Reference· Weeks 1–2 protocol· Print me

The 29 Missions

Not my invention. This is the actual task list from a published, preregistered intervention that measurably reduced fear of rejection in 286 people — reproduced so you can run it rather than approximate it.

Most of what gets written about approaching strangers is somebody's opinion. This isn't. Sandstrom, Boothby and Cooney ran a five-day, app-delivered scavenger hunt in which participants found and talked to strangers matching a stated characteristic. It produced 1,336 conversations and moved the beliefs it targeted, with effects still present a week later. Sandstrom, Boothby & Cooney 2022, JESP 102, 104356

Read these numbers before you start

Day 1 of the study

Only 40% of participants believed the first person they approached would talk to them.
92% did.

Across the whole week, of 1,336 conversations, 1,164 — 87% — happened with the very first person approached. Not the third. Not after a few knock-backs. The first.

And here is the part that should reframe the entire exercise for you. Predicted rejections fell steadily over the week (p<.001). Actual rejections did not change at all (p=.37) — they were near-zero from day one and stayed there.

The intervention did not teach anyone to be rejected less. It calibrated their beliefs to a reality that was already favourable. Nothing about the world changed. Only the estimate did — and the estimate was the thing that had been stopping them.

What moved

MeasureStartEnd +1 weekEffect
Expected rejections
"How many people will I need to approach before someone talks to me?"
0.890.230.32 ratio 3.82, p<.001 — held at 2.81 a week later
Conversational ability4.344.77 4.81d=0.37 — and rose further after the study ended
Awkwardness3.783.34 3.45d=0.33
Enjoyment4.344.70 4.58d=0.34
Positive impression made4.254.62 4.66d=0.38

Controls didn't move on rejection expectations (0.99 → 1.11), conversational ability (d=0.07), or awkwardness.

One design detail that matters for you

The authors state the belief change happened gradually, day by day, and explicitly conclude that a shorter version would not work. Repeated experience is the active ingredient, not insight. You cannot read this page and get the effect. That's the whole reason the curriculum runs six weeks rather than a weekend.

The 29 missions

Verbatim from the study's materials. Each one reads: find someone who [X] and talk to them for a few minutes. Tick them off — one per day minimum, and the order is yours.

MissionFind someone…
Al Frescooutdoors
All Smileswho seems friendly
Artsywho looks artistic
Blue Moodwho looks sad
Bossy Pantswho looks like a leader
Caffeination Stationa barista or server
Coffee Breakwho's drinking a coffee
Do Gooderwho seems like a nice or kind person
Fashionistawho's accessorising — a scarf, a hat
Fun Fabricwearing a stand-out print — stripes, animal print
Graphic Teewearing an interesting shirt
Hotwhom you find attractive
Hungrywho's eating
Inked Upwho has a tattoo
Insideindoors
Jocksporty
Kickin' Itwearing interesting shoes
Line Upwho's waiting in a queue
Manscapewho has a beard or goatee
Minionwearing a uniform
Nailed Itwho has funky nails
Next Genfrom a different generation than you
On Topwearing a hat
Ray of Sunshinewho looks happy
Sexywhose gender differs from yours
Skin Deepwhose skin tone is different from yours
Twinswearing the same thing as you
Unicornwith eye-catching, dyed, or unusual hair
Wild Card ×2anyone of your choosing

All 29. Tap as you complete them.

Why the mission structure does work a bare instruction wouldn't

It supplies the pretext. This is the thing you said you lack — no natural reason to speak. A mission converts "I should talk to someone" into "I need to find someone wearing a hat," which is a task, and tasks are the format your work conversations already run on. You're borrowing the scaffold back.

It removes target selection. You don't choose who to approach — the mission does. That strips out the deliberation window where your veto lives.

Two missions worth noticing

"Hot — someone you find attractive" and "Sexy — someone whose gender differs from yours" are both in the published list. Your hardest case is a sanctioned, ordinary part of a validated research protocol. Save them for days 3–5 rather than day 1, but don't quietly skip them — that's the category where your prediction error is largest, and therefore where the reps are worth most.

How to run it

  1. Five days, Monday to Friday. Minimum one mission per day. Do more if it's going well.
  2. Each morning, before anything: write your prediction for the day — "how many people will I need to approach before one talks to me?" That's the study's actual outcome measure and it's the number you're moving.
  3. After each conversation: log what happened in one line. Ten seconds.
  4. Friday: compare the week's predictions against the week's outcomes. That comparison is the intervention. Everything else is delivery.

Honest limitations

Assignment was at cohort level rather than individual; 30% of starters were excluded; the sample was students averaging 20 years old, mostly women (209 of 286). Controls also improved on enjoyment and positive impression, and several treatment-vs-control contrasts were non-significant despite significant within-group change. The rejection and conversational-ability findings are the robust ones. A 44-year-old man in Kuala Lumpur is not the studied population — but the mechanism it identifies (belief miscalibrated against a favourable reality) is the same mechanism the commuter, liking-gap and depth-barrier studies all found in adults.

The Number Weeks 1–2

40% expected yes.
92% got one.

87% of 1,336 conversations happened with the first person approached.
Predicted rejections fell all week. Actual rejections never moved — they were near zero from day one.

You are not learning to be rejected less. You are correcting an estimate that was already wrong.

Primary source

Sandstrom, G. M., Boothby, E. J., & Cooney, G. (2022). "Talking to strangers: A week-long intervention reduces psychological barriers to social connection." Journal of Experimental Social Psychology, 102, 104356. Free PDF · Preregistration on OSF · Original mission list

Read the Results section, particularly the predicted-versus-actual rejection comparison. Sandstrom's book Once Upon a Stranger (HarperOne, March 2026) is the long-form version by the same researcher.