All shows / Come Back Soon

Come Back Soon cover art

Come Back Soon

The Ending Has an Expiry Date

Doctors once made a colonoscopy longer on purpose so patients would come back; here's what that says about the mirror moment, why its pull fades before the next appointment, and the one thing to do before your client stands up.

Priya and Hector · 22 min · 13 Oct 2026

FactUnion VerifiedEvery fact on the episode's fact sheet was checked against its sources. Wrong facts were corrected. What this meansFacts checked 13 Oct 2026

Chapters (9)

About this episode

In a Toronto trial, doctors left the scope in for about three extra minutes at the end of a colonoscopy, and more of those patients came back for the next one. Behind that trial is the peak-end rule: we remember an experience mostly by its high point and how it ended. Priya, a behavioural scientist who once did brows at a mall kiosk, and Hector, a barber with eighteen years behind the chair, look at how strong that evidence really is, where it breaks down, and why a haircut is more like a holiday than an ice bath.

Then comes the twist. Research suggests ending-driven memory starts to fade somewhere between three and seven weeks, which is about one haircut, colour or lash-fill cycle. So your mirror moment probably won't bring a client back six weeks from now. Its real job is the booking you ask for before they stand up. We say plainly what is proven, what is promising, and what no one has tested in a salon yet.

Hosted by Priya and Hector, performed by synthetic voices. Supported by Lynqly.

Facts and sources

Every fact used in this episode, with the sources it was checked against. Download as data

174 of 174 recorded lines matched the checked script.

  1. Redelmeier, Katz and Kahneman’s colonoscopy trial, published in Pain in 2003, randomized 682 patients.

  2. In that trial, 53% of patients in the extended-ending group returned for a repeat colonoscopy, compared with 48% of the control group.

  3. After controlling for prior colonoscopy, procedure indications and abnormal findings, the extended-ending group had 41% higher odds of returning (P = 0.038), while the unadjusted 18% increase was not statistically significant (P > 0.20).

  4. The extended ending in the trial meant leaving the colonoscope in place for about three additional minutes.

  5. A 2022 meta-analysis in Organizational Behavior and Human Decision Processes (vol. 170) pooled 174 effect sizes and found a large peak-end effect of r = 0.581 (95% CI 0.487 to 0.661).

Show all 22 factsShow fewer facts

  1. The 2022 meta-analysis found the peak-end effect stronger than the effects of the beginning, the low point and the duration of an experience, with duration showing essentially no effect.

  2. A 2019 Frontiers in Psychology paper reports that for complex, heterogeneous everyday experiences, average feeling often predicts retrospective evaluations better than the peak-end rule does.

  3. A 2016 paper is titled ‘Questioning the end effect: Endings are not inherently over-weighted in retrospective evaluations of experiences.’

  4. Talya Miron-Shatz of Princeton University describes evaluating multi-episode events as a boundary condition for the peak-end rule.

    Reworded to match its sources. The paper has one author, Talya Miron-Shatz, and it lists her at Princeton's Woodrow Wilson School, not as a Wharton researcher; Wharton only hosts the PDF.

  5. A test of the peak-end rule using extended autobiographical events such as holidays found limited predictive power for the rule.

  6. A 2024 Journal of Business Research study (vol. 185) by McCullough and colleagues analysed more than 5,000 customer stays at a hotel over a five-year period.

  7. That 2024 hotel study found first impressions were the most significant contributor to overall satisfaction and to evaluations of frontline staff, and became more important as the time between the stay and the evaluation grew.

  8. Neal, Wood and Quinn (2006) describe habits as repeated behaviours triggered automatically by cues in the context where they usually happen, rather than by deliberate intentions.

  9. In Martin et al.’s 2012 NHS study, having patients write down their own appointment details reduced missed appointments by 18% compared with the previous six-month average.

  10. The Martin et al. study was not a randomized trial: it ran from February to May 2011 at two Bedfordshire practices (Wheatfield Surgery and Toddington Medical Centre) and reported percentage changes rather than raw no-show rates.

  11. In the same study, asking patients to repeat their appointment back to staff cut missed appointments by 3.5%, and combining these steps with posters highlighting how many patients attended produced a 31.7% reduction.

  12. A 2023 pragmatic trial in US Veterans Affairs primary care and mental health clinics found that behavioural nudges had no significant effect on missed appointment rates.

  13. Research on implementation intentions (Gollwitzer) proposes that linking a specific future situation to a planned action makes follow-through more automatic.

  14. A Cochrane review of eight randomized trials with 6,615 participants found text reminders increased attendance from 67.8% to 78.6%.

    Reworded to match its sources. Source corrected: the page first cited does not state this; the sources listed do.

  15. 2026 industry reporting says about 71% of solo beauty professionals operate as booth renters or independent contractors, up from 58% in 2019.

  16. A 2026 Lutily blog post, labeled as an illustrative story with a fictional narrator, describes a four-chair salon owner whose rebooking rate rose from 47% to 71% over about seven months after the rebooking conversation was moved into the final five minutes of the service.

    Reworded to match its sources. The post says its narrator and experiences are fictional, so it is not a real owner's self-report, and nothing in the evidence confirms Portland.

  17. There is no peer-reviewed controlled evidence that booking the next appointment before a salon client leaves increases return rates; the benchmarks commonly cited come from vendors and consultants.

Removed during checking (1)

  • Geng and colleagues (2013, Journal of Behavioral Decision Making) identified roughly 3 to 7 weeks as the watershed between short retention intervals, when people rely on peak and end feelings, and long intervals, when they rely on both specific memories and general knowledge.

    Why it was removed: The page gives Geng et al. 2013 in JBDM, says people used peak and end affects only over a short interval, and puts the watershed between 3 and 7 weeks; the check adds that people use episodic and semantic information over long intervals.

If you like Come Back Soon, you might also like