# We had six months left  ## Episode metadata - Episode title: We had six months left - Show: Adam's Morning Walk - Owner / Host: Adam Wathan - Episode publish date: 2026-01-07 - Episode AI description: A candid discussion reveals the pain of downsizing a talented team due to declining revenue. The emotional toll of layoffs takes center stage as the host reflects on public criticism and the personal impact of letting go of colleagues. Optimism shines through plans for a leaner team and innovative product ideas. Using AI for troubleshooting emerges as a potential solution to operational challenges. The host pays heartfelt tribute to the engineers affected, sharing their skills and offering support for future opportunities. - Duration: 33:26 - Episode URL: [Open in Snipd](https://share.snipd.com/episode/092055b6-48ed-4551-88d6-568560c88ec6) - Show URL: [Open in Snipd](https://share.snipd.com/show/e2dc97ec-2a4d-4eb8-9f32-29045e69b647) - Export date: 2026-02-17T23:20:51 ## Snips ### [Slow Decline Masks Imminent Risk](https://share.snipd.com/snip/84f7d654-5994-4848-be73-ee1a9d2107d7) 🎧 01:11 - 04:15 (03:04) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/d06cf1d2-300a-46e0-b204-92a817dc2d76" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Adam realized revenue was declining steadily each month and that trend compounds into insolvency if unchanged. - Proper forecasting showed they had about six months before payroll became impossible without change. #### 💬 Quote > If absolutely nothing changed then in about six months we would no longer be able to meet payroll obligations. > — Adam Adam on forecasting and discovering a six-month runway #### 📚 Transcript **Adam:** And I've shared some of the ideas we have for turning things around. And, you know, I'm still optimistic that we can do that. I'm excited about these ideas. But over the holidays, I finally did what any responsible business owner should probably would probably have already been doing, which is do some proper forecasting of revenue versus just going based on gut feel. um yeah situation was just like a significantly worse than i realized so revenue has been like trending down slowly for a couple of years um but it's been almost like so slow and so steady that it's like a boiling the frog sort of situation where you almost like don't really notice you just get used to the lower revenue being sort of normal and you can oh still pay all your bills and stuff but um almost just got hit by a car backing out of the driveway so forgive me for uh being a little uh here but yeah what was I saying the boiling the frog situation every month it feels like it's kind of fine because you're still able to pay your bills still have a little bit of money left over for yourself and yeah you know it seems okay and you just get used to each month being a little lower than the previous one. And you kind of forget that the previous one was a little bit higher. And the story I've been telling myself in my head the whole time is that, you know, things are low. But, like, they're not really declining anymore. We've sort of hit, like, a stable point. Because it sort of feels like revenue has been the same for a long time now but yeah i finally went and did like proper forecasting looking at the actual data and charting things and calculating what the average drop in revenue is each month and um was sort of caught off guard to realize that no it's not stabilizing at all and um revenue is consistently down by pretty much a fixed amount every month which means it's an ever increasing percentage of revenue and forecasting it out like the next year if nothing changed at all which of course is not the plan because we're trying to grow the business we're trying out new ideas and stuff like that so the intention isn't that the trend will continue but um i can't guarantee that it won't right it's not in my control i can try to influence it but i don't control the outcome um if absolutely nothing changed then in about six months we would no longer be able to meet payroll obligations, which was a pretty shitty thing to realize. --- ### [Difficult Layoffs And Team Reset](https://share.snipd.com/snip/f928a4c4-c3ca-468c-8c98-299d3c443689) 🎧 05:07 - 10:03 (04:56) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/b18fa319-7b5b-4660-9810-b64c9ad71803" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - On Monday Adam let go 75% of his engineering team to match the business's cash runway. - The team went from four engineers to one plus three owners handling the work. #### 💬 Quote > On Monday, we officially let go 75% of our engineering team. > — Adam Adam recounting the layoffs and team composition after the cuts #### 📚 Transcript **Adam:** So what's today, Wednesday? Yeah, so on Monday, we officially let go 75% of our engineering team in order to get the business into a place where our declining revenue can comfortably support the staff that we have. Now, the company is not huge, so 75% isn't tons of people. It's three people that we had to let go. But I do think the percentage is important to share because to say that oh we have to let go three people it doesn't sound like we have to do much but in reality we have four engineers on staff and now we have one so it is a big change um so yeah that uh i don't know uh i don't it fucking sucks you know like these are three people