There’s a chart floating around that bears are going to love.

Revenue per employee.
$LMND vs. $ROOT vs. seven major incumbents.
2020–2025.

And at first glance?

It’s ugly.

In 2025:

Dead last.

Bears stop reading here.

They shouldn’t.

$LMND acquired 3 million customers in:

The cheapest policies in insurance.

And they were also the cheapest customers to acquire.

Young. Digital-native. First-time buyers.
No incumbent agent was cold-calling 25-year-olds for a $15/month renters policy.

So yes — revenue per employee started low.

Because they intentionally filled the system with low-ticket policies.

But now look at the curve.

$90K ➝ $576K in five years.

That’s a 6x improvement.

And it’s accelerating.

2024 revenue per employee growth: 25%
2025 revenue per employee growth: 35%

Now compare that:

$LMND is accelerating into the band.
Everyone else is decelerating inside it.

At this rate, $LMND enters the incumbent band by 2027.

That’s not hype. That’s slope math.

Those 3 million customers?

They’re aging.

They’re buying:

The most profitable insurance products on earth.

And every cross-sell happens at near-zero CAC.

AI Maya already has their data.
No new marketing spend.
No new agent.
No new employee.

That’s where this gets interesting.

Car IFP grew 53% to $187M.
Over half of new car sales came from existing customers.

Loss ratio:

Scaling faster. Getting more profitable.

That combination is rare.

Today average revenue per customer is roughly ~$400.

If that moves to $800+ over time?

Revenue per employee doesn’t just improve — it explodes.

Because $LMND already has a massive customers-per-employee advantage.

When monetization catches up to scale, the efficiency curve flips fast.

Chart 1: Revenue per employee (they look broken).
Chart 2: Customers per employee (they look dominant).

The market is staring at Chart 1.

I’m modeling what happens when Chart 2 monetizes.

No.

$ROOT is the cleanest comparison. Same era. Same DTC P&C model. Public. Clean data.

2025 customers per employee:

$ROOT is already in the incumbent band.

Why?

Because telematics helped them price better.

It didn’t reduce how many humans they need to operate.

Last year:

That’s not incremental improvement.

That’s architectural difference.

$ROOT used tech to price better.

$LMND used AI to run the company.

AI Maya underwrites.
AI Jim handles claims.

The marginal customer costs almost nothing to serve.

And every new customer trains the system.

The moat isn’t just AI.

It’s 3 million customers training the AI.

I started buying $LMND at $30.

Not because it was trending.

Because the underwriting model and AI architecture were structurally different.

Then loss ratios blew up.

The market punished it.

Fair.

But now?

The math is different.

And I’ve recently started adding again.

Not chasing strength.

Adding into what I believe is structural mispricing.

Gross profit growth: 73.5%
Annualized gross profit: $444M
Enterprise value: $4.04B

EV / Gross Profit / Growth = 0.13x

Compare:

That makes $LMND the second cheapest in a basket I already consider undervalued.

You don’t often see 70%+ gross profit growth trading like this.

IFP: $1.273B, growing 31%

Assume 25% CAGR for 10 years.

→ $11.6B IFP

Assume stabilized GLR of 70%.

→ $4.06B gross profit

Assume operating margin midpoint of 13.5%.

→ ~$1.56B operating profit

Apply a 30x multiple.

→ ~$47B valuation

Assume 1.5% annual dilution → ~81.2M shares.

→ ~$579 per share.

Long way away.

But the math works.

And it works using conservative assumptions.

The market is still debating survival.

I’m modeling scale.

When sentiment and math disconnect, that’s where I press.

I started buying at $30.
I’ve recently started adding again.

If you want to see exactly how I’m positioning my portfolio in real time — including buy and sell notifications as they happen — subscribe below.

When I move, subscribers see it.
Before the market does.

— Connor
Alpha Before It Prints

🔗 Click to Connect with me

© 2026 Alpha Before It Prints
Unsubscribe