Episode 399 | How Derrick Reimer is Validating His Ambitious Third SaaS Application

Show Notes

In this episode of Startups For The Rest Of Us, Rob talks with Derrick Reimer, co-founder of Drip, about his new SaaS application Level. They talk about what inspired the idea as well as ways Derrick went about trying to validate it.

Items mentioned in this episode:


Rob: In this episode of Startups for the Rest of Us, I talked with Drip cofounder Derrick Reimer about what it’s like building and launching his next start up Level. This is Startups for the Rest of Us episode 399.

Welcome to Startups for the Rest of Us, the podcast that helps developers, designers and entrepreneurs be awesome at building, launching, and growing software products. Whether you’ve built your first product, or you’re just thinking about it. I’m Rob.

Derrick: I’m Derrick.

Rob: We’re here to share our experiences to help you avoid the same mistakes we’ve made. How are you this week sir?

Derrick: Well, I am deep in the process right now of building out some mockups of my new app called Level, to hopefully get some more concrete things in front of some of my early access list folks.

Rob: That makes sense. If folks are interested just right from the top, level.app is the domain name. You have a really tight landing page. If you’re listening to this and you want to see–this is maybe 400 to 500 words of text–and it is on point. You’ve got the headings, you can skim it, at the bottom is fomo, the reserve your handle today, you can claim a little slice of real estate on level.app. Then you have the social proof, right? 3,542 people who have reserved their handle including me, and so if your name is Rob, I’m sorry but I got level.app/rob. How’s the landing page working for you?

Derrick: It’s working really well. That is actually a live query now showing the current count. It works surprisingly well. I wasn’t sure exactly what to expect when I first built that but I figured that the scarcity play would be effective. I and had a few conversations with folks at MicroConf too, about this. I saw their eyes light up, and they’re like, “Oh yeah, make sure to let me know before you do that.”

Rob: That’s when you know you’re on to something.

Derrick: This is a good sign. They can appreciate it from a from a marketing ploy and then they’re like, this kind of instinct of, “By the way, I really want to know when you do that so I can get mine.”

Rob: For those listening, if you don’t know what Level is, it’s an alternative to real time chat designed for the software development workflow. I just pulled that right from your landing page but it’s ostensibly a competitor to Slack but a much less interruptive experience. More asynchronous, not the blinking red dot, I guess it’s not blinking in Slack but everything’s synchronous by default. In Slack and you’re going for the opposite in essence, and this stems from?

Derrick: From really my experience with Slack. My first time using Slack was with Drip when we were a team of like two or three. It was not painful at all at that stage, but gradually, as our team size grew, it started to get a little bit more cumbersome. I remember actually, the first time I was in a “big Slack,” team was in our building back in Fresno. That was like an early warning that, “Hmm, I think this maybe doesn’t scale so well as your team grows.”

The building chat was full of just folks tossing around various pieces of information, most of them not urgent but often pushed through with an @channel or an @here, “Hey, there’s some cookies in the in the kitchen.” And it’s like, “Do I really need to get pulled out a flow to hear that piece of information?” That was like an early sign that Slack seems kind of broken especially when there’s a lot of people chatting back and forth.

That experience was reinforced after we were acquired by Leadpages and we joined the broader company wide chat. There was about 150 people in there, and a really similar experience where a lot of well-intentioned people who weren’t trying to interrupt each other, but the tool just kind of failed as in essence and made it really hard not to accidentally interrupt people’s flow. Level is really kind of a reaction to that problem. I just observed this over the course of years and I just couldn’t get it out of my head, so now I’m building my take at a solution.

Rob: Yup. You wrote a manifesto a few months back and published it on your blog derrickreimer.com. I see now you’ve you republished it at level.app/manifesto. In essence, you’re describing the pain points you’ve just indicated—of the interruptive stuff in your story.

