At customer number 1, support is a conversation. At customer number 100, support is a to-do list. At customer number 500, support is a job. At customer number 1,000, support is a system.

Most founders don't plan for this progression. They handle support personally until it overwhelms them, then scramble to set up systems under pressure. The scramble is expensive and stressful.

Here's how to handle each stage before it becomes a crisis.

Stage 1: 0 to 100 Customers

At this stage, you are support. Every email comes to your personal inbox. Every chat message pings your phone. You answer everything yourself.

This is correct. Don't hire a support person. Don't set up a help desk. Don't automate anything.

Why? Because every support interaction at this stage is product research. When customer #37 is confused by your onboarding flow, that's not a support ticket. That's your onboarding telling you it sucks. When customer #62 asks "can your product do X?" that's a feature request from someone willing to pay for it.

You need to feel this pain personally. Reading a monthly support summary is not the same as reading the actual messages from confused, frustrated, excited customers.

What to build at this stage: nothing. Answer emails from your inbox. Keep a list (spreadsheet, Notion, sticky notes) of recurring questions and feature requests. That's your future FAQ and product roadmap.

Time investment: 30 minutes to 1 hour per day.

Stage 2: 100 to 300 Customers

Volume is picking up. You're getting 5 to 15 messages per day. Some are repetitive. You're starting to copy-paste the same answers.

This is when you build the basics.

Write an FAQ page. Not 50 questions. The 10 questions you've answered most in the last month. Put it on your website. Link to it in your welcome email. This cuts 20 to 30% of your incoming volume.

Create saved replies. The 5 to 10 responses you type most often become templates. Gmail has canned responses. So does every email client. You're not automating anything. You're just typing less.

Set up a dedicated email address. Move from your personal inbox to support@company.com. Forward it to your inbox if you want, but give customers a consistent point of contact.

Start tagging tickets by topic. Even a simple spreadsheet column noting "billing," "bug," "how-to," or "feature request" gives you data about what your customers actually need. This takes 5 seconds per ticket and pays off later when you're deciding what to automate.

Still don't hire. At 10 messages per day, support takes 1 to 2 hours. You can handle it. And you should, because the product feedback is still directly informing your decisions.

Time investment: 1 to 2 hours per day.

Stage 3: 300 to 500 Customers

Now it's getting real. 15 to 30 messages per day. Some are complex (integration issues, billing disputes, bugs). The FAQ handles the simple stuff but the volume still grows.

This is the decision point: do you hire, or do you automate?

If your product is complex and your customers need hand-holding (B2B software, technical products), hire a part-time support person. They handle the routine stuff. You handle escalations and complex issues.

If your product is simple and most questions are repetitive (e-commerce, consumer app, SaaS with a clear use case), set up AI classification. Supp costs $0.20 per message with no base fee. At 20 messages per day, that's $120/month. A part-time support hire costs $1,500 to $2,500/month.

AI doesn't replace you at this stage. It triages. Simple questions get auto-responses. Complex ones get routed to you with the context already gathered. You spend less time on the easy stuff and more on the stuff that matters.

Set up a simple help desk if you haven't already. Shared inbox at minimum. Something with ticket tracking if you can.

Time investment: 2 to 3 hours per day (less with AI handling the simple queries).

Stage 4: 500 to 750 Customers

You need a dedicated support person. Part-time or full-time depends on your volume and complexity.

At this stage, build the systems your support person needs:

A runbook covering the top 20 ticket types. What to do for each one, step by step. Your support person shouldn't have to ask you how to handle a refund request.

Basic metrics. Track tickets per day, response time, and resolution time. You need to know if things are getting better or worse.

An escalation process. Your support person handles the straightforward stuff. Complex issues, VIP accounts, and technical bugs come to you or the relevant team member. Define what triggers an escalation.

Quality checks. Review 5 to 10 tickets per week. Are the responses accurate, clear, and on-brand? Catch problems early before they become habits.

Time investment for you: 30 minutes per day reviewing escalations and quality checks.

Stage 5: 750 to 1,000 Customers

You're now handling 30 to 60+ messages per day. Your support person (or AI + you) is handling the volume, but you need to think about what happens next.

This is where you professionalize support:

AI classification becomes essential if you haven't set it up already. At 40+ messages per day, manually triaging is a waste of human time. AI classifies and routes. Your team handles the rest.

Build your knowledge base for customers. Go beyond the FAQ. Write help articles for common scenarios. Link to them from your product (contextual help). Good self-service deflects 30 to 50% of volume.

Invest in proactive support. Onboarding emails that answer questions before they're asked. In-app tips for common confusion points. Release notes that explain changes before customers encounter them and panic.

Consider your second support hire if volume justifies it. Usually around 50+ tickets per day, or if your first hire is consistently working overtime.

Track CSAT. You now have enough volume to get meaningful satisfaction data. Start measuring it and acting on the results.

Time investment for you: 15 to 30 minutes per day on escalations and weekly reviews.

The Common Mistake at Every Stage

At every stage, the mistake is the same: waiting too long to build the system for the next stage.

Founders who handle support personally until customer 300 burn out. The quality drops. Customers notice. By the time they hire, the backlog is enormous and the new hire inherits a mess.

Founders who hire at customer 50 spend money they don't need to spend and miss the product feedback they should be getting directly.

The right cadence: build the system for your current stage, and start planning for the next stage when you're at 70% of the trigger point. If you plan to hire at 500 customers, start writing the runbook and thinking about candidates at 350.

Support scaling isn't a crisis management exercise. It's a planning exercise. And the plans are simple. FAQ at 100. Saved replies at 200. Help desk or AI at 300. First hire at 500. Systems at 750. Professionalization at 1,000.

Each step builds on the last. And if you build each one at the right time, you never hit the "everything is on fire" moment that most founders associate with support scaling.