Build an ugly MVP and watch your startup win with Taylor Halliday

In this episode of Delivered, you can learn why speed and customer feedback matter more for startup success than building the perfect product.

We sat down with Taylor Halliday, co-founder and CEO of Ravenna, and a serial entrepreneur. Under his leadership, Ravenna raised $15 million in funding and gained recognition for its innovative Slack-native platform that streamlines support for IT, HR, and revenue operations teams. Before Ravenna, Taylor founded multiple companies, served as Director of AI Engineering at Zapier, and completed a Y Combinator fellowship.

Taylor believes a startup’s only real advantage is speed. Instead of building in stealth, he advocates shipping an early MVP, learning directly from customers, and proving value fast. In a candid conversation with our host, Taylor shares hard-earned lessons that every early-stage team building in AI, SaaS, or operations tech should hear.

Key learnings:

• Understand the advantages of a lean startup over a stealth startup strategy
• Learn how Taylor raised $15M and built Ravenna through rapid iteration and customer focus
• Discover how to structure your startup to move faster and stay focused
• Explore how Ravenna is reinventing internal support with AI

Thanks for tuning in! If you enjoyed this episode, hit that follow button so you don’t miss our next delivery.

Delivered newsletter
Delivered episodes are recorded live. Subscribe to our newsletter to stay informed about upcoming live events. When you join a live event, you can ask our guests questions and get instant answers, access exclusive content, and participate in giveaways.

About Infinum
Delivered is brought to you by a leading digital product agency, Infinum. We've been consulting, workshopping, and delivering advanced digital solutions since 2005.

Let's stay connected!
LinkedIn
YouTube

Transcript

Taylor Halliday is a tech entrepreneur and engineering leader based in Seattle with a background that includes roles such as co-founder at Flickr, director of engineering at Zapier, and founder of Mesh Studio. Taylor has a proven track record of building successful startups and leading product development teams. Currently, Taylor serve as the co-founder and CEO at Ravenna, which we're going to talk more about in today's conversation. Welcome Taylor. How are you? Pretty good.

How are you

Doing? Where does this podcast bind you? Where are you about?

I'm inhaling from Seattle, Washington in the United States.

So Seattle, Washington. That's not on the, are you on the east Coast then or are you on the west coast?

I'm on the west coast. I'm on the tech side of the coast, if you will. Yeah, yeah, okay. Yeah, got it. Where Microsoft was founded, Amazon is founded.

That's right. Yeah, it was the Washington that threw me out, threw me, threw me off. I was like, oh, Washington dc That's the city. Of

Course, I know. It's kind of confusing. Was Washington and one the east coast and there's a Washington and the West Coast. It's a, yeah.

Yeah. Okay. European figuring out topology of America. Sorry about that. So let's get into it. So for those who don't know Taylor and Ravenna and your story so far, can you tell us a little bit about yourself and how you got here?

Absolutely. Sure. So bet about myself and I'll start before Ravenna. Prior to Ravenna was running new products and then most recently AI engineering at Zapier for about good four and a half, few years. Zapier, pretty well-known integration platform, lots of good stuff there. Learnings that informed some of the decisions and why we ended up doing Ravenna, having to get into that a little later. And then prior to that, actually founded several companies who is my now current co-founder Kevin Coleman between 2020 and 2015. Went through Y Combinator at one point with him. Some good companies, some bad stuff, but yeah, lots of good experiences all built into there. And then prior to 2015,

I was like a joke, just professional nerd.

You said you, you started several companies. How many companies have you started?

Well, it's funny, I don't know how exactly people define what a company is. If it lasted for two weeks, is that a company, if it lasted for a year, there's a lot of shots, put it that way. And some actually did turn up decently well. But yeah, between 2020 and 2015, been a lot of time grinding, a lot of time, trying a lot of different things. And then actually we were talking a bit before the show ended up, which I think happens more often than not in the ashes of a failed product company, your back's against the wall and you kind of rolled into a service-based company. That happened for me and Kevin kind of unintentionally, frankly, that was not an intentional thing, but just happened to the way that we could try to figure out money when we just were running out of it and then ended up having growing actually business there. Ended up growing this consultancy, it took about 20 folks and yeah, lots of good lessons, lots of tiers, lots of happiness, everything

Between what type of a service business was it? Did you have a specialty or focus?

