Growth & Strategy

What "Do Things That Don't Scale" Actually Means (And Why I Stopped Following This Advice)


Personas

SaaS & Startup

Time to ROI

Short-term (< 3 months)

When I first heard Paul Graham's famous advice to "do things that don't scale," I thought I understood it. Hire people manually, do customer support one-on-one, write personal emails. Standard startup stuff, right?

Wrong. After working with dozens of SaaS startups and ecommerce businesses over the past few years, I've realized that most founders completely misinterpret this advice. They think it means "be inefficient on purpose" or "avoid automation at all costs." The result? Businesses that burn out their founders and never actually scale.

Here's what I've learned from watching startups succeed and fail with this approach: "doing things that don't scale" isn't about being manual forever. It's about building the right foundation before you automate. But there's a crucial timing element that nobody talks about.

In this playbook, you'll discover:

  • Why most startups misapply "do things that don't scale" and hurt their growth

  • The real strategic purpose behind manual processes in early-stage companies

  • How I helped clients transition from manual to automated without losing what made them special

  • A framework for deciding when to scale vs. when to stay manual

  • Real examples from SaaS acquisition and ecommerce optimization projects

Industry Reality

What every startup founder thinks they know

Walk into any startup accelerator or read any growth blog, and you'll hear the same advice repeated like gospel: "Do things that don't scale." The typical interpretation goes something like this:

The Standard Advice:

  • Manually recruit your first users one by one

  • Provide white-glove customer service to everyone

  • Write personal emails instead of automated sequences

  • Avoid building automation tools early

  • Focus on high-touch, high-effort solutions

This advice exists because it works—for the right reasons and at the right time. Early-stage companies need to understand their customers deeply, validate their assumptions, and build something people actually want. Manual processes force you to get close to your users and learn things you'd never discover through analytics alone.

But here's where most founders go wrong: they treat "unscalable" as a permanent philosophy rather than a temporary strategy. They think being manual is inherently virtuous, that automation equals losing touch with customers. Some even wear their inefficiency as a badge of honor.

The problem with this interpretation? It completely misses the strategic purpose. You're not doing unscalable things because they're better—you're doing them because you don't yet know what to scale. Once you figure that out, staying manual becomes self-destructive.

I've seen too many promising startups plateau because they couldn't make the transition from manual to scalable. They built great early traction but couldn't sustain growth. That's not startup success—that's startup purgatory.

Who am I

Consider me as your business complice.

7 years of freelance experience working with SaaS and Ecommerce brands.

The wake-up call came when I was working with a B2B SaaS client who had taken the "unscalable" advice too literally. They'd been manually onboarding every single trial user for eight months. Personal demo calls, custom setup sessions, individual check-ins. Their conversion rate was impressive—about 35% trial-to-paid.

But they were drowning. The founder was spending 6 hours a day on onboarding calls. Customer success was a bottleneck. They couldn't take on more trial users without hiring more people, but they couldn't afford more people without more customers. Classic startup death spiral.

When we dug into their analytics, something interesting emerged. The features they spent the most time explaining manually were the same ones that confused users in the product. The questions they answered repeatedly in personal calls were identical across 80% of users. The "personalized" setup they provided was actually following the same pattern every time.

They weren't doing things that don't scale—they were doing things that should scale but hadn't built the systems yet. There's a huge difference.

This pattern showed up again with an ecommerce client. They were manually managing customer support tickets, personally writing follow-up emails, and individually handling every return request. They thought this "personal touch" was their competitive advantage. In reality, it was preventing them from growing beyond $50K monthly revenue.

The breakthrough came when I realized we were asking the wrong question. Instead of "How do we stay personal as we grow?" we should have been asking "What did we learn from being personal that we can now systematize?"

My experiments

Here's my playbook

What I ended up doing and the results.

Here's the framework I developed after working through this challenge with multiple clients. I call it the "Learn, Systematize, Scale" approach, and it's based on understanding that manual processes are data collection tools, not permanent solutions.

Phase 1: Manual with Purpose (Months 1-3)

The key is being manual strategically. Every unscalable action should be an experiment designed to teach you something specific. For the SaaS client, we turned their onboarding calls into research sessions. Instead of just helping users get set up, we documented:

  • Which features caused confusion

  • What questions came up repeatedly

  • Where users got stuck in the product

  • What made successful users different from churned ones

Phase 2: Pattern Recognition (Month 2-4)

After 50+ manual onboarding sessions, clear patterns emerged. We discovered that 85% of successful users followed the same setup sequence. The most valuable part of the "personal" onboarding wasn't the customization—it was the forcing function that made users complete their setup immediately.

