Is the salon mirror premium worth paying?

A media owner has quoted you a CPM for mirror displays in salons, several times what you pay for a screen in the same salon, and it looks like a premium. Before you argue about the number, check the premise — because the two are not the same product, and the mirror format has no published price anywhere in the world.

That changes the question. It is not “should I pay a premium over the salon rate?” It is: what does this format actually deliver, what is that worth per qualified person, and what is a fair price for something with no comparable?

This piece answers those three, in that order, from published prices and stated arithmetic.

The short answer

Three different things get called beauty DOOH, and only the cheapest of them has a published price. A screen somewhere in a salon lists at $7.84 (Adomni — US national average, all-in list price, undated). A mirror-integrated screen is a different product and trades in a single market, in a $40–50 band, with no published benchmark anywhere (single-market observation).
Across ten major US markets in Q2 2026, indoor inventory is quoted at office/elevator $10–22, gym $9–18, bar/restaurant $7–15, retail media $8–25, airport $18–50 (AdQuick — marketplace rates by city, Q2 2026). Nothing indoor is published above $50 except the largest airports.
Indoor pricing barely moves between cities — office-lobby midpoints run $15–17 across five markets, a 1.13× spread, against roughly 2.7× for outdoor premium in the same source. An indoor rate card is portable.
The mirror combines three delivery properties no other indoor segment has together: effective frequency inside one visit, a near-unique audience across the month (~222 different people per 1,000 impressions, against ~11 in a gym), and a scheduled return across the quarter (BDOOH model).
On the unit that decides campaigns — cost per effectively-reached in-market person — a broad buy runs about $0.40. The format’s current band runs $0.23–0.25, and parity is around $79 (BDOOH model).
The honest catch: no one has measured attention on a beauty in-venue screen. Buy it conditionally, on the terms in §6.

1. Check the premise: you are comparing two different products

Three distinct things get sold under the name “beauty DOOH”, and conflating them is the most common error in pricing this category:

	What it is	Where it exists as sellable inventory
A screen in a salon	Reception, waiting area, wall-mounted	Widely — Brazil, the US, most markets
A chair-side screen	A terminal per chair, positioned in front of the mirror	Japan — roughly 11,000 terminals as of July 2021, reaching 1.1m people a month (Digital Garage’s “Sakizaki Teruko”, via Dentsu — trade, figure now five years old)
A mirror-integrated screen	The display is the mirror the client faces during the service	A single market. No published price anywhere

Only the first has a published price. That is the $7.84 figure quoted as “the beauty CPM” — a US national average for a screen somewhere in a salon, with no format specified (Adomni — all-in list price, single platform, undated). Brazil’s largest beauty network — 7,680 salon, barbershop and gym screens now available programmatically — is the same category at scale (trade report).

A wall screen and a mirror screen differ on exactly the property this whole argument rests on: whether the line of sight is fixed by the service. Pricing one off the other is a category error, and every multiple computed that way is wrong before it starts.

Where the format actually sits

For context, here is the published indoor market. Across ten major US markets in Q2 2026 (AdQuick — marketplace rates and benchmarks by city):

Indoor category	Band across the ten markets	Cities splitting it out
Airport	$18–50	10
Office / elevator	$10–22	5
Gym / health club	$9–18	2
Bar / restaurant	$7–15	4
Retail media in-store	$8–25 in all ten	10
Programmatic open exchange	$4–15	10
Screen in a salon	$7.84 national average	—

Two things stand out. Retail media is quoted identically in every market, which makes it a national template rather than a researched rate — treat it with suspicion. And indoor prices barely move between cities: office-lobby midpoints sit between $15 and $17 across five markets, a 1.13× spread, while outdoor premium in the same source varies about 2.7× between New York and Houston. An indoor rate card is portable; an outdoor one is not.

Against that grid, the mirror format’s $40–50 band sits above everything published except the top of the largest airports. That is the honest starting position — not a premium over $7.84, which is a different product, but a price at the top of the indoor market for a format nobody else sells. The rest of this piece is about whether it is worth being there.

2. What the segment actually delivers

Three properties decide it, and only the first is about time.

Effective frequency arrives inside a single visit. A 15-second spot in a loop returning roughly every ten minutes, across a service of about 45 minutes, delivers ~4.5 contacts to the same person in one appointment (BDOOH model — loop arithmetic). In a broad-reach plan frequency is something you buy, with all the duplication that implies. Here the dwell supplies it.