Derrick: Yeah. The manifesto was kind of step one actually for introducing the idea to the world. I didn’t really waste any time after moving on from Drip. I think I’ve published it on my last day at Drip, is when it kind of soft launched. Then the following Monday, I really did a big push in the manifesto. I was just sort of itching to get this out into the world. That promotion was probably, I think it worked really well because it sort of made a splash, it made a bold statement. I was able to gather quite a few email addresses off of that. I think it was maybe 300 to 400 within the first couple of days, and so that gave me a nice launching pad of folks to start engaging with. Obviously, people sharing it around on social helps too. I think that was a really effective strategy for making the first intro to the world.

Rob: II know you had kind of been noodling on it for awhile just in the background in your head and because we’re faced with the limitations of Slack everyday working on the Drip team, and you left Drip was it February of this year or was it March? I guess it was three or four months ago?

Derrick: Yeah, three or four months ago. Beginning of March, yeah.

Rob: Because that’s the more important thing for someone who’s listening to this two years from now, they don’t care what month it is. You hit the ground running. When you published the manifesto, I remember you wrote it up in a Google doc and then you had a couple of friends come in and edit and make suggestions and all that stuff, and then you published. What is it Twitter that gave you most of your sign ups? Because I remember, it didn’t go up in Hacker News, right? It got a few or something but it didn’t make it to the front page which I was frankly surprised by. How do you think you got those 300 or 400 sign up for someone else who’s thinking about doing it?

Derrick: It was predominantly Twitter. I think that was effective particularly for me because this is a product that’s marketed towards developers, and that’s kind of the premier place where designer, developer types hang out online. I think that was where most of my strategy was focused was trying to get people to retweet it, to like it, to share it. I did assemble a list of folks who are in my inner circle, friends of mine who also have a decent following on Twitter and did manual reach out to those folks just to tell them, “Hey, I’m going to be sharing this thing, if you wouldn’t mind, could you please tweet it out to your followers?”

Most people are really eager to help out. I think it’s where a lot of the lift came from. I did actually try running some Twitter ads that day. I didn’t really tuned the audience too much. I just let Twitter auto choose that by hitting the promote button. That drove probably 30% of the impressions for it but zero activity. I would say majority, I think I put in maybe $100 into that just to see if this is going to help provide any additional lift. I think a majority of it was really organic shares on social. I also emailed my own personal newsletter list which is pretty small just like a couple hundred folks on there at the time. I think I generated some shares off of that as well.

Rob: I could imagine that. I remember there were some comments, I think it was on Twitter but my favorite comments are always the ones that completely missed the point of an article. I would spend eight hours writing something. I know you spent a lot of time writing this, and then you have a one sentence that has something that is debatably, maybe factual, and someone rips into that, and you’re like, “Dude, that has nothing to do with the point.” There was one of these sentences we’re talking about, “Yeah, once we started using Slack, it was great. This is not news now, but five years ago, it was pretty groundbreaking.” There were somebody like, “Well, 12 years ago or 15 years ago, I was using IRC.” That’s not really the point, what are we even talking about right? Did you did you have much of that or was it the minority?

Derrick: It was very much the minority. I think it was on Reddit actually, and maybe I sort of attracted some of that flaming because I did post my own manifesto on there which I know is a little bit against the way you’re supposed to use the service. I was like, “Oh, what the heck. I’ll just post it in the developer subletter or whatever.” Some people took issue with the timing of when Chat was invented. The whole notion of arguing that this whole argument is invalid because I got the year wrong on when it came to the forefront, it was just laughable.

Rob: Yeah, it is what it is. Luckily, that tends to be the minority thing but it does always side track. That’s actually one of the reason that I completely disabled comments on my blog. I got tired of those conversations going on. There were good comments, and then just the ones that were irritating, or ill-informed, or just obviously looking to nitpick stuff. That’s just not helpful. That’s been my own personal journey with that.

You posted the manifesto, you got a few hundred emails, and you’re on the Art of Product Podcast with Ben Orenstein, mutual friend of ours, and you’ve been talking about your process through that, so I’m sure that that has helped get other developers interested in what you’re up to and probably slowly built your list over time. Once you had that list, what were your next steps? Because I know that you and I had talked about this a while back. This is a very ambitious project, and it’s either going to be awesome, or you’re just going to get smoked by Slack. They’re just going to stomp you, or not even notice you, and no one’s going to switch because of the high switching costs. It’s one or the other, and you’re fully aware of that upfront. What were your early steps of trying to validate the idea a bit more than just positing, “Hey, I can build a better Slack.” It’s like, “Okay, step two is…” What are you up to next?