In large part it had to do with Kubernetes Cloud Consulting, but that all said again, your back's against the wall. You're trying to make anything work. And so while I would always let myself customers know I was known for the X, Y, Z thing, I very much all admit it that we were happy to take on any technical problem under the sun. I've worked in some pretty interesting languages for that period of time. Some pretty interesting code bases, stacks, what have you. But you got to do what you got to do sometimes to make it work. And it doesn't matter if someone approaches you asking you if you can help them with their Rx, Java and try IT app when you're a cloud consultancy and you're still just trying to make ends meet. So you just got to do what you got to do. But like I said, lots of learnings there.

Very cool. So now you are

Operating as co-founder and CEO of Ravenna. So tell us about Ravenna. Why did you start it? What's the problem? You're solving all those stuff.

So right off the bat, what is Ravenna? Ravenna is building a modern AI enabled help desk. Okay. So the canonical example of a help desk frankly is the IT help desk. If you've been at a company larger than a hundred employees, 150 50 employees,
There's this notion of this thing called a help desk effectively. And the best way to think about that, the way I talk to my mom about it is it's nearly the same thing as customer service. Everyone's familiar with customer service, right? Software, it's like internal customer service. You got it. You got it. Exactly. The one big difference is that the end user is not like an external customer, it's a colleague, it's an employee, right? That is in large part the best to think about this class of software. There's a lot of similarities there, but it's not just the IT help desk. As a company grows, they bring on more and more frankly just people. And with more people you have just a lot more just inner inter company issues that come up, coordination. And there's a need for basically more and more operations groups. And so typically the next group to need a help desk is human resources, the human resources help desk, and then frankly speaking, at certain size and scale engineering, marketing, revenue operations. Our customer base is actually across all the vertical that I just said. We try to hone in for the most part on it. That's our biggest persona after that is hr. And then after that it's niche Revenue ops is making a run for the third spot right now.

Okay. So what's the ideal customer for Ravenna? Is there any typical organization that gets the most out of it or is it for every business over a certain size or

Totally. Totally. So the ICP we're working within is about a hundred to 1500 employees to two B SaaS. Okay. And a couple of reasons for that. The range, the floor being 100 is, well I mentioned it before actually, you just don't have the need for a help desk if you're below a hundred ish, it changes depending on the company, but a hundred ish employees, if you're 10 people, you just don't need to help desk internally. You get to talk to each other. You're like, Hey, I need this, I need that. It's kind when triaging coordination and really delivering great service to the broader organization, it starts to become difficult. That's when you start to see the need for the software arise. And then 1500 employees is pretty much the cap we're at right now, 2000. Again, these are fuzzy numbers but roughly around there. And the reason why we are just sitting at that point is because upstream from that employee headcount, you're bringing on some pretty serious enterprise classes, software Workday or a DP and those are just some deeper integrations and we just have had time to bite off at this point.
So that's kind of what declares a batting range. And then the B2B SaaS, it's funny, I always kind joke that we're doing the same thing a lot of other venture funding companies do. We venture funding companies, a other venture funding companies and just trade money around. So that's very much we're AP to right now. But the thing that's always fascinated me about this market, or sad me frankly, is the need for the software is truly just the correlated to employee headcount straight up. It does not care if you're commercial, it does not care if you're tech, it does not care if you're big box, if you're running a company and you just have over a certain threshold of folks, you need to get basically software in there to help you. Again, I said coordinate triage, but at the end of the day, it's really about just service.
If I'm an operations person, if I'm it, the best customers we work with, that's the way they think about their job effectively is providing great service to the rest of the organization and that creating great efficiency inside of there. And I dunno about you. I don't know what I've thought of. I guess I look back at Zapier and think the IT person I thought is the IT person. They view themselves as being an enabler, again of service and efficiency at the org, the really key point that I think through and the software allows 'em to create that efficiency allows them to create that amazing service.