that I really love working with. Three of the most talented people I've ever worked with. People that I was really excited about continuing to build new stuff with going forward. And I had all sorts of ideas for things that we could build together. the reality is like i either had to ask them to work here for free or let them go and find another job where their employer actually could pay them every month you know i think like i don't know if you've never run a business people sort of have a lot of lacking perspective about what it's like to actually lay people off or why people do it and stuff like that. I saw, you know, Jeffrey from Laracast had to do some layouts this week, too, and he sort of got dragged through the Through the mud by a couple people on Twitter now thankfully, 90% of people understood and didn't pile on. But yeah, some people just think like when you lay people off, it's because you're being greedy or because no one gives a fuck about you or whatever. And I don't know, that's just sort of something that that's a burden you fucking carry as an employer, unfortunately, is that the world fucking hates you and thinks you're evil. Um, but yeah, so I don't know. It, it sucks. Um, sick to my stomach about having to do it. Feel like a failure for having to do it, building this fucking open source thing that's insanely successful um but the business success is like inversely correlated with the um popularity of the open source right now which is really uh really rough so i don't know i don't know i just feel like i wanted to sort of share that update in the spirit of the sort of transparency of this podcast. But yeah, again, I don't have a really clear bulleted list of stuff to go through here. I'm just sort of talking out loud to myself, almost for therapeutic reasons and just getting some of this stuff off my chest but yeah it's uh not good good morning so um yeah i don't know i don't know where what else to talk about here um i got a few things i guess like one is, um, I don't know, like, what do we do going forward? I guess, you know, I'll, I'll say I am still optimistic. I sort of like, my job requires me to be optimistic. And I am, you know, I think the ideas that we had that we've been working on since before the break are good ideas. And once we get more of that stuff out there, I really believe it's going to make our products better. And I hope a better product attracts more customers and can make things more sustainable. Peter on our team has been doing an amazing job finding partners for our partners program, which has really helped to diversify revenue and give us more predictable recurring revenue, which has been great. So we're hoping to grow that some more too. But yeah, admittedly, it's going to be an uphill battle with a very small team. To be clear, the remaining team is the three owners of the business, which is myself, Steve, of Refactoring UI fame, who's always done all our design stuff. And then Jonathan Rennick, who helped me create Tailwind initially and also created Inertia.js. So there's us three who own the business. --- ### [Prepare For Backlash During Layoffs](https://share.snipd.com/snip/3cb566ad-50f4-4417-b147-c8c7ec47e7d9) 🎧 06:57 - 07:17 (00:20) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/1573429a-b019-436f-8650-933cb03af82b" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Expect public criticism when making layoffs; prepare clear communication and humane severance. - Focus on treating affected people well because reputational backlash is likely regardless. #### 💬 Quote > When you lay people off, it's because you're being greedy or because no one gives a fuck about you or whatever. > — Adam Adam on public reactions to layoffs and employer burdens #### 📚 Transcript **Adam:** the Through the mud by a couple people on Twitter now Now, 90% of people understood and didn't pile on. But yeah, some people just think like when you lay people off, it's because you're being greedy or because no one gives a fuck about you or whatever. --- ### [AI As Problem And Productivity Multiplier](https://share.snipd.com/snip/b137003a-e0eb-455e-b249-9256210c5da1) 🎧 11:42 - 13:54 (02:11) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/722d644d-8cfc-4d99-9a75-65f1bfb58d75" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Adam sees AI as both a cause of revenue shifts and a necessary tool to scale with fewer people. - He used Claude to help diagnose a server issue and speed up troubleshooting. #### 💬 Quote > We just like SSH'd into the server and installed Claude code and just started treating Claude like a devops engineer. > — Adam Adam on using AI (Claude) to troubleshoot server problems #### 📚 Transcript **Adam:** So yeah, I guess that is a bit of a a silver lining if i'm trying to be optimistic i i am looking forward to sort of having no choice but to really just get in there and do a lot more stuff um myself i think um the whole ai thing is a clearly a double-edged sword because i think ai is a huge reason why our business is struggling even though it's making Tailwind more popular than ever but I also think it's an amazing technology and I'm excited about it in the ways that it can help me you know and I think in the position that we're at now we're sort of going to be forced to figure out how to leverage some of this stuff to stay on top of all