Derrick: I was definitely leery of making the classic mistake of taking a little bit of early indication that we’re on to something, and just running with it and not talking to anybody—which is a mistake that I’ve made before in the past—and so many folks do. I really wanted to air on the side of having too many conversations with people to try to asses out is this, “Am I on the right track with this? Is this actually going to sell well in the market?”

I think it was within a few weeks after launching the manifesto, I sent an email out to the list. I decided to email the entire list which looking back was probably should have started with a smaller slice just to gauge what the response rate would be. But I basically emailed out and said, “Hey, thanks so much for signing up. I want to have a conversation with you and hear more about what your pain points have been with real time chat in the workplace. At this point, I’m not trying to propose any solution to you beyond what I’m kind of alluding to in the manifesto. I really just want to hear what particularly about real time chat isn’t working for you enough that you gave me your email address.”

I sent this out and I sent a Calendly link with that so that people could book a 20-minute slot on my calendar. I kind of reserved it to afternoons only, thankfully, to reserve the mornings for productive time. I think I got around 40 people booking time on my calendar.

Rob: Dope. That’s both really good news and also like, “Oops,” Were they 20 minutes each? What was that? 16 hours?

Derrick: Something like that. It turned out to be basically three and a half to four weeks of afternoons booked pretty solid. At first I was like, “Oh boy, this is a lot.” But I think, at the time, it’s early enough, I’m like, “This is probably where my time is best spent. Talking to people and hearing in their words what problems they’re having.” That did help guide the way I would thought about the product. I would say influenced where I thought the biggest emphasis should be.

I wasn’t sure if the emphasis should be on reducing interruptions, or just organizing content better, or should I be focused around really optimizing for asynchronous, where are the pain points actually. It helps clarify my thinking. One of the things that we’ve talked about a lot that helped in the early days of Drip was just getting feedback from people and looking for patterns. What are the things that we’re hearing over and over again, and those are likely to be things that we should be paying attention to. I did spot some of those themes and patterns. Looking back, it was helpful to have a decent sample size. If I’d only talked to 10 or 15 people, then that may not have been enough to spot patterns. I think it was good overall.

Rob: Yeah, I know it was a lot of time. It seems like you were learning quite a bit as you were going, and at that time, you didn’t have any mockups right? You really were just talking through how would this sound, or how would you use this, or that kind of stuff.

Derrick: Yeah. I tried hard not to actually tip my hand on what I was thinking. I just wanted to hear unbiased people’s take on like, “You know, if there’s one thing I could change about Slack.” or I ask questions like, “Do you use Slack threads and what do you think about them?” Or, “Do you use search heavily in Slack?” Just to get an idea of like how much are people relying on the tool to be their source of the repository of historical information, versus how much is just ephemeral conversations that get transferred into project management tools. There was a lot of things I was just trying to learn and SaaS out from people without tipping my hand too much on my thoughts of a solution.

Rob: Right. In the back your mind, obviously, you know that you need some kind of differentiation, pretty strong differentiation from Slack because everyone’s going to immediately compare you to Slack. One thing you’ve chosen is to niche down to developers, right? You know that interruptions piss developers off. It breaks your flow. You and I have experienced this first hand as our development team grew.

I used to tell people snooze your Slack for two hours, do not disturb it during this time, just try to get focus so that we can continue ship code at high velocity. You have that but there was there was one other thing, that early on, you made a decision to do that is potentially risky but it’s another differentiator. You want to talk a little bit about that?

Derrick: Sure. You and I noodled this little bit when I was just in the idea stage, and I think we were having drinks one day and I was like, “Alright. I think I figured out the one thing that’s going to really make Level stand out. Let’s open source it.” You spit your drink all over my face. There are examples of this happening, there’s Discourse, there’s Ghost, there’s GitLab, so there are companies that are already doing this.

