# Staying

or: the only skill that never shows up in a launch thread.

By Jon C. Phillips · October 3, 2026 · https://joncphillips.com/staying/

Everyone's obsessed with starting. Nobody wants to talk about the boring middle part where most projects quietly die and take your confidence with them.

In 2012 I started a photography site, and for the first two years it made basically no money.

**And it wasn't for lack of trying.** We were publishing three or four times a week, and I was paying freelance writers and editors out of my own pocket to keep that pace.

The traffic actually showed up pretty quickly, which somehow made it worse. I had real money going out and decent numbers on the dashboard, but almost none of it turned into income.

Somewhere in year 3, with growth flat and a notes file full of better ideas, I came very close to shutting the whole thing down.

I stayed, mostly out of stubbornness I'd love to reframe as wisdom.

That site was [Contrastly](https://contrastly.com). It passed 100,000 subscribers before I sold it in 2021, and the [newsletters that grew out of that](https://emailgrowthguide.com) same decade still pay me every month, fourteen years after I published a post that nobody read.

There was no secret. I was just still there when the slow years finished doing their thing.

**The staying was the entire trick, and it's about to become the only trick left.**

## Starting Got Free

Getting something off the ground has never been easier.

Stripe Atlas data shows the share of new companies reaching a paying customer in their first month [jumped from 8% to 20% since 2020](https://stealwhatworks.com/blogs/news/indie-hackers-going-extinct).

AI writes the boilerplate and the landing page takes an afternoon, so you can go from idea to charging money over a weekend. Thousands of people do exactly that every week.

That's great, but look at where all those products actually end up.

| Where indie products land | Share |
| --- | --- |
| Exactly $0 | 54% |
| Under $500/month | \~70% |
| $1k–$5k/month | \~18% |
| Over $10k/month | \~3% |
| Over $50k/month | 1–2% |

These come from different datasets, so don't add the column up. The shape is what matters, and the shape is brutal. That top row comes from [937 Stripe-verified products](https://www.buildmvpfast.com/blog/saas-only-way-out-indie-hacker-dream-desperation-2026) on Indie Hackers, and over half of them never made a single dollar.

The median product across everything listed makes [about $30 a month](https://www.jenariusganlary.com/blog/how-much-indie-hackers-actually-make).

So the hard part moved. It used to be hard to start and hard to continue.

Now twenty people get a paying customer in month one, and the market asks all twenty the same question every month for years.

**Are you still here?**

## The 26-Project Guy

There's a founder who went viral for shipping [26 projects in four years](https://www.indiehackers.com/post/4-years-26-projects-115k-lessons-from-an-indie-hacker-7ab46733da) and making $115K. Everyone read it as a persistence story. Look at all that hustle.

The actual number was sitting right there in his own breakdown. His average project got abandoned in under two months.

And his biggest winner came in year four, from the one thing he kept working on.

If you ask me, it's really a story about staying with one of them.

## Two Rooms That Look Identical

People usually leave for a pretty understandable reason.

Every project that survives its launch eventually hits a stretch where growth stalls and nobody says anything nice about it. The feature you were proudest of sits unused in your analytics.

Support tickets keep showing up. The work turns into maintenance and iteration, which is a polite way of saying it turns boring.

That description fits two completely different situations.

One is a healthy project in its unglamorous middle, the long flat stretch that comes before basically every outcome worth having. The other is a project that already died and hasn't told you yet.

**From the inside they feel the same.**

They both feel like failure, and they both make the new idea in your notes app look absolutely radiant, because unbuilt things always do.

**So people quit healthy projects constantly, in enormous numbers, because they read boredom as a verdict.**

I've done it. I've also seen the other version enough times to trust it now. The photography audience I almost killed in its flat third year went on to fund a decade of my life. The verdict I nearly wrote was wrong, and the only reason I know is that I didn't write it.

Now don't get me wrong, quitting is a legitimate move. Some projects deserve a shutdown email and I've sent a few.

**The mistake is quitting without knowing which room you're standing in.**

And to be clear, this is a different problem from [living in the maybe zone](https://joncphillips.com/living-in-the-maybe-zone/). The maybe zone is not knowing whether the thing works. This is the stage after that, when the numbers quietly told you it works, slowly, and you want to leave anyway. Uncertainty breaks some builders.

**Boredom gets most of the rest.**

## Reading the Room

Your feelings won't help much here. Boredom shows up in both rooms with identical intensity, so stop consulting them.

What differs is what the project does when you stop pushing it.

Open your analytics before you keep reading, because you'll answer from memory otherwise, and memory always votes for whatever you already wanted to do.

You're probably in a dead end when:

- Every unit of growth takes your personal effort and stops the second you stop
- People try it once and never come back
- You've asked for money clearly, more than once, and the answer stays no
- Nobody would email you if it vanished tomorrow

You're probably in the middle when:

- A renewal shows up that you didn't chase
- A stranger replies about something they found on their own
- Usage holds flat instead of decaying
- Someone gets annoyed when a feature breaks

That last one matters more than it looks, because people only get annoyed when losing something would actually cost them.