the things that we need to stay on top of and I don't know what that fully means yet but I have some ideas like it'd be interesting to see like how can we throw AI at some of the github issue triage and stuff like that, that has historically taken up a bunch of our time. Even yesterday, you know, like we had an issue with a server that was running out of memory and running out of CPU. And it would have been, I mean, the funny thing is ultimately the way we managed to fix it was uh Robin had an idea for a change to get the CPU down um but even like sort of diagnosing the issue we just like SSH'd into the server and installed Claude code and just started treating Claude like a devops engineer and asking it questions and asking it to try and diagnose things. And it's the thing that at least helped us figure out that, okay, like the reason this thing is busted is because the load on the thing is too high, not because of any like specific bugs in the code or anything like that. But, you know, we were able to sort of troubleshoot things a lot faster because of that than we would have been by hand trying to look up what commands I'm supposed to type at the terminal and look at all this data. --- ### [Docs Format Affects Product Discovery](https://share.snipd.com/snip/2d41a619-912a-4053-bdc1-8f15391f2d59) 🎧 16:44 - 18:24 (01:40) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/3cf9d031-d5ee-45ed-b05a-89c06bbd2b21" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Documentation changes that make content easier for LLMs to consume can reduce users discovering paid products. - Adam worries markdown endpoints may allow models to surface answers without exposing paid-product signals. #### 💬 Quote > The only way people find out about our paid products is through the docs that's our entire distribution. > — Adam Adam on how docs traffic drives paid product discovery and revenue #### 📚 Transcript **Adam:** there's a good chance that working on this is actually going to make us less money. And we're not in a position to prioritize that type of work. So closing this for now. And of course, a bunch of people got irritated in the comments. Wow, this is so bad for the community. All you want to do is make money. Blah, blah, blah, blah, blah. Really disappointed. And I wrote a a big long response to that last night basically saying okay assholes like we had to lay off three people at the beginning of the week because revenue is down so much every single second that i spend reviewing a pr um to add uh lm documentation to the site is a second that I don't spend trying to turn around revenue at the companies so we can keep paying the people that are still here. And this feature actually has a risk of decreasing revenue even more. As much as I like the idea of having the feature, I think if we're going to implement it we need to solve this problem of how do we implement it in a way that doesn't further sort of like um bury and hide the existence of our commercial products like and i i don't know to me the only way i can think to do that in a way that llms are actually gonna not just discard it is figure out how to sort of integrate that information into the docs in a way that's helpful. So people actually, so LLMs like are actually going to say to people, you know, okay, like, by the way, it looks like this thing exists, which could be helpful if, you know, you need more help with this. You know, that would be the dream. --- ### [Prioritize Revenue-Driving Work](https://share.snipd.com/snip/ab75c14f-6d5f-4287-b014-39a80f878436) 🎧 16:48 - 19:34 (02:46) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/38cab192-e4b1-4869-8bea-9b5d8899fec0" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Prioritize work that directly drives revenue when runway is tight and cut lower-ROI community features temporarily. - Avoid spending scarce time on changes that might reduce discoverability of paid products. #### 💬 Quote > Right now we have to focus on doing things that help the company make money otherwise this framework is not going to be sustainable. > — Adam Adam explaining why he closed a docs PR to prioritize revenue work #### 📚 Transcript **Adam:** And we're not in a position to prioritize that type of work. So closing this for now. And of course, a bunch of people got irritated in the comments. Wow, this is so bad for the community. All you want to do is make money. Blah, blah, blah, blah, blah. Really disappointed. And I wrote a a big long response to that last night basically saying okay assholes like we had to lay off three people at the beginning of the week because revenue is down so much every single second that i spend reviewing a pr um to add uh lm documentation to the site is a second that I don't spend trying to turn around revenue at the companies so we can keep paying the people that are still here. And this feature actually has a risk of decreasing revenue even more. As much as I like the idea of having the feature, I think if we're going to implement it we need to solve this problem of how do we implement it in a way that doesn't further sort of like um bury and hide the existence of our commercial products like and i i don't know to me the only way i can think to do that in a way that llms are actually gonna not just discard it is figure out how to sort of integrate that information into the docs in a way