That's what we're here for. Gotcha. Yeah, no,

We've been around 20 people for these years, but then we joined Infinum, which been much larger organization and now we're about 400 people and I'm getting used to this workflow myself. It's like, oh, there's a person for that that helps me with this or this, you need anything? It's a software installed or some, what's the policy for this? Or normally everybody came to me, but now there's people somewhere on Slack that knows these things. So I can see the value.

I'll say go ahead. I was going to mention too, I was just going to mention a little more, but kind of like a why now of the company also too, but so the help desk thing is nothing. It didn't get inventive yet last year, right? It's been around for like 30, 40 years. And so obviously some of the investments in AI allow you to do a lot more interesting things about meeting people where they are. There's two priors to this equation. There's the requester. Let's say you request something and I'm the person who's IT or HR operations leveraging a lot in modern ai. You can cut down a lot of what was the back and forth. A lot of what was basically every time you ask for a new piece of software, I know I need to ask about your manager, which region you're in, these different things kind of stuff. You can cut down a lot of that. And then finally too, I would say that another big thing that we are focusing on is at this company, everything works backwards from customers. This was from the early days of us just interviewing a lot of folks before we even had a pain cut before we had software. Frankly we were doing a lot of interviewing around,

There's a lot of pain around people saying, look, I was asking people if I were to build software in the space, what would make you switch?

I heard a lot of pain around Slack integration, team chat teams integration. And I remember being like, it's 2024, you can't tell me that these other incumbents don't have the Slack app. Whatcha talking about? It's not so much that there's a gap of not having the app, it's that the focus of these integrations for all the larger incumbents. It focuses on meeting you where you are and then have you click a button to then pull you up to a third party application. It's not where you work. You have to opt into their OAuth, into their whatever kind of stuff. There's a lot of paper cuts along the way end up just DMing the person. I'm just trying to say that all the user experience have been geared to get you out of where you work and increasingly where you work is a team chat based application. So we're really focused on doing everything I talked about, but that all said, delivering it in a way that is completely just Slack native teams native that you can get the service where you need work and not have to ever leave.

Gotcha. So I mean looking at your mission for the company, are you going to be almost like the brain of the company or the concierge that has all the company information available to provide it to the employees when they need it, where they need it? Is that what you're going for or what's the mission you're on?

The mission we're on, I'll say

Before is to help the folks who are

In these operations roles provide incredible service to the rest of the organization. That's kind of really the mission effectively. Okay. We do that through automating a lot of things again that people are normally involved in. You mentioned the brain organization. There's a couple things that I think

That, I dunno if you want to use the word obvious, but

Things that probably make a lot of sense once you're integrated in order to provide that great service from a software standpoint, they expect us to be integrated with their HR systems. With other MDM systems, you start to build quite interesting graph of knowledge around the employee, around the policies, playbooks, what have you. So I think that's what you're going towards. Yes, that's very much, I think that that is a very interesting thing to us about what you can do and once you have basically all that knowledge, that graph, what have you, but I wouldn't say the mission is to become the brains per se. I'll keep on going back to it. The mission is to help enable the folks who are trying to create just great efficiencies across the organization and just get you as an employee, make sure that you can just easily get, at the end of the day, it's just you're asking for help. That's really what it is. And just trying to get you that help in the most efficient way that minimizes the amount of redundant rot things that we do as humans filling out the same forms, asking the same questions.

Gotcha. Right, thanks. So you mentioned before the call that you've been through the fundraising and all of those things. Can you tell us a little bit about that journey and what you're going to use the means for?

Yeah, sure. So a bit of just run down the company. Been doing this for about 15 months now. We started, me started in July 15th of last year. And at that time we raised $3 million from Coastal Ventures, another group of angels in Seattle and also just had some great other investors on board. The founders of Zapier, Guillermo, CEO, Guillermo from Versa, Brazil, a lot actually. Great folks. And so super lucky to be having to have all them on board. And so that really kicked off the company and kind got us moving. And then we ended up doing another raise for $12 million about four and a half, five months later with Madrona Ventures, who is a extremely prominent venture capital team over here in Seattle, Washington.

