Attention benchmarks across media

Attention is becoming the currency advertisers actually care about — but the real numbers surprise people. Per-exposure attention across every measured medium is counted in seconds, not minutes, and it takes startlingly little of it to build a memory. This benchmark gathers the verified cross-media attention figures, shows where OOH sits, and is honest about the one number that doesn’t exist: a measured in-venue beauty figure.

Key findings

Per-exposure attention is ~1–2 seconds across virtually all media — and OOH receives the longest dwell of the channels measured (Lumen/JCDecaux — directional).
It takes very little to land: ~1.5 seconds of genuine attention can encode a memory when distinctive brand assets are present; enduring effects build above ~2.5s (Amplified Intelligence — directional).
Frequency compounds recall — illustratively, 1 exposure ≈ 9% spontaneous recall, 5 ≈ 56% (Lumen — directional).
Premium large-format DOOH has been measured at ~5× the attention of online formats and ~2.5× the recall of online video (Ocean Outdoor/Lumen — vendor-commissioned, roadside).
The caveat that matters: every figure here is general or large-format OOH. No verified in-venue beauty attention number exists — and dwell is not attention.

The headline: attention is measured in seconds

The single most clarifying fact in attention research is how short attention is. Eye-tracking across desktop, mobile and OOH finds average per-exposure dwell sits between one and two seconds, with OOH typically at the higher end — pedestrians and large formats hold attention longer than scrolled digital (Lumen/JCDecaux — directional, vendor-commissioned). For context, only about 30% of viewable digital ads are actually looked at (eMarketer — directional); “viewable” (50% of pixels for a second, per the MRC standard) is not the same as “seen.”

So the realistic unit of attention is a second or two of genuine looking — and the design question is what you can do with it.

How little it takes to build a memory

The encouraging counterpart to “attention is short” is “short can be enough.” Real-world attention research finds that as little as ~1.5 seconds of active attention can encode a memory — if the brand’s distinctive assets are present and recognisable — with more durable memory effects building above roughly 2.5 seconds, and active attention driving up to ~3× more recall than passive exposure (Amplified Intelligence — directional, measured in digital feeds). Separately, a couple of seconds of OOH attention has been linked to a measurable lift in brand recognition (Lumen — directional).

The practical reading: the first frame has to carry the brand. With only seconds of attention, recall comes from distinctive, instantly-recognisable assets — colour, logo, character, typeface — not a slow reveal.

Frequency compounds

A single short look is one thing; repeated looks are another. Attention accumulates across exposures, and recall climbs steeply with frequency — illustratively, one exposure ≈ 9% spontaneous recall, rising to ≈ 56% at five exposures (Lumen — directional). This is why a high-dwell environment that delivers many short glances at the same loop can outperform a single longer exposure: the frequency does the work the individual glance can’t.

Where OOH sits versus digital

OOH’s attention case, on the verified evidence, is comparative strength on a per-impression basis. A vendor-commissioned large-format study put premium roadside DOOH at roughly (Ocean Outdoor/Lumen — directional, OOH-commissioned):

Metric (premium large-format DOOH vs online)	Finding
Attention volume	~5× more than online formats
Holds attention vs online display	~8.2× longer
Holds attention vs social	~5.5× longer
Holds attention vs online video	~1.6× longer
Brand recall vs online video	~2.5× higher

Treat these as the directional attention case for the medium, not a salon number — they’re measured on premium roadside billboards and commissioned by an OOH owner. The emerging way to compare across channels in a single metric is an attention unit (e.g. Adelaide’s “AU,” an outcome-trained media-quality score that benchmarks OOH alongside other channels) — a sign the industry is converging on attention as the comparison currency (Adelaide — directional).

The in-venue caveat

Here is the number you’d most like and cannot have: a measured attention figure for beauty in-venue screens does not exist. Every benchmark above is either cross-media (Lumen/Amplified) or large-format roadside (Ocean/Lumen) — none is in-salon. Worse, the intuitive substitute is wrong: a salon’s long dwell is not attention. A client seated for a 60-minute appointment is in the screen’s exposure zone for an hour, but their attention returns to the screen in short bursts between conversation, phone and the service itself. Long dwell buys many short, repeated, in-context glances — which, given frequency compounding and the 1.5-second memory threshold, is a genuinely strong proposition. It is simply not “an hour of attention,” and claiming so is the fastest way to lose a sophisticated buyer. We unpack exactly this in Why long dwell ≠ long attention.

Methodology & caveats

Directional (all measured off-channel — transfers in principle, not as a salon number): per-exposure attention ~1–2 seconds with OOH longest, the brand-recognition lift, and frequency-to-recall (1 exposure ≈9% → 5 ≈56%) from Lumen / JCDecaux; ~1.5s to encode memory with distinctive assets and the active-vs-passive ~3× from Amplified Intelligence (measured in digital video feeds); the premium large-format DOOH attention/recall multiples (~5× attention, ~2.5× recall vs online video) from Ocean Outdoor / Lumen — vendor-commissioned and measured on roadside large-format, not in-venue; the attention-unit (AU) approach from Adelaide. Verified (primary): the MRC viewability standard (50% of pixels, 1–2s) underlying “viewable ≠ seen,” and ~30% of viewable digital ads actually looked at (eMarketer — directional). Not asserted — none exists: any measured attention figure for beauty/salon/spa/nail-bar screens. The in-venue attention case is argued on first principles (high dwell → repeated glances + frequency + low memory threshold), not from a measured beauty study.