Building Your First Customer Success Function From Seed to Series A

Scott Weinstein

I've built the first CS function at three different companies. Each time, I walked into the same situation: a product that customers were paying for, no formal process for making sure those customers succeeded, and a growing sense that something needed to change before churn became a crisis.

The temptation when you're building from zero is to try to build everything at once. Hire a team. Buy a CS platform. Create playbooks for every lifecycle stage. Design a health score. Launch a QBR program. The reality is that most of this can wait, and trying to do it all in the first 90 days guarantees that none of it gets done well.

Here's the sequence that's worked for me, organized by what to build first and what to defer.

Month 1: Talk to every customer

Before you build anything, you need to understand what's actually happening with your customers. Not what the founder thinks is happening. Not what the sales team reported during the handoff. What's actually happening.

Schedule a 30-minute call with every active customer. If you have 200 customers, prioritize the top 30 by revenue and a random sample of 20 others. Ask three questions: What made you buy this product? What's working? What's frustrating? Take detailed notes and look for patterns.

This feels slow when you're under pressure to "build the function." But those conversations will save you months of building the wrong things. At one company, I discovered during these calls that the biggest churn driver wasn't product issues at all. It was that customers didn't understand a specific configuration step, and nobody had ever shown them how to do it. We fixed that one gap and retention jumped immediately.

Month 2: Build the onboarding process

Of everything you could build first, onboarding has the highest ROI. A customer who gets to value quickly is dramatically more likely to retain than one who struggles through the first 30 days. And unlike most CS initiatives, onboarding improvements compound. Every new customer who goes through a better onboarding process is one less at-risk account you'll deal with six months from now.

Your first onboarding process doesn't need to be sophisticated. Define what "first value" looks like for your product. Map the steps to get there. Build a checklist. Assign yourself to walk every new customer through it personally. Track how long it takes and where people get stuck.

Month 3: Create a health signal, not a health score

A full health scoring model is premature when you have limited data and no historical churn patterns to calibrate against. But you need something that tells you which customers need attention right now.

Start with a single leading indicator. Based on your customer conversations and onboarding data, identify the one behavior that most strongly correlates with success. It might be weekly active users. It might be completion of a specific workflow. It might be engagement with a key feature. Pick one signal, set a threshold, and create a weekly report that shows which customers are above and below it.

This isn't a health score. It's a triage tool. It tells you where to focus your limited time. As you gather more data, you can add signals and build a proper weighted model. But in month three, one reliable signal is better than five unreliable ones.

Months 4-6: Build the expansion motion

Most early-stage CS functions focus exclusively on retention. This is understandable but incomplete. Retention keeps your existing revenue. Expansion grows it. And at an early-stage company, net revenue retention above 100% is what convinces investors that you have product-market fit, not just product-interest fit.

The expansion motion at this stage is simple. Identify customers who've exceeded the "success" threshold you defined in your onboarding process. They're using the product heavily. They're getting value. They're the natural candidates for upselling to a higher tier, adding more seats, or expanding to a new use case.

You don't need a formal expansion playbook yet. You need to build the habit of looking at your healthiest accounts and asking: what else could we help them with? The playbook comes later, once you've done it enough times manually to understand what works.

What to defer until after Series A

There's a long list of things that are important but not urgent in the first six months. A CS platform (a spreadsheet is fine until you have 5+ CSMs). Formal QBR templates (have conversations, not presentations). Customer advisory boards (you don't have enough customers yet). Detailed segmentation models (you'll rebuild these three times anyway). Automated lifecycle emails (personal outreach converts better at low volume).

All of these things become necessary eventually. But building them before you have the volume to justify them means you'll spend time maintaining systems instead of talking to customers. And in the first six months, every hour you spend with customers is worth more than every hour you spend in a spreadsheet.

The sequence matters more than the speed

The founders hiring you want results fast. The board wants to see NRR numbers next quarter. The pressure to build everything at once is real. But the companies I've seen build the most durable CS functions are the ones that followed a deliberate sequence: understand first, onboard second, monitor third, expand fourth.

Each step builds on the one before it. You can't build a good onboarding process without understanding your customers. You can't build a useful health signal without onboarding data. You can't identify expansion opportunities without knowing which customers are healthy. Skipping steps doesn't save time. It creates rework.

The goal for your first six months isn't to have a fully built CS function. It's to have the foundation that a fully built CS function can scale on top of. Get the sequence right, and everything else follows.