The Project Is Half Done. Continue or Stop?
You're months into something. The money is out, the thing is nearly finished, and the uncomfortable question finally surfaces: does going on actually pay off? And underneath it, the quieter one, how should you even think about all the effort already spent?
The way to get out of it is to start smaller. Where would the next month's time and money do the most work? Once the future is laid out on the table, the past stops feeling like a verdict and starts looking like background.
Look at what happens next, not what you've already paid
Say a photographer has spent two months and 30,000 yuan building a course and is trying to decide what comes after. The numbers below are set up for illustration, both options take roughly the same working hours, and everything gets settled within the coming month:
| Next step | New income | New costs | Left over |
|---|---|---|---|
| Keep promoting and delivering the course | 18,000 yuan | 12,000 yuan | 6,000 yuan |
| Pursue photography jobs available for discussion | 25,000 yuan | 16,000 yuan | 9,000 yuan |
If this month's cash is what matters, the jobs look better. But if the course keeps selling past this month, don't stop the comparison there. Forecast the follow-on sales, the customer questions, the updates, the promotion, and write down your reasons for each estimate. Compare both paths using their expected future income, costs, and workload.
Meanwhile, that 30,000 yuan is already gone under both options. It records what it cost to reach this point; it can't be recovered either way. What you can still decide is what to invest next, and how much. That is exactly the job of separating sunk and opportunity costs.
Some of the past investment isn't dead weight. The course can still sell. The equipment can still shoot, or be sold. The feedback from early buyers can still shape the product. Put each of those into the future plan and estimate what it can actually do. And if you have already taken client money and promised a delivery, count the work that promise requires. It's a real cost, not a sunk one.
Stopping doesn't mean burning what you learned
The team behind Slack spent years before Slack building an online game called Glitch. By their own account, it never attracted enough players to cover the cost of developing and running it, and staying the course would have drained their remaining funds within a few months.
In 2012 they shut the game down and turned to something new. What carried over was a set of collaboration needs they had felt firsthand: the team needed one shared conversation, a way to find past messages quickly, and a way for their other tools to plug in. Those needs became the starting point for Slack.
Notice what happened. They abandoned one business and reused what it had built, two separate judgments. The market conditions for the old product needed re-examining. The experience, code, content, and customer understanding it left behind got to shop around for new uses.
When you're unsure, buy one answer
Sometimes committing another six months is too hasty, and quitting is too hasty too. A focused trial can help resolve a question that matters to the decision.
For the photographer, the question might be: do clients want the finished course, or would they rather pay someone to critique their work? Writing more chapters does little to answer that. Showing a target customer two clear offers with prices does.
Before starting the trial, write down a short agreement with yourself: what exactly you're checking, how you'll watch for the answer, the ceiling on money and hours, the date you'll look, and what you'll do depending on what you find.
The photographer could offer the course alongside a limited number of paid critiques, recording for each option where each client came from, what they actually paid, how long each delivery took, whether anyone wanted to come back. After that round there's material to compare against the course: buying reasons, delivery hours, and the money left after costs. If demand clearly concentrates in one direction, reshape the product. If customers look at a concrete offer and still don't adopt it, go back and check who you're asking, your pricing, and what they currently buy instead.
One main question per round. Otherwise you change the audience, the product, the price, and the channel all at once, and when the results come in you can't tell which change caused anything. For how to record this kind of test, see willingness to pay and demand checks.
Write down which decision you actually made
"Continue" means you have a reason to allocate the next round of resources. Sales keep closing at the current price, and the delivery cost and workload sit where you expect them.
"Adjust" means the feedback points at something more specific, a better client type, a lighter delivery method, a clearer package, and you commit to a next check for it.
"Stop" means the next investment offers little, and the same time and money have a better place. Finish what you've promised, keep what's still useful, and move the rest on.
Each of these can be written down with a deadline: the next week, the next batch of orders, the next invoice. Then, at the end of the round, ask the question again: what do we know now that we didn't before, and which option got stronger because of it? Keep pushing inside the cash you actually have and the payment dates you actually have, available funds and when they arrive, and every round earns its place in the one that follows.