Lesson 3.5 · StrategyGuide · 10 min readFree · No signup

Sunk Cost Effect: past investment makes it hard to walk away

Part of the Psychology of Design learning path. The cognitive biases and psychology principles behind every click, scroll, and conversion.

L3 · How people act over time · Lesson 5 of 2610 min read for this one

What you'll understand by the end of this lesson

  • Why people continue with something they've started even when stopping is the better choice
  • How past investment in a flow increases the likelihood of completing it
  • Why "you're almost there" messaging works when users are near the end
  • How to design onboarding flows that use progress visibility to reduce drop-off

The principle in plain English

A sunk cost is something you've already spent — time, money, effort — that you cannot get back.

Rational decision theory says sunk costs should be ignored when deciding what to do next. The money is gone. The time has passed. The only thing that matters is whether continuing forward is worth it from this moment.

But that's not how humans behave.

People factor sunk costs into future decisions all the time. They stay in a bad film because they've already paid for the ticket. They continue a failing project because so much has been invested. They complete a form they no longer need because they've already filled in 7 of 9 fields.

The Sunk Cost Effect is the tendency to continue an investment not because future value justifies it, but because stopping would feel like wasting what's already been spent.


A simple example

You're applying for a job. The application has 12 steps. You're on step 10. You've just realised the role has a commute you're not comfortable with.

The rational choice is to stop and move on. But most people finish the application — because stopping now means all the effort from steps 1 to 9 was "for nothing." The sunk cost of those 9 steps isn't relevant to whether you want the job. But it feels relevant, so you continue.


How it shows up in onboarding flows

Users who have completed 80% of an onboarding flow are dramatically more likely to finish than users at 20% — even if the remaining steps are technically harder.

This isn't because users at 80% are more motivated or more interested. It's because they've invested more. Walking away from an 80%-complete flow means accepting that the time and effort spent on the first 80% accomplished nothing. That feels wasteful, so they continue.

This is why progress indicators matter in onboarding. Showing users how far they've come is not just a UX nicety — it's a retention mechanism. Every step completed is sunk cost that makes the next step more likely to happen.

If you show progress in an onboarding flow, always show the count in a way that emphasises what's been done, not what remains. "Step 4 of 5" does more work than "1 step left" — because it surfaces the four completed steps as visible investment. The sunk cost is right there on screen.


"You're almost there" messaging

The most explicit application of Sunk Cost Effect in conversion copy is the "you're almost there" message.

When a user reaches step 4 of 5, step 8 of 10, or 80% complete on a progress bar, a message that explicitly acknowledges their investment — "you've done the hard work, there's only one step left" — converts significantly better than a neutral status update.

The message does two things simultaneously:

  1. It acknowledges the sunk cost (you've invested a lot)
  2. It reframes stopping as wasteful (abandoning now would forfeit everything)

Long-form application flows — loan applications, insurance quotes, detailed account setups — use this heavily because the drop-off cost of an abandoned application is high and the margin of completion at late stages is worth optimising.

Sunk Cost Effect is descriptive of a real cognitive bias — not a mandate to exploit it. There is an ethical version: reminding users of genuine progress toward something they actually want. And a manipulative version: engineering unnecessarily long flows specifically so users feel too invested to stop. The test is whether the thing they are completing genuinely serves their interest. If it does, surfacing the sunk cost is helpful framing. If it does not, you are using bias to coerce completion.


The CRO audit

Look at your key flows — signup, onboarding, checkout, application — and ask:

1. Is your progress indicator showing investment or remaining effort?

"2 steps left" highlights what remains. "Step 3 of 5" highlights what's been done. The second framing activates sunk cost. If your progress indicator is not visible, or not showing completed steps, you are leaving a retention lever unused.

2. Do you have "almost there" copy at the 70–80% mark?

Check whether your flow explicitly acknowledges investment near the end. A simple line — "you're nearly done, just one more step" — at the penultimate step can reduce last-mile drop-off noticeably. Most flows skip this entirely.

3. Are your flows as short as they genuinely need to be?

Sunk cost should work on flows that users want to complete. If the flow is longer than it needs to be, sunk cost is propping up a friction problem rather than supporting a good experience. Audit field counts and step counts to confirm every step earns its place before relying on sunk cost to carry users through.



Q1

A checkout flow has a 40% abandonment rate at step 4 of 5. A CRO specialist recommends adding copy at the start of step 4 that says: 'Nearly there — you've done the hard part, just your payment details to go.' What principle is this applying, and why is it likely to reduce abandonment?

Think about this

You've seen how past investment makes stopping feel costly. Now — what about the decision-making quality at the end of a long flow? When users have made many decisions already, what happens to the quality of the final one?