## Flat Is Not Dying

Flat is the most misread number in indie software.

It feels like death because we've been trained on hockey sticks, and a straight horizontal line looks like there's no business there at all.

What flat actually means is that people keep choosing the thing month after month, with nothing new pushing them.

Decay means the market is leaving, while flat usually means it's waiting to see what you do next.

Your users answer the staying question before you do. If they're sticking around, the project earned more of your time whether or not it earned your excitement.

If they left six months ago, no amount of grit votes them back.

## The Slot Machine In Your Head

Even when the signals are good, staying is hard, and it's worth being honest about why.

Starting something new pays out like a slot machine, with that fresh dopamine hit and strangers congratulating you on the launch.

An established project pays out slowly and predictably, and [predictable rewards are exactly the ones our brains discount](https://joncphillips.com/no-natural-limit/).

That pull toward the next new thing is just a reward loop doing what reward loops do, and treating it as information about your project is how good projects die.

That's also why staying works so well as a filter. Reaching $10k MRR commonly takes one to three years, and nobody's excitement lasts one to three years.

> Everyone's enthusiasm expires around the same point, well before the outcome arrives. The people who get the outcome are the ones who kept working after their excitement left the building.

I'd love to tell you this gets easier. It gets more familiar, which is different. Twenty years in and I still feel the pull every time something I run goes quiet for a month. I've just stopped treating the pull as data.

## Schedule the Quit

The other trap is renegotiating with yourself every day.

Ask "should I quit this" every morning and you'll eventually get a yes, and here's the thing, the yes won't come from your data. The numbers aren't ambiguous at this stage.

It'll come from whichever shiny new idea happens to be sitting in your notes app that Tuesday. At that point the question has turned into an escape hatch.

So I take the decision away from daily me, because daily me can't be trusted around new ideas:

1. Pick a review date, six or twelve months out
2. Write down what the pull signals look like today
3. Close the question until that date
4. Compare notes and decide with evidence

Everything I've shut down got shut down at a review. None of them haunt me. And between reviews there's a strange relief in the question being closed, because the only job left is the work.

## What It Actually Looks Like

It sounds kind of heroic when you say it out loud, but in practice it's pretty dull.

Staying is a publishing schedule kept through months where the open rate says nobody cares. It's answering a support email on a Tuesday afternoon in year six about a feature you built so long ago you have to go read the code to remember how it works.

**Nobody [screenshots that part](https://joncphillips.com/monthly-recurring-profit/).**

I've written weekly for photography audiences since 2012, through stretches where the list grew slower than my patience.

The developer newsletter gets the same treatment now. Most weeks nothing happens, and the whole thing depends on nothing happening for a very long time while the numbers creep in one direction.

Two things made that survivable.

The first was pace. People mix up staying and grinding all the time, but they're pretty much opposites.

**A pace you can hold for five years beats a pace you can hold for five months, every time, in every arena I've competed in.**

Those seventy-hour weeks people brag about are usually a leading indicator of the quit, because nobody holds that through a flat year.

The second was lowering the bar for what counts as a good week.

If progress requires growth, a flat year is fifty-two failures, and nobody survives fifty-two failures. If progress means the work shipped and the system ran, that same flat year is fifty-two wins that happen to look boring.

It's the same year with the same numbers, and one framing keeps you going while the other one ends you.

## Why This Is Worth More Now

Every force in this industry pushes toward more starting and less staying.

AI made beginning trivial. Feeds reward launches and completely ignore year four. The portfolio approach has real advocates, and cheap experiments have their place, right up until one of them works and the correct move becomes [concentration instead of another launch](https://joncphillips.com/the-second-product/).

That means sticking around is getting rarer while the payoff keeps going up.

**When anyone can rebuild your product in a weekend, the product stops being the moat.**

What's left is the four years of published work, the list built one reply at a time, the customers who've renewed enough times that leaving would actually cost them something, and a name people recognize because it [kept showing up](https://joncphillips.com/hey-nobodys-coming/).

Every one of those needed the same thing, which is someone who was still fucking there.

Plenty of people can build something great, but very few stick with anything long enough to find out if they did.

I've watched genuinely talented builders lap themselves for a decade, restarting from zero every six months with more skill and the same result, while less talented people who picked one thing and stuck became impossible to compete with.

The site I almost killed in year three paid for a large chunk of my life. Every good outcome I've had has that same shape. The win showed up years after the point where quitting felt completely reasonable.

The only thing I can take credit for is that when the boring stretch came, I checked which room I was in, saw people still using the thing, and went back to work.

That's the whole skill. It doesn't demo well and it'll never trend, but it's also the one thing no model is going to generate for you.