From Burnout to 3 Products Live: My 6-Month Rebuild Story

Six months ago, I couldn't write a single line of code without feeling like the walls were closing in. I'd sit down at my desk, open my editor, stare at the cursor for twenty minutes, then close the laptop and go split firewood instead. The firewood didn't judge me. The firewood didn't have deadlines.

I was burned out. Not the trendy kind of burnout that people post about on LinkedIn with a sunset photo and a caption about self-care. The real kind. The kind where you've spent thirty years building software and you wake up one morning genuinely unsure whether you want to do it for even one more day.

This is the story of how I went from that place to having three live products generating revenue in six months. It's not a fairy tale. It's messy and honest, and some of what I did was probably wrong. But it worked.

---

Month Zero: The Breaking Point

I need to back up and explain how I got here, because burnout doesn't happen overnight. It's erosion.

For the previous two years, I'd been grinding on contract work while trying to launch side projects. The contract work paid the bills — good enterprise rates, interesting enough problems — but it consumed my best hours. By the time I finished client work, I'd try to shift gears to my own projects and produce absolute garbage.

I was working twelve to fourteen hour days. Not because anyone was forcing me to, but because I'd internalized this idea that if I wasn't shipping, I was falling behind. Behind whom? I have no idea. Some imaginary competitor who lived in my head and never slept.

The tipping point was a Tuesday in September. I was debugging a race condition in a client's payment processing system at 11 PM, and I realized I hadn't eaten dinner. I hadn't gone outside. I'd been sitting in my cabin in Alaska — surrounded by some of the most beautiful wilderness on the planet — and I hadn't seen any of it because I was staring at log files.

I finished the bug fix, submitted the PR, and then I just stopped. Closed everything. Didn't touch a computer for two weeks.

---

Month One: Doing Nothing (On Purpose)

Those two weeks of doing nothing extended into a full month. And here's the thing nobody tells you about burnout recovery: the first phase feels like failure. You're not producing anything. You're not shipping. You're just… existing.

I chopped wood. I hiked. I read actual books — not technical books, novels. I fixed a broken fence on the property. I cooked meals that took more than five minutes. I slept eight hours a night for the first time in years.

And somewhere around week three, something shifted. I started having ideas again. Not forcing them — they just showed up. I'd be walking through the birch trees behind my cabin and think, "You know, there really should be a better way to look up vehicle history data." Or I'd be organizing my pantry and think about how many homesteaders struggle with the same thing.

The ideas weren't the important part. The important part was that I wanted to build them. Not because I felt obligated or pressured, but because the spark was back. That intrinsic motivation I'd ground down to nothing through two years of overwork was slowly regenerating.

What I learned: You cannot optimize your way out of burnout. You can't hack it with a new productivity system or a different task manager. You have to actually stop. Your brain needs time to defragment, and you have to give it that time even though it feels irresponsible.

---

Month Two: The Rules

When I came back to code, I made myself a set of rules. I wrote them on an index card and taped it to my monitor. They're still there.

1. No more than six hours of code per day. Not negotiable. When the timer goes off, I stop, even if I'm in the middle of something.
2. No client work. I had enough savings to cover six months of expenses. I was going to use that runway to build my own things or admit that I didn't actually want to be an entrepreneur.
3. Ship small. No more nine-month development cycles building perfect products nobody asked for. Get something live in weeks, not months.
4. One thing at a time. I'm not allowed to start project B until project A is either live or deliberately killed.
5. Mornings are for building. Afternoons are for marketing, writing, and operations. Evenings are for living.

Rule number two was the scary one. Walking away from contract income when you live in Alaska and heating oil costs real money requires a certain kind of recklessness. But I knew that if I kept one foot in client work, I'd never fully commit to my own products. The safety net was also a trap.

---

Month Two-Three: Product One — AutoDetective.ai

The first product I built was AutoDetective.ai. I'd had the idea for months — a tool that lets you look up detailed vehicle information using AI to synthesize data from multiple sources. The used car market is massive, buyers are anxious about getting scammed, and the existing tools like Carfax are expensive and clunky.