Within a month, the people are different people. This is the property buyers miss. Most indoor inventory resells the same population: an office worker takes roughly forty lift rides a month; a gym member visits three or four times a week (Captivate; fitness trade press — directional). The impressions accumulate and the audience does not change. A salon client comes in on a monthly cycle, so a month of impressions is close to a month of distinct people (BDOOH model — contacts per visit × visits per month):

Segment	Contacts per person per month	vs effective frequency of 3	Beyond the threshold	Unique people per 1,000 impressions
Salon mirror	4.5	1.5×	33%	222
Retail media / grocery	12	4.0×	75%	83
Airport	18	6.0×	83%	56
Bar / restaurant	20	6.7×	85%	50
Office / elevator	40	13.3×	92%	25
Gym / health club	90	30.0×	97%	11

A thousand salon impressions reach about 222 different people; a thousand gym impressions reach about eleven. And everything past three memory-forming exposures is over-frequency you paid for — the salon over-delivers by a third, the office by more than nine tenths.

And they come back, on a schedule. The usual price of a unique audience is that you cannot build frequency against it. Here you can: salon clients return monthly, so a quarter re-contacts the same person about three times, in clusters spaced weeks apart, without a single duplicated impression. A quarter of office inventory delivers ~120 contacts to one person; a gym, ~270. Recall compounds steeply with frequency — illustratively from about 9% spontaneous recall at one exposure to 56% at five (Lumen — directional) — and here it compounds on separate occasions rather than on a loop someone walks past forty times.

No other indoor segment has all three. That is the segment definition, and it is what a price should be argued from.

Why the contacts should also land better

Two structural facts, neither requiring a study. The line of sight is fixed by the service, not by the ad — a seated client faces the mirror because the treatment requires it, and no other indoor placement obtains its viewing angle as a by-product of what the customer came to do. And the room pre-qualifies the audience: everyone in the chair is in-market for hair or beauty and mid-consultation with a professional they pay for, so there is no data fee, no segment to build and no waste to strip out. The venue is the targeting — the advantage described in the cookieless advantage of DOOH, with category selection in which beauty categories convert on salon screens.

What this is not is an hour of attention. Long dwell buys many short repeated glances, not sustained looking, and the distinction is the whole of why long dwell is not long attention. The claim here is narrower and survives scrutiny: a higher share of contacts should clear the memory bar than in a scrolled feed.

3. The unit that decides it

You do not buy impressions. You buy people, reached often enough to remember you, and in the market for what you sell. So price both sides on that.

What one qualified person costs you today. Three published findings compound. Only about 30% of viewable digital ads are actually looked at (eMarketer — directional); 85% of digital ads receive under 2.5 seconds of active attention, below the durable-memory threshold (VCCP Media, 2025; Amplified Intelligence — directional); and most of what remains lands outside your category. At an effective frequency of three (BDOOH model):

Step	Result
Impressions per memory-forming exposure (~15% clear the bar)	~6.7
Impressions to reach one person 3×	~20
Media cost at a $6.53 CPM (Place Exchange, H1 2025 — primary)	$0.131
Divided by the ~33% who are in-market	~$0.40 per qualified person

Your sticker CPM is roughly a thirtieth of what you actually pay to move one qualified person. And it moves with your own inputs:

Share clearing the memory bar	25% in-market	33%	50%	70%
10%	$0.784	$0.594	$0.392	$0.280
15%	$0.522	$0.396	$0.261	$0.187
25%	$0.313	$0.237	$0.157	$0.112

What one qualified person costs in a salon. At an assumed 60% of contacts clearing the memory bar, three memory-forming exposures need five impressions — all delivered to someone in your category. One visit supplies 2.7 of the three. Now the CPM appears, as an input:

Mirror CPM (input)	Cost per qualified person	vs a broad buy at ~$0.40
$20	$0.100	−75%
$40	$0.200	−49% — bottom of the format’s band
$45	$0.225	−43%
$50	$0.250	−37% — top of the format’s band
$60	$0.300	−24%
$79	$0.396	parity — the premium stops paying
$100	$0.500	+26%

There is a specific number — around $79 on these inputs — where the two are identical. Above it you are overpaying. Below it, the sticker shock is an artefact of the unit. Recompute it with your own in-market share in thirty seconds.

And the format’s current $40–50 band sits comfortably inside that. At today’s prices a mirror reaches a qualified person for 37–49% less than a broad buy, with about 1.6–2.0× of headroom before the efficiency argument runs out.

4. Holding the two halves together

The tension is real but narrower than it looks, and it is worth stating precisely.

§1 says the format sits above every published indoor price except the largest airports. §3 says it is still cheaper per qualified person than a broad buy, with headroom to about $79. Both are true, and they are not in conflict — they are measuring different things. One is a position in a price list; the other is an outcome per dollar.

