?
Aditya: Sure, so for user generated content, try and explain it like if you're talking to 5-year-old,
[00:25 - 00:26]
Aditya: explain it like when your friends make up their own stories about your toys and you share it with everyone and then they share it with everyone and then they share it with everyone and everyone's making up their own stories about toys. So that's like a hint, well almost a story, almost a lie-fiber, yeah.
Orlando: Okay. so I guess I'll start with the first one, A-B testing.
Aditya: Okay. Yeah. If you want a countdown or you can start whenever you're ready, take as much time as you want to.
Orlando: Okay, so five-year-old, A-B testing. You have two different types of cookies that you think people are gonna love, right? You wanna be able to sell these cookies, but you really wanna know which of these cookies are gonna be better and who's gonna like enjoy, or which of these cookies is gonna make you more money and bring people more joy. So what you're doing, is you're putting both of these cookies out there and you're selling them to your friends.
[00:26 - 00:27]
Orlando: You're going to your teacher, you're going to your friends, and you're giving these types of cookies to different people and you're measuring their reactions. You're watching them and you're saying, did you smile? Did you frown when you had my cookie? And you're measuring the results. And essentially through that, with some time, you're going to be able to see which of the cookies that you made is going to be better for you to sell and be the best cookie.
Aditya: That's perfect. I have to applaud that. That's one good cookie of an explanation. I think my three-year-old would get that too. But yeah, good one, good attempt at that, yeah.
Orlando: Thanks.
Aditya: That's it. Any last thoughts? What's been really interesting about selling a developer tool and something that's maybe been an interesting insight over the last year in this space?
Orlando: Yeah, it's especially in this space and like in the AI space things are just changing so much so it's extremely important
[00:27 - 00:28]
Orlando: that if you're selling a developer tool that you just keep yourself up to date on the latest news, the latest product releases, like talk to the engineers that you're selling to, not only to try to sell your product, but to understand what they need because then they give you insights that you didn't even realize that you had to incorporate into your strategy. So, especially when everything's just so rapidly changing and growing as an industry.
Aditya: Very nice. Yeah, I think that's so true, right? And like we all talk about it, like talk to your customers, talk to your audience. But I think with developer tools, have to just ratchet that up, ratchet the dial of that all the way up to 100 and then a little more till the engine is getting over hot because there's just so much happening and there's just pretty much any developer, every developer is at the cutting edge of their space, or at least reading up on the cutting edge of their space and we need to be able to understand where they live.
Orlando: Yeah, yeah, exactly. it's like, at this point, and especially me transitioning from a completely different career,
[00:28 - 00:29]
Orlando: there are so many books and podcasts and it's impossible to get through all of them. So I just really rely on all my peers and my coworkers to make their best recommendations because I also just lean on them to keep me up to date as well.
Aditya: Yeah, absolutely. I've been a developer marketer three different companies in the last five companies that I've worked and every time it just keeps me humble, right? Like the amount that moves in every quarter, something's moving and something's different. And it's such an interesting space to be working with developer tools. So yes, with that, I think we'll wrap this up. Orlando, it was an absolute joy and pleasure to to talk to you about Unstructured.io and Intent Signals and Cookies and so many other things and Military Intelligence as well. Great to have you here.
Orlando: It was great to be here. I really appreciate you having me on the show. And yeah, I'm really looking forward to seeing everything that Reo releases and to be continuing to making actionable data actually usable
[00:29 - 29:20]
Orlando: by the Go-To-Market team. So thanks again.
Aditya: Thanks so much Orlando. Talk to you soon. Bye bye.
Orlando: Alright, bye.