I gave myself four weeks to get an MVP live. Here's what the core looked like.

var express = require('express');
var router = express.Router();

router.post('/api/vehicle/lookup', function(req, res) {
    var vin = req.body.vin;

    if (!vin || vin.length !== 17) {
        return res.status(400).json({ error: 'Invalid VIN' });
    }

    // Fetch from multiple data sources in parallel
    var sources = [
        fetchNHTSAData(vin),
        fetchRecallData(vin),
        fetchMarketData(vin)
    ];

    Promise.all(sources).then(function(results) {
        var vehicleData = mergeResults(results);

        // Use AI to generate a plain-English summary
        generateReport(vehicleData, function(err, report) {
            if (err) return res.status(500).json({ error: 'Report generation failed' });
            res.json({ vehicle: vehicleData, report: report });
        });
    }).catch(function(err) {
        console.error('Lookup failed:', err.message);
        res.status(500).json({ error: 'Lookup failed' });
    });
});

The key insight was that the value wasn't in the data — most vehicle data is publicly available if you know where to look. The value was in the synthesis. People don't want to read a NHTSA recall database. They want someone to tell them, "This 2018 Tacoma has two open recalls, the market price is about $28K, and here's what to watch for."

I shipped the MVP in three and a half weeks. It was ugly. The landing page was basic. But it worked, and I put it in front of people immediately.

Revenue by end of month three: $127. Not impressive, but not zero. People were paying, which meant the problem was real.

---

Month Three-Four: Product Two — Grizzly Peak Software

While AutoDetective was getting its first users, I rebuilt grizzlypeaksoftware.com from the ground up. The old site was a static page that hadn't been updated in two years. I turned it into a content hub — articles, tutorials, a jobs board, a library of resources.

This wasn't a "product" in the traditional sense, but it was a product in the business sense. Content generates traffic. Traffic generates authority. Authority generates opportunities. Every article I write here is a long-term asset that compounds over time.

// Article route with clean URLs and SEO
router.get('/articles/p/:slug', function(req, res) {
    var slug = req.params.slug;
    var shortId = slug.split('-').pop();

    fetchArticleByShortId(shortId, function(err, article) {
        if (err || !article) return res.status(404).render('404');

        var canonicalUrl = 'https://grizzlypeaksoftware.com/articles/p/' + slug;

        res.render('article', {
            article: article,
            canonicalUrl: canonicalUrl,
            structuredData: buildArticleSchema(article)
        });
    });
});

I committed to writing consistently — not every day, but on a schedule. Two to three articles per week. Topics I genuinely knew about: API development, AI integration, career advice for experienced engineers, business strategy for solo developers.

Traffic by end of month four: Around 2,000 unique visitors per month, growing steadily. Not viral, but sustainable organic growth from search.

---

Month Four-Five: Product Three — The Content Pipeline

The third product emerged from necessity. I was writing articles for Grizzly Peak and realized I needed a system to manage content at scale. Research, drafting, editing, SEO optimization, publishing — the workflow was chaotic.

So I built a content creation pipeline. Scripts that handle research aggregation, article generation with AI assistance, quality control, and publishing to the CMS. The pipeline itself became a product — not one I sell externally, but one that makes everything else I do more efficient.

// Content pipeline - process an article from draft to publish
var pipeline = require('./pipeline');

function processArticle(articleId, callback) {
    var steps = [
        function(next) { pipeline.research(articleId, next); },
        function(data, next) { pipeline.draft(articleId, data, next); },
        function(draft, next) { pipeline.optimize(articleId, draft, next); },
        function(optimized, next) { pipeline.publish(articleId, optimized, next); }
    ];

    runSteps(steps, function(err, result) {
        if (err) {
            console.error('Pipeline failed at step:', err.step);
            return callback(err);
        }
        console.log('Article ' + articleId + ' published successfully');
        callback(null, result);
    });
}

This is the unsexy kind of product that nobody writes about on Hacker News. It doesn't have a landing page. It doesn't have users. But it's the infrastructure that makes the other two products possible. Without it, I'd be spending twenty hours a week on content. With it, I spend about six.

---

Month Six: Where Things Stand

Here's the honest scorecard at the six-month mark.

AutoDetective.ai:

• Monthly revenue: $840
• Monthly active users: ~320
• Growth: slow but consistent, roughly 15% month-over-month
• Status: profitable (covers its own hosting and API costs)

Grizzly Peak Software:

• Monthly traffic: ~5,500 unique visitors
• Newsletter subscribers: 380
• Revenue: modest (Amazon affiliate commissions, book referrals) — about $200/month
• Status: growing asset, not yet a significant revenue source

Content Pipeline:

• Internal tool, no direct revenue
• Time saved: approximately 14 hours per week
• Articles published: 80+ in four months
• Status: functioning well, occasional maintenance needed

Combined monthly revenue: ~$1,040

Is that a lot? No. It's less than I made in a single day of contract work. But here's what the revenue number doesn't capture.