Wow. Alright. And how does the team look like now then? You're only 15 months in and already raised a bunch of money and how are you using that?

Are you using it? We're using it to try build this thing as fast as possible. We're using it to try to get as much customers as much value as fast as possible. I mean everything ladders up to that I guess. So anyways,

That's kind of high level how are we using it? So company started midway through 2024,

The rest of 2024 was kind of three things. I feel like you just kind of do it that stage. It's like build talk customers, recruit b boom. I mean that's pretty much what you're doing, right? You're trying to do as fast as you can to build, I'm always my ethos and stuff like building the minimal, the MVP, the minimal thing, whatever you want say kind of stuff. And getting the ugliest, smallest thing out the door that provides a modicum of value to your customer. That's the way I work and I'm very big on getting there as fast as possible. And the only way you do that is you spend a lot of time talking to customers. You spend a lot of time building and you create a very tight feedback loop as best as you can kind of stuff. And then get out, in my opinion, the ugliest duckling that provides the tiniest amount of value kind of stuff. And this is in contrast, I know folks, there's a lot of things that everyone has an opinion. I'm not trying to say mine's right or wrong compared to others, but I know a lot of friends even that build stealth startups for a long time and go into a hole.
Maybe it's just a skill. I don't don't know what it is. It's just not how I know how to build products. I have this

Core belief that not that smart. I think none of us are that smart but not that smart. And I

Think that what I worry about with even I learned, lessons learned, when you go and you build into a whole or a steal area, when you don't have that feedback loop and you're building you, I think you're not setting yourself up for success because you're just creating a bigger delay between what the is and getting in front of customers. And one of the cool things about customers in the market, the market is the most shrewd thing in the planet. It's awesome. It's like the shrews mechanism out there. It does not care about it. Anything about you or blah blah, blah, blah, blah. The only thing that cares about is whether or not you're providing value a hundred percent. And so I guess when I think about sitting here and building stuff, which is, don't get me wrong, I have a lot of fun sitting here and building stuff. I can do it all day long, but the amount of time you delay from when you start building to meeting that shrewd check, that opinion that it will check your reality kind of stuff. I think you're just delaying longer potentially. You're worried about just bullshitting yourself. I don't know. I say this and I know Steve Jobs went into a hole for however many years and built the iPhone now it's the greatest product of all time. Good for him. I don't know how to do that at.
I very much am focused on just get opinions as fast as possible. Don't do the goal. I'm going to build for two years and not tell anyone about it for open house stuff. And that's how we very much have done this. And so anyways, I kind of went off on a tangent there. So 20, 24, just try to just build as fast as possible, get this thing in front. Customers be like, oh, it's a toy, it's toy, it's a toy, it's a toy until it's not. And so I'd probably say around January, February, my goal was to actually onboard people during January of 2025. That was aspirational. That didn't work. We get it out there, crash and burn, but we keep on doing this thing kind of boom, boom, boom. And then I think we were lucky in the sense that we had some great design partners, folks who were willing to just give us advice along the way who just had fun.
Hey guys, here's the thing we're building again, the toy, they're poke at it. That's cute, that's cute. They did these things kind of stuff in, I come back like, okay, cool, we'll come back. So did that motion until we still did that motion to a degree, but that was going on probably until March or so, March, April started to actually be like, okay, wait a minute, some these folks who were giving us advice are actually starting to use this thing in a real way. That's cool. They're actually starting to run parts of their business on this thing. I can't believe they're doing that. What are they thinking? Like, oh my gosh. But it's awesome, it's awesome, it's stressful, it's everything. And then I'd probably say around May-ish juneish, just being like, okay, I think most of the people we've been working with at that point, we've known for a while it's safe space, but I think we might be ready to go out there and try to find people we've never met before and actually just expect to turn something around and sale, if you will. So that around late spring and yeah, was able to turn that corner. I'll remember the first time I hopped on the phone with somebody for a demo call

At the end of, they're like, cool, we'll take it. I'm like, are you sure?