For the ecommerce client, we tracked every customer service interaction and found that 70% of inquiries fell into just 6 categories. The "personal" responses were actually templates in disguise.

Phase 3: Systematic Automation (Month 4+)

This is where the magic happened. Instead of scaling the manual process, we scaled the insights. For the SaaS client, we built:

  • An interactive onboarding flow that mimicked the successful manual sequence

  • Automated check-ins triggered by user behavior, not calendar schedules

  • A help system that surfaced answers to the most common questions contextually

  • Personal outreach reserved only for high-value users showing specific success signals

The result? Their trial-to-paid conversion rate actually increased to 42% while reducing founder time spent on onboarding from 6 hours to 30 minutes daily. They'd learned what mattered from the manual process, then built systems that delivered the same value at scale.

The ecommerce client saw similar results. We automated the 70% of support that was predictable while keeping human intervention for the 30% that required actual problem-solving. Customer satisfaction scores went up because response times improved, but they still felt "heard" when it mattered.

Key Insight

The goal isn't to avoid automation—it's to automate the right things based on what you learned manually.

Documentation Method

Track every manual interaction as data. Pattern recognition is impossible without systematic recording.

Transition Timing

Start building automation when you can predict 80% of interactions. Keep 20% manual for edge cases and learning.

Strategic Questions

Ask 'What did this teach us?' not 'How do we keep doing this?' Manual work should have an expiration date.

The results from this approach consistently surprised clients. The SaaS company I mentioned went from onboarding 20 trial users per month (their manual capacity limit) to 200+ trials per month within two months of implementing systematic automation. More importantly, their conversion rate improved because the automated system delivered the best parts of the manual experience consistently.

Revenue grew from $12K MRR to $45K MRR in four months—not just because they could handle more trials, but because they'd learned exactly what drove conversions and built systems around those insights.

The ecommerce client saw their customer support costs drop by 60% while Net Promoter Score increased by 18 points. They redirected their "manual" energy toward the 20% of customers who had complex issues, making those interactions even more valuable.

But the most interesting result was cultural. Once founders stopped spending all their time on repetitive tasks, they had bandwidth to work on actual growth challenges. They could experiment with new acquisition channels, improve their product, and build relationships with higher-value prospects.

The manual work hadn't been building competitive advantage—it had been preventing them from building it.

Learnings

What I've learned and the mistakes I've made.

Sharing so you don't make them.

Here are the seven key lessons I learned from helping startups navigate the "unscalable" transition:

1. Manual work is market research, not customer service. Every unscalable action should teach you something you couldn't learn any other way. If you're just repeating the same process, you're ready to automate.

2. Document everything, even when it feels obvious. Patterns only emerge when you track interactions systematically. Most founders rely on memory and miss crucial insights.

3. Automate the process, not the outcome. Don't just digitize your manual workflow—rebuild it based on what you learned about what actually drives results.

4. Keep a manual override for everything. Automation should handle the predictable 80%, but you need humans for edge cases and continuous learning.

5. Scale insights, not activities. The value isn't in doing things manually forever—it's in learning what matters so you can systematize it.

6. Time-box your unscalable experiments. Set a deadline for moving from manual to systematic. Otherwise, "learning" becomes an excuse for avoiding the hard work of building systems.

7. Measure leading indicators, not just outcomes. Track what behaviors predict success during the manual phase, then optimize for those behaviors in your automated systems.

How you can adapt this to your Business

My playbook, condensed for your use case.

For your SaaS / Startup

For SaaS startups implementing this approach:

  • Manually onboard your first 50 users with detailed documentation

  • Track which product features drive activation vs. confusion

  • Build automated onboarding that replicates successful manual patterns

  • Reserve human intervention for high-value prospects only

For your Ecommerce store

For ecommerce stores applying these principles:

  • Personally handle customer service for first 100 orders while documenting common issues

  • Create automated responses for predictable inquiries

  • Build systems that surface personal attention when it matters most

  • Focus manual effort on repeat customers and high-value segments

Get more playbooks like this one in my weekly newsletter