I have a habit of killing projects after a slow first quarter, and I have a habit of being wrong about which ones deserved killing. Three of the things I almost scrapped in year one went on to be the projects still paying the bills four years later. The launches told me almost nothing about the life of the work. That is the part of the late-2026 box office story I cannot get out of my head. A film that opened to a 31 percent score in 1998 sat at number one twenty-eight years later, and the conversation flipped overnight. The point is not the movie. The point is what the movie’s path tells us about how audiences attach to things, and how badly the launch score misreads the curve.
Most of us are not making movies. We are shipping products, blog posts, side projects, tools, and small businesses. The launch score looks different for each, but the failure mode is the same. A quiet first year gets read as a verdict. If you build anything that takes longer than a coffee break to ship, the slow curve is the only curve that matters, and the launch is the worst measurement of it.
A messy story is not a broken story
Practical Magic was hard to explain in 1998 because it was three pitches at once. Romance. Witch story. Dark comedy about grief. Studios do not love that. Buyers want a single sentence they can repeat at lunch. The original could not give them one, so the marketing machine wrote one anyway, and it was the wrong one.
What changed between 1998 and the sequel’s release is not the story itself. The genre blend is the brand now, not a confusing liability. Audiences learned to want the thing the studio could not describe. The sequel pitches itself the same way and people line up, because the audience that grew up with the original is now the audience that buys tickets on opening night.
This is editorial advice, not a measurement I ran. The data points come from the source material; the framing is mine. If you build something that is hard to describe in one sentence, do not panic. The people who love it will describe it for you, badly and often, and that is enough.
The slow burn is not a marketing plan you can copy
The part that should change how you think about reach is not the sequel at all. It is the twenty-eight years between the two releases. The original did not become a hit through re-releases or a rebrand or a paid campaign. It became a hit because streaming platforms pushed it into a top-10 shelf where a generation of viewers could find it. That is not an acquisition strategy. That is recommendation systems doing the work that 1998 marketing could not.
Cult followings do not pay the bills in year one. They rarely pay the bills in year three either. They pay the bills in year six, when a generation of viewers, readers, or users has been quietly recommending what you built to people who would never have heard of it through your launch channel. The original earned its place slowly, through accumulated goodwill, and that is the version of the curve you should plan for if you are betting on the long game.
If you are the kind of person who likes to write down the rules, here is what I have landed on after watching my own projects rise and fall.
- Stop optimizing for week one. Most of what kills a launch is the gap between launch expectations and launch reality, not the work itself.
- Optimize instead for the audience that finds you in year three. They will not see the launch, and they will not care.
- Treat every quiet win as an asset. A single blog post that someone bookmarks is a recommendation that compounds for years.
- If you cannot describe your work in one sentence, treat that as a problem of marketing, not of work. The product might be fine.
Why the sequel gets to spend what the original earned
The sequel is not a guaranteed hit. The honest version is that it gets to spend twenty-eight years of goodwill earned by word of mouth, recommendation systems, and quiet discovery. The audience that grew up with the original will sit through a long runtime because they trust the world. That trust is the asset. The runtime is the test.
The question I cannot shake is what would have happened if the studio had pulled the original from circulation after the bad score. We would have lost a story that quietly built a twenty-eight-year following and primed an entire sequel release. That is a real cost, and it is invisible if you only watch opening weekends.
What I would say to past me, who was ready to scrap three projects after a slow first quarter
A few specific lessons, addressed to past me.
- The opening weekend is almost always the worst measurement of a thing’s real value. The slow curve is the only curve that matters if you are building anything that lasts.
- If you cannot describe your product in one clean sentence, that is not necessarily a bug. The audience will write the sentence for you, badly and often, and that is enough.
- Streaming and recommendation systems are how most long tails get found. Build for the slow surface, not the splashy surface.
- If the work survives year three, the people who found it through the slow surface will be your loudest evangelists. Treat them like the asset they are.
Trade-offs
The long game is not free in time, and it is not free in patience. Cult followings do not pay rent in year one. Streaming algorithms are out of your control, and you cannot buy your way onto the top-10 shelf. The slow burn does not survive every product, every story, or every team. Most things that try the long game quietly die in year two and nobody notices.
In our case, the practical version of this is: bet on the work, not the launch. If you are making a thing that takes longer than a coffee break to ship, plan for year three, not week one. Your math will be different if you are shipping short-form content or features with a hard shelf life, where the long game is structurally less useful. For a one-time product with no compounding audience, the lesson is not applicable at all.
The honest version is that I cannot tell you which projects will pay off in year six. I can tell you that the ones I almost killed in year one are the ones still paying the bills. The opening-weekend score is a measure of marketing, not of value. Build for the people who will find you later, and you will outlast every launch that did not.
If you only do one thing from this article, stop reading opening-weekend reviews as if they are quality scores. They are launch-channel scores. They tell you how well the studio marketed the thing, not how good the thing is. The same is true for your own launches. Stop reading week one like it is the verdict. It is the marketing report. The verdict comes later, and it usually comes from people who were not in the room at launch.