So was that your first customer and what was it? I was curious when you talked about building fast and the minimum, the MVP and what was the first thing that someone saw value in that they were willing to pay for? You remember that?

Yeah, it's actually really funny. So we ended up getting our first customer during the middle of that period. I told you about February when we were working with some big design partners. So Zapier was a design partner, a home base design partner. I think there's two big ones at that time kind of stuff. And we're working with pretty large teams. These are sizable organizations. So trying to bite off a lot. I think it was through some LinkedIn posts I had done or something like that. We made a little bit of noise, but not a ton. We had Josh, it was an IT manager for a venture capital firm called Air Tree that's

Over in Sydney, Australia. Maybe a quick pause

When you're founder, you get a lot of people reaching out to you trying to do stuff and people are trying to get ahold of you. Not going to lie. I kind of thought this was a way for someone to reach out from the firm or something like that kind of thing. And no, Josh was like, look man, I saw what you're doing. I saw, I think it was a LinkedIn post and I just really liked the vision, what you're talking about. Yeah, the software in the space kind of sucks. Looks like you guys have the ability to, I don't know, it is one those things that is really cool because he, I get it, it's not there, but I'm looking for just something that's like, I don't know, a breath of fresh air into this market. Because one thing I glossed over is when we even were approaching the market, it became very apparent that a lot of the longstanding incumbents, people who are, I don't want say this nicely. A lot of the software is not very inspiring in this market. Put it that way. Okay. No, yeah, I'll say it like that. How about that? Okay.

And

Very much, I think one of the things that Kevin and I got interested about was like, I'll say this too. There was no notion of this space. There's no figma of this space. There's nothing that I could find that it's all

Ugly, slow and yeah,

Well, if we're being blunt,

Necessary

Evil. Yeah. I mean, yeah, exactly. And so it's a couple funny because we talked about lessons learned from building and stuff, and

This

Might not just be me because the first thing I thought about was like, wow, is there space for a notion? Is there space for a figma of this category? And the software romantic to me, like the builder of romantic, I wish I want that to be true kind of stuff that people will rip and replace. There's software because I built such a more beautiful kind of stuff, beautiful new solution. I don't let myself believe that's the edge. I think you got to find something deeper than you have pretty software. You know what I'm saying?

Yeah. I mean it used to be this thing that in that case, it needs to be 10 times better than the competitor you're replacing. Otherwise the switching costs is going to be too high. Right, exactly. So you need to find a differentiator other ways.

Exactly. And I say that and I don't know if the linear people agree with that fully seem to go around and do a pretty good job. I
Mean, I guess, I don't know. I've never been pitched by them. I don't know what their pitch is in particular, but it is very much apparent like, wow, you have prettier project management software. Beautiful. And it's a joy to use, want that same feeling, but didn't want that to be the only thing wanted to be. In fact, actually we don't really lead with that today at all. It very much we tried to lead some deeper pain points around automation, around insights, analytics around one thing we chatted about too, a team chat kind of stuff, how you appear. That's a big one. It's something that is just something I didn't expect going into this. And it's interesting and it makes sense how big of a deal it is about. We talked about before, the whole point is delivering service to your colleagues. People have this, there's this funny discussion that goes on nowadays when people are like RTO return to work since the pandemic. What's changed since the pandemic? I'm not really here fight that discussion. What I think is irrevocably changed is ever since the pandemic companies are just operating more increasingly on team chat based applications, that's just where work happens regardless of where you're sitting physically. And that onslaught, I think just basically brings a different paradigm about how you want to get service and also how to deliver service to people.

Very cool. So I was curious. I mean, time just flies here, but I was curious to understand, you mentioned that you're a builder yourself engineer. Do you still write code?

Yes.

Oh, okay. Very cool.

Although a co-founder sometimes yells at me for doing it,

But spend your time on something more.