Rob: Not in the chat space, those are in other spaces but it’s a mockup.

Derrick: Yeah. I will say there are other open source chat tools but I don’t think they’re really making any kind of headway as a business. They’re just open source only. The model of open source, the core code base, but then charge people for hosted version of the service is basically the model that I’m going for. The thinking behind it is that one, since this product is marketed towards developers, a lot of developers sort of appreciate when things are open source, when they can look at the source and see what’s happening with their data, and just sort of have that transparency, and also be able to download it, and stand up a cluster of servers, and manage it themselves if they really want to do that.

I’m sort of banking on the fact that most companies that have sufficient budget to pay for a tool aren’t going to want to go through the hassle of managing their own servers, and patching them, and keeping them up to date and all that kind of stuff. They’re just going to want to pay me for the hosted version. But if you don’t have the budget or if you’re bootstrapped, and you’re really scrappy, then by all means, download it, stand it up on your serve, and when you’re ready, you can transfer the data into the hosted service.

Rob: This was, in essence, around that you and I kept talking about, how you can have a free plan because I think you need one. We both thought you need one because that’s kind of part of the course in the space, and how are you going to do that as a bootstrapper, and not get killed by hosting costs, or not get killed by support and all that, and this is a way to do two things. That’s pretty ingenious if it works, and it’s, like you said, developers or more tech-oriented folks, this is in essence the free plan in addition to the benefit of it being open source which most people like, and then they can pay you for whatever the hosting, and the support, and as an almost SaaS app.

Derrick: Yeah, exactly. It’s not only the free plan but it also kind of paves the way for potentially offering on-prem if I’m sort of optimizing for the ability to setup the entire service from the ground up easily as opposed to just running a SaaS app or maybe you kind of cobble together your own hosting situation that’s not easily replicable. I think building it in this way paves the path for companies that don’t want to run a hosted service, or trust another company to run a hosted service for them, they can download it, and run it inside their firewall too.

Rob: Yeah, and that’s really interesting. Can you do that with Slack?

Derrick: No, I don’t think so.

Rob: There’s no on-prem version as far as I know. That could be an interesting enterprise play. I know that wouldn’t necessarily be the market you go after first, but if you get traction, you get name brand, and people are like, “Hey, not only is this open sourced–” which big companies tend to. They either hate it or love it, but if they understand it they’ll love it, and then like you said, you can do the on-prem play, and those are super expensive, that would be a really nice, high-end revenue source for you if you’re willing to put up with the headaches.

Derrick: Right. That probably would be down the line where I have traction, and a team, and I could kind of establish a team to run that end without me having to do all the sales and all that kind of stuff.

Rob: Totally, yeah. After you got all of the information from those Skype calls, I know you and I then met. I know you did a bunch of thinking on your own for a couple of weeks and then you were at a point and you said, “You know what, I have thoughts, it’s not ready to go into mockups yet. It’s not ready to build a UI. I have a bunch of ideas about different message types and how to structure these, let’s do wide port session.” And so you and I met at the local library, actually, in a nice little room, and you wanted to talk about I think, the value of that, or what that felt like and just the point of doing that.

Derrick: A lot of times there are these key points when you’re designing products where it’s like, there’s a lot of information scattered about and kind of coalescing that information into something actually tangible. It’s hard to hold that all in one person’s head, I feel like. I tried doing a little bit of solo whiteboarding, and I jotted down in a notebook, and I probably could have arrived at similar conclusions, but I think it would have taken a lot longer and probably wouldn’t have been quite as crisp and clear.

We whiteboarded for probably an hour and a half or two hours. I think we came away with some really concrete takeaways. It kind of started with like, “Okay, I have all these types of messages that people send in Slack. There’s water cooler type chat. There’s people shelling their work. There’s people announcing things to their team. There’s synchronous discussions maybe around an incident, or something happening in production where we’re needed to go back and forth quickly, and there’s people requesting something from someone else, or maybe things blocking their work.”

