---
title: "Projection Content Length and Loop Structure"
url: https://americanguerrillamarketing.com/articles/projection-content-length-and-loop/
date: 2026-11-22
modified: 2026-11-22
lang: en
author: "Millie Phillips"
description: "How to set projection loop length from the street: nobody watches from the start, passers by versus dwellers, anchor moments and night scheduling."
categories:
  - "Guerrilla Projection Advertising"
image: https://americanguerrillamarketing.com/wp-content/uploads/2026/08/Guerrilla-Projections-7-2.jpg
word_count: 3112
---

# Projection Content Length and Loop Structure

Almost every projection brief I receive asks the wrong question about duration.

The question that arrives is "how long should the content be". The question that matters is "how long is a viewer going to look at it", and those two numbers are not related in the way people assume.

A brand team sits in a room and watches the content on a monitor, start to finish, with full attention. Ninety seconds feels brisk. Then the same piece goes on a wall and the actual audience is a person crossing at the lights who looks up for four seconds, or someone sitting outside a bar who glances at it eleven separate times across an hour, or a group walking past who see it from behind for twenty seconds while talking to each other.

None of those people watch it the way the room did. Build for the room and the piece underperforms. Build for the street and a shorter, simpler loop will outperform a longer, cleverer one.

This is how our crews actually think about loop length, structure and playback across a projection night.

## Nobody watches from the beginning

This is the foundational fact and it changes every decision downstream.

A projection has no start time from the viewer's perspective. People arrive at a random point in the loop. There is no title sequence, no establishing shot, no building of anticipation, because the first thing they see is whatever happened to be on the wall when they looked up.

That means narrative structure is largely wasted. If understanding frame sixty requires having seen frame ten, most of your audience will not understand frame sixty.

It also means every segment of the loop has to stand alone. At any arbitrary moment, somebody is looking at the wall for the first time, and what is on it right then is the entire message as far as they are concerned.

The design implication is that a loop is better thought of as a sequence of complete statements than as a story. Each statement makes sense cold. Transitions between them are pleasant but not load bearing.

When I explain this to creative teams, the useful test is to pause the content at twenty random points and ask whether each frozen frame communicates something on its own. If most of them are transitional mush, the piece is built wrong for the medium.

## Two kinds of viewer, two kinds of dwell

Projection audiences split into two groups and the loop has to serve both.

Passers through are the larger group. They walk past, they glance up, they look for a few seconds and they are gone. They will see the wall once, from a changing angle, often while talking to somebody. What they need is a single clear message delivered in whatever moment they happen to catch.

Dwellers are the smaller but more valuable group. People at a bus stop, outside a bar, waiting for someone, sitting on a terrace, queuing for a venue. They will see the content repeatedly, potentially many times across an evening. What they need is something that does not become irritating on the tenth viewing, and ideally something with enough variation that later viewings add a little.

A loop that only serves passers through can be relentlessly repetitive and will annoy the dwellers. A loop that only serves dwellers tends to be long and layered, and the passers through get a fragment of nothing.

The structure that serves both is a short loop with a clear anchor moment, plus some variation around it. The anchor is the same every cycle so the glancing viewer always gets the message within a short window. The variation keeps it tolerable for the person who sees it twenty times.

Knowing which group dominates a given location is a survey question. A transit corridor is mostly passers through. A plaza with seating, a venue queue or a bar strip has far more dwellers. That observation should shape the content, and it is one of the things crews record during the night survey.

## How long the loop should actually be

I will not give a single number, because the right answer depends on the location, and anyone giving you a universal figure has not surveyed your site.

What I will give is the reasoning.

Short loops favour the glancing viewer. The shorter the cycle, the shorter the maximum wait before the key message appears. If somebody looks up at a random moment, a short loop guarantees they see the anchor quickly.

Long loops favour the dweller and allow more content, but they increase the chance that a passing viewer sees only an unimportant portion.

The practical compromise most locations want is a loop short enough that the key message appears frequently, with the anchor held for a meaningful stretch within it rather than flashing past.

That last point is where most content goes wrong. A loop can be the right total length and still fail because the logo or the proposition occupies a second and a half of it. The proportion of the loop spent on the message matters more than the total duration.

So the questions I ask are: what is the single thing a viewer must take away, how much of each cycle is that thing actually on the wall, and how long is the worst case wait for someone who looks up at the wrong moment.

If the answers are "the logo", "about five percent" and "most of a minute", the content needs restructuring regardless of how good it looks.

## Structuring the loop

The shape that works consistently is simple.

Open on the anchor, meaning the strongest single statement, held long enough to be read at a glance by someone walking.

Move through supporting content, which can be more visual, more atmospheric, more playful, and which gives the dwelling viewer something to enjoy.

