Failure changes at each milestone so investors trust you

An investor asks, "So, how's it going?"

You say, "Pretty good, we're still learning," but you have no idea if you're actually failing.

Most early founders don't lack progress. They lack a clear definition of failure for where they are right now.

The Core Insight

Failure is not "the company died." Failure is "the main risk of this stage did not get reduced."

And that definition changes at every milestone. If you don't update it, you'll either quit too early or drag a dead thing around for years.

Stage 1: Idea – Can This Matter to Anyone?

At idea stage, the only question is: Is this a real pain for real people?

If you're not careful, you'll define failure here as "no one likes my solution." Wrong milestone. You don't have a solution yet. You're just trying to prove the pain is sharp enough that people already hack around it.

What failure really is at idea stage:

You talk to 30–50 relevant people. You do not hear real workarounds like "I export this every week and clean it in Excel for 2 hours." You don't hear real costs like "we pay a contractor just to keep this updated." And you don't hear real urgency like "if this broke, we'd scramble today."

You can't get anyone to say, "If you solved this, I'd want to try it."

If that's true after honest interviews, the idea has failed. Not you. This specific bet on a problem is wrong.

Make it objective:

Before you start, write: Kill/pivot if "After 30 interviews, fewer than 5 people show real pain and urgency." Continue if "10+ people say they'd try a fix and can explain why in their own words."

No revenue. No product. Just pain vs. no pain.

Stage 2: MVP / Launch – Do People Use This Without You Pushing?

Once you've found a painful problem, the question changes to: Will people use this thing to solve it?

This is where founders hide behind vanity: signups, launch upvotes, nice tweets. None of those mean survival.

At MVP / launch, failure = no real usage.

Concretely: People sign up but don't come back. They try it once, then go back to their old hack. You have to personally chase every user just to get a login.

Better definitions: For a SaaS tool, success is >30–40% of new users are still active after 4 weeks. Failure is <15% 4-week retention after multiple product attempts. For a consumer app, success is at least 20–25% of users open the app weekly. Failure is everyone disappears after day 2.

Pick one main number.

For example: "Kill/pivot if after 3 months and 100 signups, <20% are still active after 4 weeks." "Continue if ≥30% are still active and usage is growing without bigger incentives."

That's how you avoid "we just need more marketing" when the product simply isn't sticky.

Stage 3: Early Growth – Do The Numbers Work At All?

Once something is sticky for some people, the question becomes: Can we grow this without going broke?

Here, failure is not "we're not huge yet." Failure is: We can't grow users or revenue without lighting money on fire.

Mechanically this is about unit economics: CAC = Customer Acquisition Cost and LTV = Lifetime Value (gross profit from a customer over their life).

You don't need perfect precision, but you need to know directionally: If you spend $1,000 on ads / outbound / sales to acquire users, what do you make back in 12–24 months?

Rough benchmarks for SaaS-ish models (will vary): Healthy-ish early stage means LTV is at least 3x CAC on a gross margin basis. Scary is you can't even estimate LTV because retention is too weak or churn is too high. Also scary is you can't acquire customers at any scale without CAC blowing up.

Define failure here as:

"After testing 2–3 channels with real spend, we cannot get CAC under 1/3 of our best-estimate LTV." Or "Logo growth is flat for 3–6 months and every new user is as hard or harder to get than the last."

Define success as:

One or two channels where you can put in $1 and reliably get $3–5 in gross profit back over a reasonable time frame. And retention is strong enough that revenue grows even without increasing acquisition every month.

If this doesn't exist, you don't have a growth machine.

You have a product some people like.

Stage 4: Scale – Can This Run Without Breaking?

If you pass early growth, the game flips again: Can we scale without the company tearing itself apart?

Now failure is not "we missed target MRR." Failure is: execution or operations break faster than you can fix them.

Signals: Customer experience collapses with support tickets piling up, NPS tanking, churn spiking. The team can't ship or coordinate; everything is a fire drill. Margins compress as you grow, and every new dollar of revenue is less profitable.

At this stage, your kill/pivot line is usually about model vs. reality: "If at $X ARR we can't maintain Y% gross margin and Z% net dollar retention, the model is broken." Or "If we can't hire and onboard role R in <N weeks without chaos, this org design doesn't scale."

Here, "pivot" might mean changing the go-to-market model, the pricing, or even narrowing the customer segment so operations can keep up.

How To Actually Use This

A simple operating rhythm:

1. Name the stage you're in. Not aspirationally. Honestly. Are you still proving pain? Proving usage? Proving economics? Proving execution?

2. Write a 1-line failure definition for this stage. Idea: "If 30 interviews don't yield real, urgent pain, we drop this problem." MVP: "If 100 users produce <20% 4-week retention, we rethink the product." Growth: "If we can't reach 3x LTV:CAC on any channel, we don't scale spend." Scale: "If margins + retention degrade as we grow, we pause growth to fix ops."

3. Set a date to judge, not debate. Put it on the calendar: in 6–12 weeks, you either kill, pivot, or double down based on the metric you chose before emotions and sunk cost kicked in.

4. Review with evidence, not vibes. Customer interviews, retention charts, revenue, margins, fundraising reality, team capacity. No "it feels like we're close."

The "So What?" Conclusion

Your definition of failure is your steering wheel.

If you keep using the same definition from idea to scale, you'll be driving with it locked in place. Pick your stage. Write down what failure actually means right now. Put a number and a date next to it.

Then execute until you hit one of three outcomes: kill, pivot, or continue.

‍