I had this long list of things. I was like, “Okay, I need to kind of build up what’s similar about these, what’s different, how will the application know what types of message do I need to ask the user to tell the app what type of message this is, or can we infer it. Then, how does this translate into a priority and notifications?” It was sort of like a really central piece of the application. I felt like I couldn’t just start designing UI because it all kind of hinges on how each of these types of messages is treated. I think this was a really good candidate for whiteboarding, and yeah, I felt kind of like we’ve reignited some of the magic from the Drip days–it was really fun too and motivating. I would say, if you’re solo founder, I think it can be really valuable to find a friend that you have good rapport with, and can kind of brainstorm with, and bounce ideas off of, and just get there kind of two brains thinking about the problem, it can often lead to a really great result in the end.

Rob: Yeah. I was, I don’t know–concerned is probably not the right word–but going into it I was like, “Uh-oh, Derrick’s been thinking about this for months and I’m way out of that whole process. Are we going to be able to rekindle the old magic?” On Drip, we just used to whiteboard all the time, and came up with really good stuff–I thought. I use to say, with the two of us in the room, it’s like 10 times better. We catch all the edge cases and it’s like you said, two people holding the whole thing in the collective heads rather than one person trying to do it. There’s always a back and forth, there’s your sanity checking, there’s just the collaboration, it’s just night and day. If you’re standing in front of a whiteboard on your own, I think you and I really found how different that could be when we started collaborating right on Drip.

Derrick: Yeah, because you’re thinking about stuff, and you reach these points where my train of thought hits a dead end. If you don’t have someone there to kind of either pivot it or just pick up where you left off, then it can be a really frustrating process. You feel like you’re slogging through mud, and just having a pair there to help keep things moving. There were times when we would sit there and just kind of stare off in the space for a minute or two, and that’s okay too. It was a really fun exercise.

Rob: I agree. Being comfortable with silence too. We’ve talked about this when I came on your podcast. If you are going to whiteboard with someone, one or two people, I don’t know if we ever found a third who is as in our own heads as we are, as in the flow. We certainly had some good collaborators at Drip but you and I can sit there for five minutes with complete silence and not feel weird. I think that’s a big thing is you let the other person think and you’re thinking as well.

That’s the other thing I think with whiteboarding is oftentimes, if I try to do it on my own, I’ll get to a point and I just get stuck. I just can’t get past this and that’s when you will step in, and be like, “What if we think about this?” It’s like, “Oh.” Now you got us past it, and we’re still in the flow, and then we could finish the whole rest the hour. But I would have stopped there and just got hung up on it for a day, potentially. I think that’s a good point where you’re like, you could have gotten there eventually but it may have taken you weeks, and it was 90 minutes or whatever for us. That was super fun, by the way. I came out of that feeling great.

You came out of that then with kind of a mental model of message types, and not quite UI. I know we threw you why ideas around. We never sketched anything but it was always kind of hand wavy and talking as we’re like, “This alert to do this in this flow.” Then did you go straight from there into mockups or what was your next step? I know there was a vacation in there as well, right?

Derrick: Yeah, I did take a little vacation. I’ve been having this feeling of like a little bit of guilt as a product person that I should be getting some more concrete ideas in front of users to get feedback—sort of what I alluded to at the top of the show. I did try to take a stab at building some mockups. I’ve been working on them for a little while now. It’s taking a lot longer than I expected. I think part of that is that, we had some concrete ideas formed from the whiteboarding session but like we said, they’re not actual envisioning of where the pixels will sit, and how the product will actually form together. It was still a pretty fuzzy vision of it.

It felt like I had a lot more work out of my head than I actually did when I started to lay out UI elements. I think that’s probably true of any idea where it’s easy to sort of picture this thing that’s not actually real, and once you try to make a concrete, it’s like, “Oh, there’s actually a lot here to still think through and work through.” Perhaps actually, we could whiteboard on this again and maybe burst through another kind of wall, or I just need to keep slogging through it and returning back to it every so often, and incrementally building it out.