I get it. I think there's a lot of things being a founder, like, Hey, you should do this, you should do that. I should. There's some selfish personal enjoyment. I'm not going to lie. Just not going to lie. I get a kick out of you got to keep this fun, by the way too. And

I mean your company, right? You can

Do it if

You want to. Yeah,

No, I know, but I can't run around just saying it that all for everything. But I

Wanted to ask you, going from being writing code to being a leader of a company, an organization, how did you, I mean, it sounds like you're still keeping a foot in the coding still, but how that transition been for you? Was that easy to go from IC, if you will, to leader or entrepreneur?

No, no, not at all. No, it's

Tough. I mean, we're all figuring it out. Okay though. So I mean a couple of things, and I don't know what the right or wrong way to do things are, I guess, per se sometimes. And I read just made things on hacker news and blogs, and I try to do as best I can, but at the end of the day,
Here's what works for us so far. So we got 20 people. It's a fully flat organization. We're all locking arms. There's no VP or blah, blah, blah, blah, blah. I mean, sure people have functional titles, what have you, but we're locking arms and we're all in this together. I really want to create a massive sense of responsibility agency, if you will, and urgency. And the way we're doing that right now is it's a flat organization and I depend on everyone else across the company just as much as I expect 'em to depend on me. And so that's the way we're doing it right now.
I have brand teams at Zapier, so it's not quite fully IC kind of thing or full IC pass and definitely had some managerial experience there. But yeah, no managing teams, it's a thing. And I think a lot about how do we as a team move faster? How do we become more efficient? I also just have this kind ethos thing around, I think it is common, but as a startup, you have every disadvantage becomes a large incumbent you could think of, right? You don't have the money, you don't have the customer, you don't have the go-to-market, you don't, I don't know anything. The only thing you do have, frankly speaking, is just the absolute ability to just move quicker. I just don't move quicker than, I don't care how much money a company throws at a large effort from being on the other side of the fence and just seeing, I just don't care if they throw a gazillion dollars at something, they just cannot move DNA wise kind of stuff.
It just won't happen. And so I think about that a lot. And I think about effectively, I think my role is to help create, help create basically environment that's conducive to help us go as fast as possible. And so I think about that a lot. And again, I mentioned it, lots of agency urgency, responsibility. This is the way it's working right now for us. And I think it's working okay. Everything could always be better, but it's a tough deal managing folks, managing teams. It's definitely not straightforward and there's a lot of opinions, like anything else in life, there's a lot of opinions how to do it. I just got to kind of pick the one that you feel like it's going to work best for you and your culture and your team. You got to go with it,

Right? So final

Question. So since this show is called Delivered, I got to ask, can you tell us the last time you delivered something that was

Really meaningful? I could do an answer here. Not the package to someone, but more like in your

Professional career.

So I'm just trying to pick the altitude right now. There's a lot of stuff we're delivering this week per se, but
One of the things I think was interesting kind of thing is that when basically first felt basically like, Hey, look, you are truly making us more efficient. Yes, we have a larger help desk here. Yes, it's a prettier piece of software. Like I said before, that's all fine and dandy, right? But at the end of the day, what I'm here for is I want to actually know that I'm providing value in terms of helping people become more efficient and quicker. And that's very much been through a lot of work with automation, some deterministic, some non-deterministic, everyone talking to a Zapier guy. So also a guy who, by the way, running AI engineering at Zapier. So a lot of the rage right now is the LLMs non-determinism, right?
One of my beliefs here is I think that a lot of the ai, the startups and the ilk of ai, modern AI are throwing the baby of the bathwater out in terms of non-determinism. There's a lot of value deterministic workflows, frankly speaking and finding a mix between the non-determinism and determinism is a really key for it. And so hitting that stride I think is really important. Sorry, delivering to match your question. Actually being able to show, prove measure that you're making folks more efficient via basically using your software. I think that's probably the most important. Thanks.

Right. Thanks Taylor. That was awesome. Where can people find you?

One of the easiest ones is frankly just our website, ravenna.ai, LinkedIn. That works as well. Those are kind the two main fronts, what we're on, and yeah.

Awesome. Thank you so much again, and I look forward to following

What you're building.

Likewise. Appreciate it.

Cheers. Cheers. T