Creative for salon & mirror screens

The tempting mistake with salon and mirror screens is to assume a captive, seated audience means a captive gaze — that because a client sits for 40 minutes, you have 40 minutes of their attention to fill with a long story. You don’t. Across every medium measured, per-exposure attention sits at roughly one to two seconds, even where dwell is long; what a salon actually buys you is many short glances across a loop, not one continuous stare (Lumen/JCDecaux — primary). That single fact reorganises how you build for these screens: design for repeated brief glances with a brand that’s recognisable in the first frame, run it silent because the room has its own sound, build it vertical for the mirror, and earn a QR only because the dwell is genuinely high. This guide is how. And the honest caveat up front, because it shapes everything: there is no verified salon- or mirror-specific creative-effectiveness study — so these are principles to test, not a benchmark to trust.

Key takeaways

Long dwell ≠ long attention. Per-exposure attention is ~1–2 seconds even in OOH; design for repeated short glances, not a 30-second gaze. The brand must read in the first frame (Lumen/JCDecaux — primary).
Distinctive brand assets do the work. As little as ~1.5 seconds of attention can encode memory if strong brand assets are present; well-branded assets are far more effective at low attention (Amplified Intelligence — directional, measured off-channel).
Silent by default. Salons have their own ambience and DOOH video generally carries no audio track — build to work muted, with captions not voice-over (Ströer spec — primary).
Vertical is the master. Portrait 1080×1920 is the dominant DOOH build and the natural mirror format; supply a landscape variant for lobby screens. One asset gets transcoded across many screen types (Vistar, The Trade Desk — primary).
A QR earns its place here. High-dwell venues are exactly where a code has time to be scanned — but only if it’s large, eye-level, and has a reason to scan (Vistar — directional).
Dynamic creative usually resolves to static at serve time — design DCO variants as static layouts, don’t assume animated HTML5 will run (Vistar — primary).

1. Design for the glance, even when dwell is long

Start from the counter-intuitive truth. Eye-tracking across desktop, mobile and OOH finds average dwell per exposure sits between one and two seconds, with OOH often at the higher end — and a couple of seconds is “plenty to build brand recall” (Lumen/JCDecaux — primary). A salon client is in front of the mirror for the whole service, but their attention is split with the stylist, the conversation and their phone, returning to the screen in short bursts. So you’re not designing one long ad someone watches start to finish; you’re designing for repeated 1–2-second glances across a loop.

That has two consequences for the build:

The brand has to land in the first frame. Distinctive brand assets — colour, logo, character, typeface — are what convert a short glance into recall: research finds as little as ~1.5 seconds of attention can encode memory if strong assets are present, and well-branded creative is markedly more effective at low attention (Amplified Intelligence — directional; measured in social feeds, transfers in principle not as a salon number). Brand early and persistently, not in a reveal at second 25.
Frequency is your friend. Repeated exposures compound recall, and a salon loop delivers many during one sit. So you can use the long visit — not for a long ad, but to sequence a message across the loop (a brand frame, then a product frame, then a CTA frame) that a returning glance assembles over the appointment.

2. Keep it simple — the high-dwell version

DOOH creative best practice is “keep it simple”: one idea per screen, few words, high contrast, large bold sans-serif type, the logo prominent, motion used sparingly to attract without burying the message (joint DPAA/IAB/OAAA best-practices framework — primary that it exists; the specific “6-8 words” figures are vendor rules of thumb — directional). One concrete, citable spec: keep text at least ~15 px in height so it’s legible (Vistar — primary).

High dwell doesn’t license clutter — it licenses sequence. The glance is still short, so each individual frame stays simple; what the long visit adds is the chance to tell a slightly longer story across frames (storytelling, product education, a sequential narrative) rather than cram it into one. The discipline: every single frame must pass the few-seconds test on its own, even though the loop as a whole can say more than a roadside board ever could.

3. Build it silent

Assume the screen is permanently muted. Salons own their own ambience — music, chatter, dryers — and DOOH is silent by convention anyway: platform specs are explicit, e.g. Ströer’s public-video spec is MP4, H.264, no audio track, 5–30 seconds (Ströer — primary). So:

Carry the message in type and captions, not voice-over. Write a caption script, not a narration script; pace edits to reading speed.
Make it work with the sound off, because it will be. Silent-video research from social and web (most video watched muted; captions lift completion) supports designing for sound-off — apply it by analogy, since it’s not measured on salon screens (off-channel — directional).

A spoken-word idea that only works with audio is the wrong idea for this screen.

4. Build vertical for the mirror, landscape for the lobby

Portrait 1080×1920 (9:16) is the dominant DOOH build — used by the large majority of panels — and it’s the natural fit for a mirror screen at a styling station; The Trade Desk’s standard vertical format (D6) is exactly this (Vistar, The Trade Desk — primary/directional). Practical rules:

Make portrait 1080×1920 the master for salon/mirror placement; supply a landscape 1920×1080 variant for lobby and standalone screens.
Keep critical content in a safe area. One creative gets transcoded, resized and letterboxed across many screen types — platforms run exactly this service — so a logo or CTA jammed to the edge can be cropped (Vistar — primary).
Treat the mirror as a standard portrait panel creatively. There’s no verified mirror-specific creative spec, and claims about half-silvered-glass contrast penalties aren’t substantiated — so design for a bright vertical screen, not an exotic surface (unverified — do not assert).

