Day 47: Why We Killed the Sales Team

San Francisco · November 10, 2025

On a Thursday in September, the CEO walks into the growth team's area and says, "Can we talk?" She doesn't say about what. She doesn't have to. I've seen the sales numbers. I know what's coming.

We sit in the small conference room — the one without windows, the one people use for hard conversations. She closes the door and says it plainly: "I'm thinking about cutting the sales team."

Not downsizing. Cutting. All six of them. The VP of Sales, three AEs, and two SDRs. The entire outbound motion, gone.

"Tell me why," I say, even though I already know.

The sales team costs $1.14 million annually, fully loaded. Salaries, commissions, tools (CRM license alone is $3,200/month), travel, the whole thing. In the last twelve months, they've closed $890,000 in new ARR. That's a negative ROI — we're spending $1.14 million to generate $890,000. You don't need a finance degree to see the problem.

But the raw numbers don't tell the whole story. The sales team isn't just closing deals; they're running discovery calls, doing demos, managing a pipeline. The question isn't just "are they profitable?" It's "what happens to the pipeline if they're gone?"

I pulled the data. Over the past year, the sales team ran 847 discovery calls. Of those, 312 progressed to demo. Of those, 89 submitted a proposal. Of those, 43 closed. Average deal size: $20,698 in ARR.

Now here's the uncomfortable part. During the same period, our self-serve signup flow generated 2,340 paying customers with an average deal size of $1,620 in ARR. Total self-serve revenue: $3.79 million. Cost of the self-serve infrastructure (product, marketing, support): approximately $640,000.

The sales team's CAC: $26,512 per customer. The self-serve CAC: $274 per customer.

The sales team's CAC was $26,512 per customer. The self-serve CAC was $274. I ran the numbers three times because I didn't want to believe them.

"But the sales customers are higher value," the VP of Sales will argue when he's told. And he's right — $20,698 versus $1,620 in average ARR. The question is whether those higher-value customers can be acquired through a different motion at a lower cost. And the answer, increasingly, is yes.

The layoffs happen on a Monday. I hate that it's a Monday. The VP of Sales — I'll call him Kevin — finds out at 8:30 a.m. He's been at the company for two years. He built the sales playbook from scratch. He hired every person on his team personally. He takes it with the kind of quiet dignity that makes me feel worse, not better.

"Is this a growth call or a cost-cutting call?" he asks the CEO.

"Both," she says. "The numbers don't work, Kevin. You know that."

"The numbers would work if we gave the team more time. We're just starting to see pipeline velocity improve. The Q4 pipeline is the best we've had."

He's not wrong about the pipeline. But pipeline is a leading indicator, and the board doesn't have patience for leading indicators when the lagging ones look like ours. We need to get to profitability, and $1.14 million in sales costs is the biggest lever we have.

I don't say any of this to Kevin. It's not my conversation to have. But I sit in the room because the CEO asked me to, and I remember every word.

Twelve days in, and the absence is physical. The sales team used to occupy the east side of the office. Now it's empty desks and orphaned monitors. Someone turned off the TV that used to display the sales leaderboard. The CRM still sends daily pipeline reports to the leadership Slack channel. Nobody's turned them off yet, so every morning I see a pipeline update for a team that no longer exists.

The practical problems start immediately. We have 34 open opportunities in the pipeline — deals that were in various stages of conversation when the sales team was cut. Someone has to own them. That someone is me.

I've never closed a deal in my life. I've run growth teams, built funnels, optimized landing pages, analyzed cohort data until my eyes bled. But I've never sat across from a VP of Operations at a mid-market company and asked them to sign a contract. The skills are related but different, like the difference between a war correspondent and a diplomat.

Of the 34 open opportunities, I manage to close 8 in the first month. Total ARR: $142,000. The other 26 go cold. Some we could have saved with a real salesperson. Some were never going to close anyway. I'll never know the exact split, and that uncertainty bothers me more than the losses.

We need a new motion. You can't just remove the sales team and expect the revenue to replace itself. PLG — product-led growth — is the strategy, but PLG isn't a switch you flip. It's a complete rethinking of how your product acquires, activates, and monetizes users.

Here's what we build in the first month:

An in-product upgrade flow. Previously, upgrading required talking to a salesperson. Now, you click a button, pick a plan, enter a credit card. It sounds obvious, but we'd never built it because the sales team wanted to control the upgrade conversation. "If they talk to us," Kevin used to say, "we can upsell them to a higher tier." True. But we were also creating friction for the 70% of customers who just wanted to pay and move on.

A usage-based expansion trigger. When an account hits 80% of their plan limits, they get an in-app notification and an email offering a one-click upgrade. The notification includes specific data: "You've used 412 of 500 contacts. Teams your size typically need our Professional plan."

A self-serve enterprise tier. This was the controversial one. Enterprise deals were always done through sales — custom pricing, negotiated terms, security review. We standardized the enterprise plan at $799/month with a fixed feature set and a self-serve checkout flow. No negotiation, no custom terms, no three-month procurement process.

It's been 47 days. Here's what I know:

Self-serve revenue is up 23% month-over-month. Some of that is organic growth that would have happened anyway. Some of it is the new upgrade flows. I'd estimate 60/40.

The self-serve enterprise tier has 11 customers. Average ARPU: $799/month (obviously — it's the only price). Three of those customers were in the sales pipeline before. The other eight are new. That's $8,789 in new MRR that we didn't have a month ago, acquired at near-zero marginal cost.

Total MRR: $2.94 million, down from $3.02 million before the sales team was cut. The gap is closing faster than I expected.

Customer feedback is mixed. Some customers — particularly larger accounts — have told us they miss having a dedicated person to call. One wrote: "I used to have Kevin's cell number. Now I have a help desk." That stung.

We hired two customer success managers to handle the relationship side. They don't sell — they support, troubleshoot, and identify expansion opportunities. It's a different skill set and a different cost structure. Total annual cost: $240,000, versus $1.14 million for the sales team.

"I used to have Kevin's cell number. Now I have a help desk." That stung. It still stings.

I think about Kevin. He landed at another startup three weeks after we let him go, which I know because LinkedIn exists and I couldn't help checking. I hope he's doing well. I hope the new company gives his team the runway we didn't.

I think about the 26 pipeline deals that went cold. Each one represents a company that might have become a customer, a relationship that might have turned into revenue, a story that might have ended differently if we'd had six more months.

I think about whether we made the right call. The numbers say yes — or they will, if the trends continue. But numbers don't capture the human cost. They don't capture Kevin's face in the conference room, or the empty desks, or the leaderboard TV that's still turned off.

The CEO asked me last week if I'd do it again. I said yes. But I also said this: "Next time, we don't let it get to the point where cutting is the only option. We build the PLG motion first, and scale sales only when the unit economics support it."

She agreed. Whether we'll remember that lesson when the next growth lever looks shiny and the board is asking for faster revenue, I don't know.

Day 47. The self-serve machine is running. The numbers are moving in the right direction. The office is quieter than it used to be.

I'm not sure if that's a good thing or just a different thing.