Rob: Yeah, that makes sense. It sounds like you hit—not a robot necessarily—but it’s a bit harder, a bit more challenging than you thought it would be. Mockups are often like that. You’re almost trying to invent something new and you have to thread a needle because you can’t be Slack but you can’t be email.

Derrick: Most of the difficulty that I’m having is centering around this inbox in Level. That’s one of the core pieces of the product is–it’s something that’s really missing from Slack–is that I want to have one place where I can come to and quickly see, in priority order, all the things that need my attention. Then I can step away for six hours, go into deep focus mode, come back, and feel confident that everything I need to see with relative urgency is it’s all laid out for me just as I want it.

Once you start kind of exploring, “Well, what happens when there’s 100 messages posted in this group and 50 in this group? How does the inbox mutate over times, that things don’t become overwhelming, and it doesn’t devolve into kind of the email inbox mess that so many people experience? How can I actually make good on the promise that I am threading the needle well between email and chat?” Once I overcome this, it’s going to be, hopefully, something really good and is really going to resonate with people. But until then, I don’t feel like I can just start showing these theoretical mockups where I just kind of hand wavy say, “Oh, it’ll be always manageable for you.” I want to give some proof that I’m actually on to something–some framework or some methodology for categorizing and organizing these messages.

Rob: Right. Because the challenge when you’re building something new is to validate your most risky hypothesis–that the ones that are most likely to fail where the risk is. Typically, it’s not the code itself unless you’re going to build—“I’m going to build an AI engine to predict stocks.” Well, it’s like, “Alright, that’s your first thing. Can you even do that?” But here, you know that you can write the code to make messages go in and out of a queue and come out of the things. There’s two risks left that I see. One is, can you design a UI that is novel enough, and does thread that needle between the two, and is 3X better or 5X better than the other experience for a certain group of people. Then the last risk will be, will people pay for it?

You validated that somewhat through building your email list and having this in-person conversations. I’m guessing once you have mockups or at some point you’re probably asked for whether it’s verbal commitments, or you may take money upfront, and I don’t think you’ve decided on when you’ll do that. That is a big issue—building the mockups and figuring out how is this flow going to work, and is it as novel enough and iterating on it until it is.

Derrick: I think this is arguably one of the most important pieces to work out because without this, the whole promise falls down. It’s one thing to just introduce threads for example. Slack has kind of arguably, poorly implemented threads. If I just implemented something that was just like, “Oh, everything’s a thread. That’s great but that still doesn’t solve the problem of not being a complete catastrophe when you step away from it for a long time.” It’s one of the critical moments, I would say.

Rob: Indeed. I want to switch up a little bit. We’ve talked a lot about the specifics of Level and your process of validating it, and getting the design going. I’m curious, this is in essence, kind of the third app or third startup really that you’re launching. I know that you launched a couple that came and went pretty quickly. But then, you had obviously, Drip with me. You did Codetree, codetree.com which is kind of project management that sits on top and you did that on the side. You sold that while we were, actually, right around the time that we sold Drip, you sold that if I recall.

Derrick: I sold that, we sold Drip, and I sold my house when we moved. I had three big sales.

Rob: That was crazy.

Derrick: Talk about massive liquidation.

Rob: I know. This really is, for all intents, your third act here. I’m curious if you think it’s easier this time around, is it harder, is there more pressure. Talk me through the emotional side or the mental side of where you’re at with it.

Derrick: I think it’s unique in that, with Drip, I started out as a contractor, and so I was gathering a paycheck. Then I went full time on it and then we had HitTail bankrolling our efforts a bit. It was a self-funded endeavor that was throwing off enough money for a least a decent paycheck so that I could live. Codetree was a side thing–that was nights and weekends. I was basically sacrificing my free time, but in exchange, it wasn’t really costing me actual money. Now, this is the first time that I’m really, as an adult, going off and saying like, “I’m not going to earn any salary or revenue for the whole time that I’m building this thing, or at least until I launch it.”