5. Use a QR — but earn it

This is the one place a QR code genuinely belongs in out-of-home. QR works specifically in high-dwell environments and fails on fast/roadside placements with under ~15 seconds of exposure — a seated salon client is the ideal case (Vistar — directional, on-thesis). But a code is wasted if it’s done badly. The rules that make it scan:

Size it for the distance (a rough 10:1 code-width-to-scan-distance ratio), place it at eye/camera level, and make it part of the artwork rather than a sticker in the corner.
Give a reason to scan with an explicit CTA — “shop the look,” “get the offer” — not a bare QR (Vistar; Media Leader caution — directional).

There’s no published salon-specific scan-rate uplift, so treat “dwell makes QR scannable” as sound logic to test, not a measured number. Done right, the high-dwell glance is what finally gives a code time to work.

6. Dynamic creative: design the variants as static

Contextual and dynamic creative (dayparting, weather, nearest-store, promos) lifts outcomes — vendor meta-analysis reports targeted creative outperforming untargeted on awareness and consideration (Vistar — directional). But the production reality constrains the build: many platforms render dynamic creative to a static image at serve time, animated dynamic creative is region-limited, and raw HTML5 is often not accepted (Vistar — primary). So design each DCO variant as a static layout that stands on its own, rather than assuming live animation will play on the panel. Dynamic relevance is upside; a clean static frame per context is the deliverable.

7. Creative specs reference

Build to spec, supply the master plus a landscape variant, and remember the universal rule — DOOH video carries no audio (primary/directional per platform):

Platform	Static	Video	Notes
The Trade Desk	D6 Vertical 1080×1920	MP4	Cross-SSP reach format
Ströer	—	MP4 / H.264, no audio, 5–30 s, ≤12 Mbps	The canonical no-audio spec
Vistar	JPG, ≤10 MB, text ≥15 px	MOV/MP4, ≤50 MB	HTML5 not accepted; transcodes/letterboxes; DCO → static image
Hivestack	JPG/PNG	MP4/MOV, under 10 MB	Common dims 1080×1920, 1920×1080, 1280×720
Broadsign	PNG/JPEG (no animated GIF)	MP4/MOV/WebM, H.264 ~8 Mbps	No published platform-wide file-size cap — per-network caps apply

A note on a figure you may see elsewhere: a “Broadsign 750 KB max” limit could not be verified in Broadsign’s own docs — caps in DOOH are usually per-network, not platform-wide, so confirm with your supply path rather than designing to an unverified number (unverified — do not assert).

8. The creative mistakes that waste the dwell

Assuming long dwell = long attention. The single biggest error; design for repeated short glances (§1).
Too much text / tiny type. Violates the few-words, ≥15 px legibility floor (§2).
Relying on sound. It’s muted; caption everything (§3).
Landscape-only art. Breaks on the portrait mirror screen; build vertical first (§4).
A tiny, reason-less QR. Wastes the one environment where a code can work (§5).
Assuming animated HTML5 will run. It often resolves to static — design the static (§6).
Weak or absent brand assets. Leaves recall on the table when attention is short (§1).

So — how do you make high-dwell salon creative work?

By designing for the glance the dwell actually gives you, not the gaze you imagine it does. Brand in the first frame with distinctive assets, because a 1–2-second look is what you get. Keep each frame simple but use the loop to sequence a longer story. Build it silent and vertical — captions not voice-over, portrait 1080×1920 master with a landscape variant. Earn a QR with size, placement and a reason to scan, because high dwell is the one place it pays off. And design dynamic variants as static, since that’s usually how they serve. Test these rather than trusting them blindly — there’s no salon creative benchmark to lean on — and you’ll make creative that converts short, repeated, in-context attention into the recall the channel is genuinely good at building. (For where these screens sit on a plan, see is beauty DOOH right for your brand?; for proving the work landed, how to measure effectiveness.)

A note on the numbers

Verified (primary): per-exposure attention of ~1–2 seconds across media with OOH often highest, and “2 seconds is plenty to build recall,” from Lumen × JCDecaux; the silent/no-audio norm from the Ströer public-video spec (MP4, H.264, no audio, 5–30 s) and DOOH-silent convention; portrait 1080×1920 as the dominant build and the transcoding/letterboxing reality, plus the ≥15 px text minimum, JPG ≤10 MB / video ≤50 MB, HTML5-not-accepted and DCO-renders-to-static facts, from Vistar and The Trade Desk; the existence of a joint DPAA/IAB/OAAA DOOH best-practices framework. Directional (vendor / measured off-channel — transfers in principle, not as a salon number): distinctive brand assets and ~1.5 seconds to encode memory (Amplified Intelligence, measured in social feeds); QR works in high-dwell / fails under ~15 s exposure and the size/eye-level/reason-to-scan rules (Vistar); targeted-vs-untargeted creative lift (Vistar meta-analysis, no n disclosed); silent-video stats (social/web research, applied by analogy); the “6–8 words” simplicity rules (vendor heuristics, not a standard). Could not verify — not asserted: any salon- or mirror-specific creative-effectiveness, scan-rate, recall or dwell number (none exists — the academic signage studies are general retail, e.g. a boutique); a “Broadsign 750 KB” file cap (not in Broadsign’s own docs — per-network caps apply); half-silvered mirror-glass contrast penalties (no spec found); “+17% contextual relevance” (untraceable). Test these principles in your own venues rather than treating any borrowed figure as a salon benchmark.