How a 5-Person Team Ran Post-Sales for 100+ Customers: Lessons Learned

There is a stage in every startup's life when customers arrive faster than the team can look after them. The product is selling, the founders are still building, and nobody's full-time job is making sure a new customer gets value in their first month. The usual answer is to hire. Some teams find another one.

Octolane, a Y Combinator W24 company building a self-driving CRM, crossed 100 customers with a five-person team and no customer engineering hire. Mycroft, a governance, risk and compliance platform, scaled its book of accounts three to four times without adding headcount. Both did it by treating post-sales as a system of workflows rather than a list of tasks for people to remember.

This article pulls the lessons out of those two Octolane and Mycroft customer stories and turns them into an operating model any lean team can copy: when to change approach, what to automate first, and where people should still spend their time.

The Breaking Point: When Manual Post-Sales Stops Working

Octolane's founding team works seven days a week on the product, and both founders are engineers. There was no dedicated customer engineer and no onboarding specialist. At fewer than 50 customers, the team could keep up manually. Once they crossed that mark, things started to break.

That pattern is common. Manual post-sales works while the founders know every account by name. It stops working when nobody can hold the whole book in their head, and the symptoms tend to look the same:

• New customers wait for a kickoff because nobody noticed the contract was signed.
• Setup steps stall quietly, and the team only finds out when the customer complains or goes silent.
• Check-ins happen with whoever shouted loudest, not with the accounts that need them most.
• Expansion happens by accident, when a customer asks, rather than at the moment they are ready.

At that point the choice, in Octolane's words, was clear: hire a customer engineer, or find a better way. The lesson is to recognise the breaking point early. The window between "too small to need customer engineers" and "too large to do it manually" is short.

Lesson 1: Treat Post-Sales as a Hire You Configure

Octolane didn't add a tool to a team. It brought on Quivly to act as its entire post-sales function, deploying agents to handle everything from the moment a customer signs to well past go-live. Today, the case study reports, 100% of its post-sales motion is covered by Quivly workflows.

The mindset shift matters more than the software. Instead of asking "which tasks can we automate?", ask "what would a great post-sales hire do every week, and which of those things can run on their own?" For most lean teams that list covers four areas:

• Onboarding: getting each new customer set up and to first value.
• Health monitoring: spotting accounts that are drifting before they leave.
• Expansion: noticing when an account is ready to grow.
• Reviews and check-ins: preparing the context for every customer conversation.

Those four areas are exactly where Octolane pointed its workflows. Each one is covered below.

Lesson 2: Make Onboarding a Workflow, Not a Calendar Reminder

At Octolane, when a new customer signs, Quivly kicks off a structured onboarding workflow: sending the right resources, checking in at the right moments, and flagging accounts that haven't completed key setup steps. The case study reports that onboarding completion rates improved significantly in the first month.

Mycroft attacked the same stage from the handoff side. Before Quivly, field engineers pieced together context from sales calls by hand, so onboarding started slowly and anyone who picked up an account later started from scratch. Now, the moment a deal closes, Quivly builds a handoff notebook that pulls the sales calls, emails, recordings and CRM data into one place and extracts the company summary, security maturity, pain points, prior tool migrations and tech stack.

To copy this, define three things before you automate anything:

1. The trigger: a closed-won deal or a signed contract.
2. The handoff: the fields every new owner needs, captured automatically from the sales record.
3. The milestones: the setup steps that predict success, with a nudge when an account falls behind. Our guide to onboarding milestones covers how to choose them.

Mycroft also runs each customer's framework implementation as a project, and Quivly's agents spin up a new project with tasks and milestones whenever a customer opts into another framework. No one builds it by hand. The case study credits the overall change with 50% faster onboarding-to-value.

Lesson 3: Let Health Monitoring Find the Problems for You

A small team can't review every account every week. It can respond to a short list of the accounts that need attention. At Octolane, Quivly tracks product engagement signals across the customer base and surfaces at-risk accounts before a customer ever sends a cancellation email, so the team gets a clear picture without digging through usage data.

The lesson is to invert the work. Instead of people searching for risk, the system finds it and people respond. In practice that means:

• Agree on the handful of engagement signals that matter for your product, such as logins, key feature use or a stalled integration.
• Combine them into a health score that updates continuously.
• Route a flagged account to a named owner with the reason attached, rather than to a shared dashboard nobody checks.

Mycroft adds an external layer: Quivly monitors news across its customers for signals that matter to a compliance business, such as a new funding round, a new CISO, a disclosed breach, an acquisition or expansion into a new market. Each becomes a reason to reach out with context rather than something an engineer stumbles on weeks later.

Lesson 4: Tie Expansion to Milestones Customers Actually Hit

Lean teams rarely have time for speculative upsell calls. The answer is to start expansion conversations when the customer has given a clear signal. At Octolane, as customers hit milestones such as connecting their email, getting their first AI-updated deal or hitting their seat limit, Quivly identifies the moment and prompts the right follow-up. The case study notes that expansion conversations now happen systematically, not by accident.

To build this for your own product, list the moments that typically come before a customer buys more. Seat limits, usage thresholds, a new team adopting the product and a successful first outcome are common examples. Attach a play to each one, with the context the account owner needs already in hand. Quivly's Plays work this way: a named motion with a trigger, steps and a channel, started when the signal hits.

Lesson 5: Automate the Prep, Keep the Conversation Human

Founders and senior engineers are often the most valuable people a customer can talk to, and the most expensive to spend on preparation. Both case studies keep people in the conversation and move the prep elsewhere.

• Reviews: for higher-touch accounts, Quivly preps the summaries, pulls the relevant activity and surfaces talking points, so Octolane's founders aren't starting from scratch when they join a call.
• Follow-ups: at Mycroft, every customer call logs automatically with a structured summary, action items and sentiment, and agents draft the follow-up emails and notes. The case study reports 8 to 10 hours saved per field engineer each week.
• Visibility: Mycroft's leadership now sees ARR and customer count by field engineer, milestone and time-to-value tracking, and upcoming renewals in live dashboards instead of spreadsheets.

The result, in Mycroft's case, is three to four times more accounts managed per field engineer, with that time going into audit readiness, executive relationships and compliance milestones instead of typing notes and chasing data.

A Lean Post-Sales Operating Model You Can Copy

Put the lessons together and a simple model emerges for a small team running a large book:

1. Unify the data first. Workflows are only as good as the account record they read. Connect the CRM, product usage, billing, support and calls into one view.
2. Automate every trigger. Signed contracts, stalled milestones, health drops, usage thresholds and renewal windows should each start a workflow without anyone remembering to.
3. Send people only the exceptions. The team's queue should be the short list of accounts where a human makes a difference, with context and a draft already attached.
4. Keep humans on the relationship. Use the time saved for the calls, decisions and judgement calls only people can make.
5. Measure it. Track onboarding completion, time-to-value, at-risk accounts caught early and accounts per person, so you know when the model needs adjusting.

As Octolane's co-founder put it: "We're a team of five, and we're moving like a team of fifty."