I own all of it. Every line of code, every customer relationship, every piece of content. Nobody can fire me. Nobody can cancel my contract. Nobody can move the goalposts on a project I've invested months in. That ownership changes everything about how work feels.

---

What I Did Differently

Looking back, several things were different this time compared to every other attempt I'd made to build my own products.

I stopped optimizing and started shipping. My previous side projects died because I kept polishing them. I'd refactor code that worked fine. I'd redesign UIs that were good enough. I'd add features nobody asked for. This time, I shipped things that embarrassed me and then improved them based on what users actually needed.

I treated energy like a finite resource. This is the biggest lesson from burnout. You don't have unlimited willpower. You don't have unlimited focus. When you pretend you do, you borrow from tomorrow, and eventually tomorrow comes due. Six hours of focused work beats twelve hours of exhausted grinding every single time.

I built for markets I understood. AutoDetective works because I buy used vehicles and I know the anxiety of not knowing what you're getting. Grizzly Peak works because I've been a software engineer for thirty years and I know what people in this field actually need to learn. I stopped trying to build for imaginary users in markets I'd never participated in.

I said no to everything else. No freelance gigs. No consulting calls. No "quick favors" for friends that always turn into two-week projects. Every hour I spend on someone else's product is an hour I'm not spending on mine. This sounds selfish. It is. That's okay.

I moved to Alaska. I'm half-joking, but only half. Living somewhere with genuine quiet — where your nearest neighbor is a quarter mile away and the only notifications are from ravens — makes it dramatically easier to focus. The lack of social obligations, networking events, and coffee meetings freed up more time and mental energy than any productivity hack ever could.

---

The Mental Health Part

I almost didn't include this section because it feels too personal for a tech blog. But I think it matters.

At my lowest point, I genuinely questioned whether I'd chosen the wrong career. Thirty years of building software, and the thought of opening an IDE made me feel physically tired. That's a terrifying thing to confront when your entire identity and livelihood is built around programming.

The recovery wasn't linear. Month one felt like relief. Month two felt like anxiety — was I wasting my runway? Month three felt like momentum. Month four I hit a wall again when AutoDetective growth stalled and I spent a week convinced the whole thing was a mistake. Month five I found my rhythm. Month six I started feeling like myself again.

What I know now is that burnout wasn't about working too hard. It was about working on the wrong things for the wrong reasons. Client work drained me because I had no ownership over the outcome. My old side projects drained me because I was building them out of fear — fear of missing the next wave, fear of falling behind, fear of not being relevant.

The products I'm building now energize me because they're authentic. They come from my actual life and solve my actual problems. There's no performance involved. I'm not pretending to be a startup founder or a growth hacker or whatever the current trendy identity is. I'm just a guy in Alaska who builds useful things.

---

What's Next

The plan for the next six months is straightforward. Keep growing AutoDetective. Keep writing for Grizzly Peak. Keep improving the pipeline. Start exploring a fourth product idea — a homesteading management tool that I've been researching and prototyping.

I'm not setting revenue targets. I tried that before and it turned building products into a source of anxiety rather than satisfaction. Instead, I'm setting process targets: ship one improvement per week per product. Write two articles per week. Talk to five users per month.

If the revenue follows, great. If it doesn't, I'll adjust. But I'm not going back to fourteen-hour days and cold dinners. The cabin, the quiet, the six-hour workdays — those aren't rewards I'm earning. They're boundaries I'm maintaining. There's a difference, and it took me thirty years and a breakdown to learn it.

If you're in the middle of burnout right now, here's the only advice I have: stop. Not "optimize." Not "reprioritize." Stop. Give yourself enough time to remember why you started doing this in the first place. The code will wait. The opportunities will wait. Your brain won't wait forever.

Shane Larson is a software engineer, product builder, and recovering workaholic writing from his cabin in Caswell Lakes, Alaska. He runs Grizzly Peak Software, AutoDetective.ai, and a homestead where the chickens are unimpressed by all of it. Learn more at grizzlypeaksoftware.com.