that's helpful. So people actually, so LLMs like are actually going to say to people, you know, okay, like, by the way, it looks like this thing exists, which could be helpful if, you know, you need more help with this. You know, that would be the dream. Problem is like figuring that out and executing that on like whatever many hundreds of pages there are on the site properly on a per page specific way is a big project and a hard project to sort of figure out and I don't know what the ROI on that would be right away so it's pretty hard to prioritize so that's like why I'm going to close this pr because it's not the version of this feature that i want to have on the site right now and again if i could wave a magic wand i would love to have it in there and figure this problem out but i can't spend the time on that right now i have to spend the time on uh you know improving 10 plus as a product trying to figure out how to find more customers for 10 plus and i don't know maybe that is one of the ways that we find more customers for it i just don't know but um everyone on github just acts like if you don't merge their pull request that you are you have something against them and their specific thing, when the reality is there's like 600 fucking pull requests, and only limited amount of time, and bandwidth to do anything. --- ### [Philip: Versatile Fast Engineer](https://share.snipd.com/snip/12b5eaf1-e8bc-4d72-8834-cd04c6ff7dbd) 🎧 22:14 - 25:14 (02:59) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/f80b152b-687b-4fb2-89bc-c0d33faac4a3" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Adam described Philip as a fast, business-minded engineer who shipped across front-end and product areas. - Philip previously worked at Sourcegraph and Facebook and brought strong product sense. #### 💬 Quote > He works extremely fast. He's extremely AI- which has been fun to watch he just lives on the cutting edge. > — Adam Adam describing Philip's skills and background #### 📚 Transcript **Adam:** Who are going to be looking for new roles soon. I mean, I don't know exactly what their plans are. but I promised them and I want to do everything that I can to help them find really great gigs that they're really excited about. And maybe someone listening to this podcast works at a cool company and is looking for extremely talented engineers. So yeah, I'll just say a few things, I guess, about each one of these dudes in case it's helpful for them. And in case you're out there hiring, maybe one of these guys would be a great fit for your company. So Philip worked with us. joined us a year and a half ago as just a staff engineer on the team. Before working with us, he was working at source graph before source graph he was working at facebook on the react core team um actually i don't know if he was directly on the react core team at facebook work on react full time i think he worked on like facebook workplace and some other stuff there but this dude is just like brilliant He was referred to me by Torsten Ball who if you're on Twitter, you probably know Torsten. He's a brilliant engineer and he told me about Philip. So Philip's one of best guys he's ever worked with. And that's how I found Philip and hired him. And, you know, he was just instantly productive at our company, working on literally anything and everything. Worked on, you know, Tailwind Core. He built out a Tailwind Plus Elements library. Recently, he's been building all sorts of crazy front-end, iFrame-y, responsive scaling interfaces for previewing components for Tailwind Plus. And this new design that we've been working on. He can basically do everything. He works extremely fast. He's extremely AI- which has been fun to watch he just lives on the cutting edge and is always curious and always learning and he's a great communicator he has a great sort of business and product sense much better than i think your average engineer. Like he's always asking the question like, which, you know, how is this going to deliver value for the business? Does this make sense to work on? Does it make sense to invest more than X amount of days in this based on the ROI we're going to get on it? You know, he is, he just thinks about things the way that any business owner would want an engineer to think about things. Just an absolute pleasure to work with. I wouldn't even be able to say that he had specific strengths in specific areas because he could do anything. He's just a smart guy who knows how to make the computer do what he wants the computer to do. And yeah, just a really great personality and just a great guy to work with. Love Philip. I'm looking forward to seeing what he does next. --- ### [Jordan: The Fearless Troubleshooter](https://share.snipd.com/snip/12a6d035-feef-4c94-af7b-a1393e96b860) 🎧 25:14 - 27:15 (02:01) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/4ec07662-6b7c-46cd-9f66-1cce185a0315" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Adam praised Jordan's encyclopedic, fearless troubleshooting across unfamiliar codebases. - Jordan contributed to Headless UI and solved obscure platform bugs quickly and reliably. #### 💬 Quote > I've never, ever in my life worked with someone who could take a totally unfamiliar code base ... and just jump into it and figure out how to work with it. > — Adam Adam on Jordan's ability to diagnose and fix obscure