I think that that introduces, in itself, a bit of pressure that I’m trying to just like always mentally overcome that. Obviously, selling Drip and Codetree helped give me a little bit of runway so that I can afford to do this. That’s arguably the whole reason why we do this is so that we can afford to work on the things we want to work on. I’ve had to do some things like, set aside my living expenses for the next year, and figure out loosely what my budget’s going to look like, and then make transfers from this account to my checking account, and just say like, “I’m paying myself for the next year. I’m not allowed to think about money or stress about it.” It’s a work in progress but working on trying to do things like that to keep myself sane and not get too stressed out.

Rob: That makes sense. I find it funny you’re like, “This isn’t the first time as an adult,” because you were an adult all the time but now you’re married. You have real expenses. You live in a city where the cost of living is not super high. But Minneapolis is definitely akin to a California city that’s not L.A. or San Francisco. It was similar to Fresno which is not super cheap. It’s not like you’re going to live on $2,000 a month here. Depending on how you live and where you live, it’s a non-trivial sum.

I’m glad you did that. I know that you were stressed early on about burning some cash. As you said, the two exits you have helped bolster that, but I think your wife also distinctly gave you permission. She’s like, “You need to do this. You earned this.” Is that the phrasing? It was something like that. Like, “That’s what you did these other ones.”

Derrick: Exactly. This is not the time now to say, “Oh, I don’t know. Maybe I should keep a job and do this nights and weekends.” She’s like, “No, no. This is the whole reason why you sell companies.” I think she has a lot of wisdom in that.

Rob: Yeah, I would agree. Cool sir. Thanks for joining me today while Mike is out of town filling in our 399th episode.

Derrick: Oh my gosh. You’re one off from 400.

Rob: Isn’t that crazy? We’re going to figure something out to do that’s like a big bang, not just Mike and I talking about five ways to wrap or something. Can we do something different? We were the worst at that. We’ve looked back and I think 300 we may have just done a normal episode, that’s kind of a shame.

Derrick: Yeah, you got to take these milestones and squeeze some juice out of them.

Rob: Yup, yup, celebrate them a bit. Cool sir. Well, if folks want to keep up with you, they can hit level.app yeah and then derrickreimer.com. That’s where you blog now and again about your experience. You’re building Level for developers out there, you’re building it in Elixir and Phoenix.

Derrick: Elixir Phoenix is the web, framework web, kind of the rails of Elixir.

Rob: GraphQL…

Derrick: GraphQL is the API layer, and then Elm on the frontend, so the Elm translates into javascript. It’s a very shiny new tech stack for me. But I’m having a lot of fun building in it.

Rob: Well, I think for most people.

Derrick: Yeah, for most people. It’s a fairly proven technology but still on the newer side. I feel it’s really exciting. It’s predominantly centered around functional programming languages. Elixir and Elm, they’re very functional. It’s a bit different from what I’ve been doing. I’ve been writing Ruby for the past eight years, but yeah, I’m having a lot of fun for it. I’m hoping that the Level codebase can be a good example of like a full-scale SaaS application that’s kind of out there in the world for people to reference.

Rob: Yeah. I obviously don’t know any of the languages you’re using but you have such a good way of organizing our codebases, and your Read Me was super clean. I was like, “Ha, if I had a few extra hours this week, I would just run these commands and try to see if I get this thing right.” It was kind of fun. It was good to see it and then get out there. Cool. Well, thanks again for coming on the show.

Derrick: Yeah, thanks for having me.

Rob: I think that wraps us up for today. If you have a question for us, call our voicemail number at 888-801-9690, email us at questions@startupsfortherestofus.com. Our theme music is an excerpt from We’re Outta Control by MoOt, it’s used under Creative Commons. Subscribe to us in iTunes by searching for Startups. Visit startupsfortherestofus.com for a full transcript of each episode. Thanks for listening. We’ll see you next time.

Twitter Digg Delicious Stumbleupon Technorati Facebook Email

One Response to “Episode 399 | How Derrick Reimer is Validating His Ambitious Third SaaS Application”

  1. Always awesome to get an update from Derrick’s side. Will be sure to check out the new project!