When should I push, and when should I pull back?
The difficult part is not learning to never quit. It is learning what deserves another attempt.
Written Feb 2025 · Updated Sept 2026
Persistence is easy to admire after it works.
Someone spends ten years building a company and eventually succeeds. The ten years become proof that they were right to continue.
But many people spend ten years on something that does not work.
From the middle, the two stories can look exactly the same.
So when should I push forward, and when should I pull back?
I do not think the answer is “never give up”. I also do not think every hard week is evidence that the plan is wrong.
The useful question is whether the reason for continuing is still true.
Difficulty is not a decision#
Some work is difficult because it matters and I have not learned how to do it yet.
Some work is difficult because the basic idea is wrong.
The feeling can be the same: slow progress, doubt, tiredness and the temptation to do something easier. If I use discomfort as the signal, I will leave good work too early. If I use effort already spent as the signal, I may stay far too long.
This is why the sunk-cost effect matters. Once we have invested time, money or identity, stopping feels like wasting the investment. But the cost has already happened. It cannot make the next month more likely to work.
The next month needs its own case.
Ask what has changed#
I would start with the original belief.
What did I believe when I began?
Maybe I thought customers had a painful problem. Maybe I thought I wanted a particular career. Maybe I thought a relationship could be repaired. Write the belief plainly enough that it can be challenged.
Then ask:
- What evidence have I learned since?
- Does it weaken the belief, or only show that my first method was poor?
- Is there a smaller and cheaper test I can run next?
- If I were not already invested, would I choose this again today?
That last question is harsh. It is also clean.
Separate the goal from the route#
Pulling back from a route is not always giving up on the goal.
If the goal is to help people learn, a particular course can fail while the goal remains true. If the goal is to build financial independence, one company does not have to carry the whole purpose. If the goal is to become a filmmaker, one rejected film is not the final test.
Research on goal adjustment makes a similar distinction. Disengaging from an unattainable goal can be healthy, particularly when it is followed by re-engagement with another meaningful goal.
The important word is unattainable. Not uncomfortable. Not late. Not embarrassing.
Unattainable means the path no longer has a credible mechanism.
The goal and the current route are two different decisions.
Decide before the bad day#
It is difficult to make a clean decision while ashamed, exhausted or excited.
So decide some of the rules in advance.
Before starting, write down:
- what evidence would make you continue;
- what evidence would make you change the method;
- how much time or money the next test is allowed to consume;
- what would make you stop completely;
- when you will review the decision.
This does not remove judgement. It stops each mood from inventing a new rule.
A deadline can also force honesty. “I will review this in six weeks” is better than asking whether to quit every evening.
Push when learning is still happening#
I would push when the original reason is intact, the next experiment is clear and the work is still producing information.
Progress is not only revenue or applause. It can be a sharper customer, a failed test that removes an assumption, a skill that is becoming usable, or a version that is plainly better than the last one.
But “I am learning” can become a comfortable excuse too. The learning should change what I do.
Pull back when only the investment remains#
I would pull back when I can no longer explain the mechanism by which the work will succeed, when new evidence repeatedly weakens the original belief, or when I am continuing mainly because stopping would damage the story I tell about myself.
That does not make the earlier effort worthless.
It means the effort bought an answer.
Sometimes persistence means staying with the work. Sometimes it means staying with the larger purpose while being willing to abandon the route.
The hard part is not learning to never quit.
It is learning what deserves another attempt.
Research sources
- Hal R. Arkes and Catherine Blumer, “The Psychology of Sunk Cost”, Organizational Behavior and Human Decision Processes (1985)
- Carsten Wrosch et al., “Adaptive Self-Regulation of Unattainable Goals”, Personality and Social Psychology Bulletin (2003)
- Carsten Wrosch et al., “Goal Adjustment Capacities, Subjective Well-Being, and Physical Health” (2013)
