1. The shop was empty. I joined. Then humanity spawned behind me
Nobody is waiting.
You think, “Great, no line,” and step up.
A few minutes later you look back.
People. More people. Somehow, a queue.
Then it happens again somewhere else. And again.
After enough repetitions, a reasonable question becomes an unreasonable-sounding one:
“Am I a customer, or a queue summoner?”
This is nowhere near proof of a personal superpower. But the underlying phenomenon—visible customers attracting later customers—fits surprisingly well with research in consumer behavior, observational learning and queue choice.
So “it is all in your head” is too simple an answer.
2. The short answer: yes, queues can attract more queues
| Hypothesis | Evidence status | Verdict |
|---|---|---|
| A visible queue can attract later customers | Field and theoretical research | Well supported |
| Unfamiliar customers rely more on other people’s choices | Field evidence | Strong support |
| Even one extra person can sometimes make the longer queue rational to choose | Formal model | Possible under conditions |
| Waiting can signal popularity or quality | Multiple experiments | Supported |
| An attractive first customer strengthens the effect | Indirect evidence | Plausible, not directly proven |
| “I personally create queues unusually often” | No personal dataset | Unproven |
“Social proof” is a useful everyday label. Academic work often describes more specific mechanisms such as observational learning and information externalities.
A newcomer may not see a line as merely an obstacle. They may read it as free information:
“Those people might know something I don’t.”
3. Why zero-to-one is bigger than “one additional customer”
An empty shop contains uncertainty.
Is it open? Is this the right entrance? Is the product bad? Where do I wait? Is the shop next door the safer choice?
One visible customer quietly answers several of those questions at once.
The place is open. This is where ordering happens. Someone has chosen it. There is now a visible behavioral script.
One person has accidentally become a signboard, a review and a wayfinding marker.
Veeraraghavan and Debo’s queue-choice model shows that when service rates are equal and known, it can be rational to choose the longer queue even when it contains only one additional customer. Their model also shows how the private signal of the customer who arrives when both providers are idle can help determine which provider subsequently thrives.[1]
That is an unusually elegant academic way of saying:
the first customer can matter a lot.
From the shop’s perspective, you are no longer a person buying lunch.
You are seed liquidity.
4. Field evidence: when people do not know the shop, they watch other people
Raz and Ert observed customers choosing between nearby similar restaurants. Queue length was positively associated with newcomers joining, and the relationship was substantially stronger in tourist settings where customers were less familiar with the restaurants.[2]
Fishman, Fishman and Gneezy studied two adjacent, outwardly similar food stands on a university campus. Early in the academic year, when many students had little prior experience, arriving customers tended to choose the more crowded stand. Later in the year, after students had acquired direct experience, that tendency disappeared.[3]
That is more interesting than blind imitation.
When people lack private knowledge, they borrow information from other people’s choices.
Giebelhausen and colleagues found across four experiments that waiting can function as a quality signal and, under some conditions, increase purchase intentions and satisfaction rather than simply damage them.[4]
Munichor and Cooke likewise found in a field study and experiments that longer waits could lead to more purchasing or consumption when the wait could reasonably be interpreted as reflecting other people’s preferences.[5]
A queue is therefore both friction and advertising.
A human billboard that requires no electricity.
5. What if the first customer is attractive? Interesting—but this is where the evidence gets indirect
There is an important boundary here.
We did not find strong direct evidence showing that an attractive person at the front of a queue causes that queue to grow faster.
Adjacent evidence does exist.
A large meta-analytic review by Langlois and colleagues found that people judged as attractive tend, on average, to be evaluated and treated more positively.[6]
Argo, Dahl and Morales ran three field experiments in real retail settings and found that products believed to have been touched by a highly attractive other received higher evaluations under specific conditions. They called this a form of positive consumer contagion. The effect had important boundary conditions, including the sex relation between the attractive source and the observer, so it is not a magical universal effect.[7]
That makes a queue-specific hypothesis reasonable: a first customer who looks stylish, clean, confident or otherwise like a useful social reference may provide a stronger signal than a completely uninformative passer-by.
But that step is an inference, not a demonstrated queue effect.
So the scientifically defensible version is:
Attractiveness boost: plausible. Direct queue evidence: not established.
Shop: “Can we use attractive early customers as free advertising?”
Research: “That is not what we said.”