Return to the anchor.

Repeat, with variation in the supporting content across cycles if the content volume allows.

The join between the end and the start of the loop should be seamless. A visible jump is distracting and it signals to attentive viewers that they are watching a repeat, which breaks the small illusion that the building is doing something.

Avoid a hard blackout at the loop point. On a facade, a full frame going dark reads as the building switching off, and doing that on every cycle is noticeable from the whole street.

Pacing should be slower than instinct suggests. At architectural scale, what feels like a measured cut on a monitor can feel abrupt, and what feels like a gentle camera push becomes a large sweeping motion across a wall. Restraint reads better and it fatigues the repeat viewer less.

## Scheduling across the night

Content length is only half the question. The other half is what plays when.

The usable window starts at genuine darkness, not at sunset, because twilight carries real light in the sky for a while afterwards. In a northern city in winter that window opens early and overlaps completely with evening street activity. In summer it opens late, after much of the foot traffic has gone. This is a scheduling fact that should influence the whole plan, and it is why I treat projection as a strong cold season format.

Within the window, street population changes considerably. An early evening audience is different from a late one in volume, composition and mood. Where a campaign has multiple content variants, matching them to those phases is worth doing. The heaviest message can sit in the hours with the most people, and lighter or more atmospheric content can carry the quieter stretches.

Run hours should match when people are actually present rather than how late the equipment could run. Lighting a facade at 2am on a corridor that empties at eleven is effort spent on nobody. I would rather concentrate the run into the hours that matter.

Venue and event contexts have their own rhythms. Arrival, interval and exit periods concentrate audiences dramatically, and content scheduling around those moments outperforms a flat run.

## Playback on the night

Operationally this is simpler than people expect, and the complications are mundane.

The crew loads content ahead of darkness and confirms it plays correctly before it matters. Opening a file for the first time on site is how avoidable problems become expensive, which is why we ask for delivery early enough to do a technical check.

Someone stays with the playback for the duration. Projection is not a format you switch on and leave. Equipment needs watching, the image needs occasional checking, and somebody has to be present to respond if anyone asks a question or if the site requires the projection to stop.

A static fallback is kept ready. If there is any playback issue, a strong still frame on the wall is far better than a dark building, and a good still is frequently what ends up being photographed anyway.

Mid run checks happen by walking. Someone leaves the projector position and looks at the wall from the opposite footway, from the approaches, and from where people are actually standing. From behind the lens everything always looks fine. The audience view is the only one that counts, and that walk routinely produces an adjustment.

Adjustments during a run are legitimate. If the anchor is not holding attention, extend it. If a section is washing out as shopfront lighting changes, reframe or reorder. Treating the schedule as fixed when the street is telling you otherwise is a wasted opportunity.

## Crew reality

A single position projection runs with a small crew, typically two to four people.

One lead on position, optics and alignment. One on playback and content operation. One on power and cabling. One handling public interface, meaning the person who talks to security staff, building concierges and curious passersby, and who holds the written permission and the authorising contact.

Setup begins well before dark. Position confirmed, power run, rough alignment done, content loaded and tested.

Then a wait for darkness, which in summer is a long and slightly dull part of the evening and in winter is barely a wait at all.

Final focus and framing once it is genuinely dark. Then the audience walk. Then the run.

Teardown is quick. Equipment loaded, cables coiled, position photographed, gone well within an hour. If further nights follow, position marking is confirmed before leaving so the next setup is faster and consistent.

## Permission and the grey area

Where projection can run is a genuine grey area. Ordinance language varies by municipality, much of it predates the format, enforcement is inconsistent between cities and sometimes between neighbourhoods, and the practical decision maker on any given block is usually a property owner, a venue operator or a district organisation rather than a specific municipal desk. I am not going to characterise any location as permitted or prohibited, and nothing here is legal advice.

Loop and duration touch permission in two specific ways.

Agreed hours are part of the permission. If an owner has agreed to a window, the run ends when that window ends even if the street is busy. Overrunning an agreed time is a small thing that costs access, and access is the thing that lets a campaign come back.

Repetition affects neighbours. A loop playing for hours is a persistent presence for anyone living or working nearby, and persistent presence is what converts mild curiosity into a complaint. Crews check for light spilling into occupied windows and adjust framing, and in residential adjacent locations I would shorten run hours rather than push them.

The format's structural advantage is that it leaves nothing behind. No residue, no removal night, no restoration question. If anyone with standing over the site objects, the remedy is instant. Switch it off, and the building is exactly as it was. That is why owners who would refuse anything applied to a facade will often consider light on it.

## Running several messages on one wall

Campaigns often arrive with more than one thing to say. A product line with three items, a tour with several dates, a brand with a primary proposition and a secondary call to action. The instinct is to put them all into one loop and let them take turns.