What is uncomfortable is that the format has no external validation of either number. There is no second seller to check the band against, no published benchmark for the format anywhere, and no measured attention study to test the efficiency claim. A single-market product with one price and one model behind it is exactly the situation where a buyer should be most careful — and where a seller should be most explicit about what is assumption and what is evidence.

Three readings, and you should decide which you believe:

The format is fairly priced and under-appreciated — expensive against a price list built for other products, cheap against the outcome it delivers.
The efficiency maths is too generous, most likely on the 60% clearing assumption — see §5. Halve it and the headroom disappears entirely.
Both, which is the safe prior. The truthful position is a range, and the range is wide because the evidence is thin.

All three lead to the same conclusion: do not pay the top of the range on an argument. Pay a number you computed, against delivery you can inspect.

5. Four objections, answered

“There is no third-party measurement for this.” Correct, and it is the category’s biggest gap. Every attention figure here was measured in online display, online video or roadside large-format — not one on a salon screen. No published eye-tracking study of a beauty in-venue screen exists; we restate that rather than paper over it in attention benchmarks across media.

“The 60% clearing rate is doing all the work.” Also correct — it is the load-bearing assumption. Halve it to 30% and parity falls to about $40, at which point a $60 quote is above indifference and you should walk. Ask the seller to evidence it, or price as if it were lower.

“The reach is too small to matter.” True, and it should change how you use the channel, not whether. Chairs and appointment lengths cap total impressions, so this is a high-intent conversion layer inside a plan, not a reach vehicle — the interaction is in why OOH amplifies digital and cross-format reach and frequency.

“I can’t verify the screen was even on.” The real risk — and entirely solvable, by you, in the contract.

Two costs do quietly vanish and never appear in a CPM comparison: fraud, ad-blocking and made-for-advertising inventory — there are no bots in a styling chair — and distance from the decision, since the ad runs while a trusted specialist advises the client on her own hair or skin. Context: brand safety, physical vs digital and QR and O2O attribution.

6. How to buy it so the premium is conditional

Do not pay for an argument. Pay for delivery you can inspect. Five requirements a serious operator can meet and a weak one cannot:

Proof of play, at the display layer — and ask for occupancy. A schedule is not evidence. Demand per-play logs showing what was displayed, when, on which screen — the whole of proof of play: scheduling vs display, and the question separating a media business from a signage estate. And know that the bar is higher than that already: the closest analogous network in the world — chair-side salon terminals in Japan — has billed programmatically on “valid seated playback count” since June 2021, using seat-detection on the terminal so an impression only counts when someone was actually in the chair (Digital Garage × Microad Digital Signage — trade). That is not attention measurement, but it is a far stronger claim than “the screen was on”, and it is a five-year-old solved problem. A seller who cannot offer occupancy-verified plays is behind the state of the art, not at it.
An uptime SLA with automatic make-goods. Every hour dark is people you paid for and did not reach.
The multiplier, in writing. What footfall source, what visibility factor, what assumed contacts per client — the inputs to §3. If the seller cannot produce them, the per-person number is unpriceable.
Dayparts as separate line items. A Saturday colour appointment and a Tuesday walk-in are different dwell and different audiences — salon daypart patterns. Published indoor sources treat 20–40% daypart premiums as normal.
A measurement condition on the first flight. Matched-market or geo-holdout, or QR/O2O against a pre-agreed baseline, with a success metric stated before it becomes a line item: how to measure effectiveness.

Structure it that way and the question stops being “do I believe the attention story?” It becomes “did they deliver what they contracted?” — answerable from a log file.

7. What to pay

The premise was wrong, and the answer is yes — inside a range you can compute.

You were not being quoted a premium over the beauty rate. You were being quoted a different product: the $7.84 figure prices a screen somewhere in a salon, while a mirror-integrated display fixes the line of sight by the service itself. The mirror format trades in a single market at $40–50, above every published indoor price except the largest airports, and has no benchmark anywhere because no second market sells it.

What it delivers is a combination no other indoor segment offers — effective frequency inside one visit, ~222 unique people per thousand impressions, and a scheduled monthly return that turns a quarter into three spaced contacts. On the unit that decides campaigns, the current band reaches a qualified person for $0.23–0.25 against roughly $0.40 in a broad buy, and stops paying somewhere near $79.

So: the band is defensible and there is real headroom above it — roughly 1.6–2.0× before the efficiency argument runs out. Pay inside that range, never at the top of it on an argument alone, and only against evidenced delivery. Size it as a conversion layer against the chairs available. Recompute the parity point with your own in-market share and hold the seller to it. And treat the first flight as a test with a stated success metric — because the one thing nobody can yet give you is a measured attention figure, and until someone commissions it, every number on this page is arithmetic rather than evidence.