Shop: “Damn.”
6. The strongest rival explanation: you did not summon the crowd; you arrived 30 seconds before the same demand wave
Customer arrivals are not perfectly independent.
Lunch breaks, dinner time, train arrivals, movie endings, traffic lights, event intermissions, rain and shared schedules can send people toward the same place in clusters.
You decide “time to eat.” Ten strangers nearby may have made the same decision from the same clock.
Then this sequence appears:
0 customers → you join → eight people arrive within two minutes.
It feels like you caused eight people to appear.
Another explanation is simply:
you reached the front edge of a demand wave 30 seconds earlier than everyone else.
There is also selection. You may be unusually quick at spotting the same attractive shop, smell, menu, price or location that will attract everyone else moments later.
And long queues are not always positive. If waiting costs are high, people leave. A line can also signal “slow service” rather than “great product.” Queue models explicitly show that the information contained in a long line depends on what customers believe about service speed and quality.[1]
A queue is advertising only until the advertisement starts physically blocking the pavement.
7. “But this has happened many times.” Does repetition count as evidence?
Yes—repetition is more interesting than a single coincidence. But what matters is the denominator, not only the memorable hits.
You need to count:
- times you joined and many people followed;
- times you joined and nobody followed;
- times you did not join yet the crowd appeared anyway.
Tversky and Kahneman’s work on the availability heuristic showed that people can estimate frequency and probability from how easily examples come to mind.[8]
“LOL, another queue formed behind me” is memorable.
“I bought food and left; nobody came” is deleted from the brain’s camera roll before dinner.
A simple N=1 field test would be much better:
- Record only places with roughly zero or one existing customer.
- Randomly alternate between joining immediately and watching from a short distance for two or three minutes.
- Record time, day, surrounding foot traffic, weather and obvious event-release effects.
- Count new customers after five and ten minutes.
- Compare arrivals per minute between the “I joined” and “I did not join yet” conditions.
Twenty or thirty observations would not prove a universal law, but it would be dramatically better than memory alone.
If the gap survives, the shop may owe you a new line item:
Seed-liquidity provision fee.
8. Final verdict: customer, or unpaid marketing infrastructure?
The best current judgment is:
Queues really can attract subsequent customers.
There are good theoretical reasons why the first visible customer can matter.
Repeated personal experiences are worth testing rather than automatically dismissing.
But:
a special individual “queue-summoning ability” has not been demonstrated.
And while attractiveness can influence social and consumer judgments, the specific claim “an attractive first customer makes the line grow” remains a plausible hypothesis, not an established result.
So the provisional job title is:
Unpaid Seed Liquidity Provider.
Shop: “Slow day today.”
You: “One of these, please.”
Three minutes later: “The end of the line is over here.”
You: “I literally came here to eat.”
Superpower or not, one thing is real: people watch who has already chosen far more than we like to admit.
References (8)
- Veeraraghavan, S. & Debo, L. (2009). Joining Longer Queues: Information Externalities in Queue Choice. Manufacturing & Service Operations Management, 11(4), 543–562 doi.org
- Raz, O. & Ert, E. (2008). “Size Counts”: The Effect of Queue Length on Choice between Similar Restaurants. Advances in Consumer Research, 35, 803–804 acrwebsite.org
- Fishman, A., Fishman, R. & Gneezy, U. (2019). A tale of two food stands: Observational learning in the field. Journal of Economic Behavior & Organization, 159, 101–108 doi.org
- Giebelhausen, M. D., Robinson, S. G. & Cronin, J. J. (2011). Worth waiting for: increasing satisfaction by making consumers wait. Journal of the Academy of Marketing Science, 39(6), 889–905 doi.org
- Munichor, N. & Cooke, A. D. J. (2022). Hate the wait? How social inferences can cause customers who wait longer to buy more. Frontiers in Psychology, 13, 990671 doi.org
- Langlois, J. H. et al. (2000). Maxims or myths of beauty? A meta-analytic and theoretical review. Psychological Bulletin, 126(3), 390–423 doi.org
- Argo, J. J., Dahl, D. W. & Morales, A. C. (2008). Positive Consumer Contagion: Responses to Attractive Others in a Retail Context. Journal of Marketing Research, 45(6), 690–701 doi.org
- Tversky, A. & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5(2), 207–232. )90033-9 doi.org
