How to know when to pivot, tighten, or push through
You’ve been working on something for a few weeks. The progress doesn’t match the effort. Something is off.
Most people frame this as a binary: either I push through, or I quit and try something else. That framing is wrong, and it’s where most of the bad calls come from. There’s a third option that’s almost always the right one.
The three options are: pivot, tighten, or push through.
Pivot
A pivot is changing what you’re doing because the underlying assumption turned out to be wrong. You thought the bottleneck was your fitness. It was actually your sleep. The first version of the plan can’t fix sleep, so the plan changes.
Pivots feel dramatic but they’re usually quiet: a one-line change to which problem you’re working on. The work doesn’t restart. Most of the previous work still applies — the priority of which work matters most just shifts.
You should pivot when:
- The data from the last two or three weeks consistently points at a problem the current plan doesn’t address.
- The thing you keep failing at isn’t the thing you set out to fix — it’s a precursor you didn’t name.
- You’ve been telling yourself “next week I’ll get to that” about the same precursor for a month.
Pivots are easy to do too late and almost never too early.
Tighten
A tighten keeps the what the same and changes the how. Same goal, better mechanism. The plan was right; the execution was sloppy.
Most stalls call for a tighten. Examples:
- Goal: read for thirty minutes a day. Stall: phone wins. Tighten: phone outside the bedroom from 9pm.
- Goal: lift twice a week. Stall: gym is too far. Tighten: bands at home, same volume, less drive time.
- Goal: weekly date night. Stall: kids’ bedtime keeps eating it. Tighten: date night becomes 9–10pm Wednesday after the house is asleep.
A tighten changes the friction without changing the destination. Most people skip tightening because it doesn’t feel decisive enough — but it’s where most of the actual progress comes from.
You should tighten when:
- The goal still feels right when you read it back to yourself, and you can name the specific friction that’s getting in the way.
- The friction is environmental (time, place, equipment, social pressure) rather than a values mismatch.
- You haven’t actually tested the goal at full effort yet — you’ve been testing the goal with a built-in handicap.
Push through
Push-through is the rarest of the three, even though it’s the most common advice. You push through when the what is right, the how is right, and the only thing missing is time.
Push-through doesn’t mean grinding. It means recognizing that some foundation work doesn’t show visible results for six to twelve weeks, and the absence of visible results in week two is not evidence that something is wrong.
You should push through when:
- The work being done is genuinely the right work, with a tight enough feedback loop that you’d see signal if it existed.
- You’ve already pivoted or tightened recently and the new version hasn’t had time to mature.
- You’re four to eight weeks into a thing that you said up front would take three months.
Push-through is only the right answer once or twice in a typical year. If it’s your default, you’re probably skipping pivots or tightens that should be happening.
A simple rubric
Each week, when something feels stalled, ask in this order:
- Is the goal still pointing at the right problem? If no → pivot.
- Is the mechanism set up to actually deliver against the goal? If no → tighten.
- Has the current version had enough time? If no → push through. If the answer is yes and nothing’s improving, you’re back at #1.
The mistake is to skip steps 1 and 2 and default to step 3. That’s how people end up six months into “just keep going” before realizing the direction was wrong all along.
What this looks like in Vision Guide
The weekly check-in in Vision Guide explicitly asks, for each commitment: continue, tighten, pivot, pause? It’s the same rubric, in the same order. The system also tracks the answer over time — three “tighten” answers in a row is a different signal than three “continue” answers in a row, and the next plan revision uses that history.
If you’d like a tool that walks you through it, the free profile and weekly check-in are here. Otherwise the rubric above is the actual content. Three questions, in order. Don’t skip to the third one.