Methodology & caveats

Primary: the definition of an OOH impression as a likelihood/opportunity to see rather than verified attention (OAAA OOH Measurement Guide, 2022); the MRC viewability standard (50% of pixels, one second); the blended programmatic OOH CPM of ~$6.53, H1 2025 (Place Exchange — a single SSP’s blended, billboard-weighted series; full run and limits in our tracker).

The format’s own price band. The $40–50 range for mirror-integrated salon screens is a single-market observation, not a published rate card — the format is not sold as third-party inventory in any second market we can find, so there is no benchmark to check it against and no second seller to compare. Two adjacent formats do exist elsewhere and are not the same product. First, chair-side terminals positioned in front of mirrors: Japan’s “Sakizaki Teruko”, operated by Digital Garage and launched August 2020, reported over 10,000 terminals at 31 Mar 2021 and approximately 11,000 as of July 2021 with ~1.1m monthly reach — one terminal per chair, in front of the mirror. Since June 2021 it has sold programmatically through Microad Digital Signage’s MONOLITHS DSP on a “valid seated playback count” basis, with seat-detection on the terminal so plays are billed only when a client is seated. (Dentsu Web Dentsu-hō, 30 Nov 2021, Kana Misawa, Dentsu OOH Media Division; Digital Garage press release, 3 Aug 2020; syncAD, 10 Jun 2021 — all trade, and the deployment figures are five years old.) Second, general salon screens, sold widely — including a Brazilian network of 7,680 salon, barbershop and gym screens now available programmatically (trade report). Brand-owned smart mirrors installed by retailers in their own stores are excluded throughout: owning a screen and selling impressions on it are different businesses.

Published prices — US only, single marketplace each. Indoor CPM ranges by city for Q2 2026 (office/elevator $10–22, gym $9–18, bar/restaurant $7–15, retail media $8–25, airport $18–50, open exchange $4–15) from AdQuick city guides across ten major US markets, collected 2 Aug 2026; these are marketplace rates and benchmarks, not audited clearing data. The venue-type grid including Salon (Women) $7.84 / Salon (Men) $8.82, office building $18.78, veterinarian clinic $12.50, dental clinic $10.00, doctors’ waiting room $8.17, grid median $8.55, from Adomni — an all-in national average list price, undated, bundling media-owner rate, platform fee, data and service. Cross-source comparison is directional: the beauty figure is one platform’s national average, the comparator bands another platform’s dated city ranges, and no city-level beauty price is published anywhere. One unresolved discrepancy is flagged rather than smoothed: Adomni’s national office average ($18.78) sits above most cities’ local office bands, which a national average normally should not.

Directional (all measured off-channel — they transfer in principle, never as a beauty number): ~30% of viewable digital ads actually looked at (eMarketer); “85% of digital ads receive under 2.5 seconds” (VCCP Media, 2025); ~1.5s memory-encoding and ~2.5s durable-memory thresholds (Amplified Intelligence); frequency-to-recall 1 exposure ≈9% → 5 ≈56% (Lumen); per-exposure attention 1–2 seconds across media with OOH longest (Lumen / JCDecaux). Visit frequencies behind the duplication table: elevator ~30–60 seconds per ride and ~10 minutes daily dwell (Captivate); gym 3–4× weekly at 45–75 minutes (fitness trade press); airport dwell 60–120+ minutes (OOH trade); waiting-room wait 13.3 minutes (PatientPoint, 2024). Several originate with firms selling attention measurement or media.

Model (ours, reproducible): cost per effectively-reached in-market person = (CPM ÷ 1000) × (effective frequency ÷ clearing rate) ÷ in-market share. Worked inputs: effective frequency 3; broad-buy clearing 15% (arithmetic complement of VCCP’s 85%, applied to looked-at impressions); in-market share 33%; salon clearing 60%; loop arithmetic of a 15-second spot returning ~every ten minutes across a ~45-minute service ≈ 4.5 contacts per visit. Duplication table = contacts per visit × visits per month, with unique-people-per-1,000 as its reciprocal; the salon monthly cycle and the weekly cycles for bars and grocery are our assumptions. Every input here is a stated assumption, not a measurement — change any and the parity point moves, which is why the sensitivity tables ship alongside.

Not asserted — none exists: any measured attention figure for a beauty in-venue screen; any audited beauty or salon CPM benchmark; any city-level published beauty price; any non-US indoor price in this analysis. This article argues the segment is under-priced relative to what it delivers, up to a computed parity point, and should be bought conditionally on evidenced delivery — not that any of it has been independently verified.