issues #### 📚 Transcript **Adam:** Next would be Jordan on our team, who he used to be like a really high-level contributor to Laravel Mix years and years and years ago. Just like another building tool, right? Some trends here is the sort of stuff we work on. But that's how I sort of met Jordan. And man, this guy has just more cryptic, esoteric knowledge hidden in his brain than any human being I've ever met in my life. If there's some obscure bug that we run into that no one understands, he always had an answer of, oh yeah, there was this bug in Safari or in the WebKit engine that caused this, this, and this that I know about. And here's the link. And this is not stuff he researches to find. This is just stuff he knew. just knew everything. And I've never, ever in my life worked with someone who could take a totally unfamiliar code base that they've never seen ever and just jump into it and figure out how to work with it. I think like compared to me, if you put me in a code base, I've never seen, it's like stressful. And I'm sort of like, I'm sort of like afraid of it, you know, and anxious and nervous. This guy has like no fear of any code base. If we ever needed to like fix something upstream in another library or needed to try and fix something in a framework or a tool to improve compatibility with tailwind he was the obvious person to throw at those things because he would just fearlessly dive into these code bases and find the problem and open a pr and have it ready to go like i feel like this guy is like ai before ai you know um yeah i've just i've never met anyone who could troubleshoot problems and diagnose problems as fast and reliably as him. And his breadth of knowledge is just insanely vast as well. --- ### [Dan: Designer Who Codes And Teaches](https://share.snipd.com/snip/938b8d5b-d45b-4866-837a-4c3d05727633) 🎧 28:28 - 31:39 (03:11) <iframe src="https://share.snipd.com/embed/obsidian-player/snip/678bd4e2-3c1f-4076-8222-b70bb023c3f7" width="100%" height="100" style="border: none; border-radius: 12px;" sandbox="allow-scripts allow-same-origin allow-forms allow-popups allow-clipboard-write" ></iframe> - Adam celebrated Dan as a designer-engineer who builds tooling and clear educational content. - Dan writes deep explanatory pieces and builds design systems and preview tools for the product. #### 💬 Quote > He is an incredibly gifted educator and writer ... I think Dan's going to take a little bit of time before his next thing because he's working on a book. > — Adam Adam describing Dan's design, engineering, and writing skills #### 📚 Transcript **Adam:** He's extremely resourceful. He never gets stuck or blocked by basically anything. He just figures things out, even if they're totally out of his wheelhouse. He's an incredibly gifted educator and writer, which is the thing that made me most excited to work with him honestly because it's such a rare skill and if you've seen dan on twitter maybe you haven't noticed dan on twitter but if you've ever seen a viral twitter thread with a bunch of isometric drawings and diagrams in it 90 chance it was created Dan. I think Dan's going to take a little bit of time before his next thing because he's working on a book, which he has up at makingsoftware.com, which is just like a big sort of like collection of independent articles and essays on how different things in computers and software work going really deep with all these diagrams like things like how llms work or how computer displays work or how oklch color system works just like tons of different stuff that he's gone deep on and figured out how to explain to people and i think it's really going to be a fun resource so i'm excited to see what he does with that um but yeah he is just a delight to work with super talented designer a designer who really gets engineering you can get his hands dirty in engineering like he is like the person at the company that we trusted to work on like our custom syntax themes for example because he knows what colors to pick because he understands that a comment is less important than a function name you know which is a hard thing for a designer to work on who isn't also an engineer but yeah dan man he he could just do like anything here and he always had an incredible attitude he would he wrote like half of the fake content for us and he has an amazing sense of humor um he designed tons of stuff here for us especially helped uh a lot on the sort of the rebranding of tailwind 4 and sort of the new look that we did there he designed all of the new p3 colors for tailwind 4 and built like entire custom tools for picking those colors and previewing those colors and stuff like that um yeah i don't know can't say enough good things about dan just finding a designer who really gets engineering and really knows how to work within the constraints of the medium and how to build their own tools to help them solve their own problems. And it's just always on the cutting edge, you know, like he's another guy who's just loves playing with AI and figuring out how to combine some of these things together. He did all the work to create all the sort of fake interview videos for the course platform that we built recently. And yeah, I don't know. I'm going to be really sad to not work with Dan too. --- Created with [Snipd](https://www.snipd.com) | Highlight & Take Notes from Podcasts