That works, but only if each message gets enough of the cycle to be received, and the arithmetic gets unforgiving quickly. Three messages sharing a loop means any given viewer has roughly a one in three chance of catching the one you care about most, and a passing viewer gets one look.

So we rank them. There is a primary message that appears every cycle without exception, and secondary messages that rotate between cycles. The primary is what the campaign is actually for. The secondaries are a bonus for the dwelling viewer who sees several cycles.

The alternative, where the location and the schedule allow it, is to separate messages by time rather than by cycle. One message carries the early evening, another carries the later hours, each getting the whole loop to itself during its block. This suits locations where the street population genuinely changes across the night, and it gives every message a clean run at a real audience rather than a third of a cycle.

What does not work is equal weighting by default, which usually reflects an internal compromise rather than a decision about what the street needs to know.

## What we talk clients out of

Long narrative content. Viewers arrive mid loop and leave after a glance. A story most people see one chapter of is not a story.

Repurposing a broadcast or social cut. Wrong aspect ratio, wrong pacing for a moving audience, and usually a dark grade, which on a projection reads as absence rather than as shadow.

Spending the loop on everything except the message. If the proposition occupies a sliver of each cycle, most viewers will not receive it.

Running until the early hours because the booking allows it. Match the hours to the people.

Hard blackouts at the loop point. The building visibly switching off every cycle is noticeable and unflattering.

Fast cutting. At architectural scale it is disorienting, and for the dwelling viewer it is exhausting.

Summer evening projections aimed at an audience that has already gone home. If a campaign needs evening reach in June, I would look at [LED Trucks](https://americanguerrillamarketing.com/led-billboard-trucks/) and [Mobile Billboards](https://americanguerrillamarketing.com/led-billboard-trucks/), which work in daylight and go to where people are, and hold projection for the darker half of the year.

Treating the loop as finished at delivery. The street teaches you things on the first night that no review can.

## Getting the length right before you produce

The sequence that avoids rework is straightforward.

Survey first, at night, at the real campaign hour. Record who is actually on that street, whether they are passing through or dwelling, how long a typical viewer has the wall in view and where they are standing.

Decide the single takeaway. One thing. Everything else in the loop supports it.

Decide how much of each cycle that takeaway occupies and what the worst case wait is for a random arrival. Set the loop length from those two numbers rather than from a creative preference.

Build each segment to stand alone, and build the anchor to be legible to someone walking past without stopping.

Produce a static fallback at the same time, not as an afterthought.

Deliver early enough for a technical check before the first night.

Then plan to adjust after night one, because you will want to.

## Frequently asked questions

### How long should projection advertising content be?

There is no universal figure, because it depends on the location and the audience. The governing numbers are how much of each cycle the key message occupies and the worst case wait for someone who looks up at a random moment. Shorter loops serve glancing passers by better, while longer ones suit locations with seated or waiting audiences.

### Does narrative content work in projection advertising?

Rarely. Viewers arrive at a random point in the loop, so understanding a later moment that depends on an earlier one is impossible for most of the audience. Build each segment to stand alone instead. A useful test is to pause the content at twenty random points and check whether each frozen frame communicates something by itself.

### How do you design a loop for people who see it repeatedly?

Use a short loop with a consistent anchor moment plus variation around it. The anchor guarantees that a glancing viewer gets the message quickly, while varied supporting content keeps the piece tolerable for someone at a bus stop or outside a bar who sees it many times across an evening. Slow pacing fatigues repeat viewers far less.

### What hours should a projection campaign run?

From genuine darkness rather than sunset, and only for as long as people are actually on that street. Winter gives a long usable window that overlaps evening activity, while summer darkness arrives after much of the foot traffic has gone. Agreed hours with the property owner are part of the permission, and we end when that window ends.

### Can you change the content during a projection run?

Yes, and we often should. Crews walk out from the projector position mid run to view the wall from where audiences actually stand. If the anchor is not holding attention or a section washes out as shopfront lighting changes, reordering or reframing during the run is legitimate. Treating the schedule as fixed wastes what the street is telling you.

### What happens if the playback fails on the night?

We keep a static fallback ready, because a strong still frame on the wall is far better than a dark building, and a good still is often what gets photographed anyway. Content is loaded and tested well before darkness rather than opened for the first time on site, and someone stays with the playback throughout the run.

## Request a campaign quote

If you are about to commission projection content, the loop length should come out of a night survey of the actual location rather than out of a creative meeting.

Send us the locations you are considering, the nights and hours you want to be live and the season. We will tell you who is genuinely on that street at that hour, whether they are passing or dwelling, and what loop structure will reach them.

Request a campaign quote and we will set the content length from the street rather than from the storyboard.