How Do I Measure
the ROI of My
Window Display Screen?
Your window screen is not unmeasurable. AI sensors now track Opportunity to See, Dwell Time and anonymous gaze patterns — giving every UK retailer the same data rigour as programmatic digital advertising. Here is exactly how it works.
For years, the standard answer to "how do I know if my window screen is working?" was a guess. Before and after footfall counts. A feeling. A sense that things were busier. That era is over. AI-powered audience measurement tools now give your shop window the same data infrastructure as a Google display campaign — counting who saw it, how long they stopped, and what happened next. This guide explains all of it, from first principles to practical implementation.
1. The Measurement Problem — Solved
The question "is my window screen worth the investment?" has historically been difficult to answer with rigour because window screens operate in the physical world — where unlike a website or social media ad, there is no natural click, no session data, no automatic attribution trail between "someone saw your screen" and "someone bought something."
This problem has been the central challenge of out-of-home (OOH) advertising for decades. And in 2025–2026, the industry systematically solved it. The IAB published a comprehensive DOOH (Digital Out-of-Home) Measurement Guide in 2025 that defines standardised metrics, measurement methodologies and privacy-compliant data collection standards — bringing OOH into full parity with digital advertising measurement.
The IAB's DOOH Measurement Guide introduced standardised definitions for Opportunity to See (OTS), Likelihood to See (LTS), Dwell Time, Engagement Rate and Attribution — from "what to count" to "how to count it." For the first time, a window screen delivers the same measurement vocabulary as a programmatic display campaign. Your window screen can now be measured, benchmarked and optimised with the same rigour as any other advertising channel you invest in.
2. The Metric Hierarchy — From Exposure to ROI
Audience measurement for a window display screen moves through a hierarchy of increasing specificity — from the broadest possible count (everyone who walks past) to the tightest possible attribution (purchases linked to screen exposure). Understanding where each metric sits in this hierarchy tells you what it measures and how reliable it is.
3. OTS — What It Means and Why It Matters
Opportunity to See (OTS) is the foundational currency of out-of-home advertising measurement. It counts every person who passes within the defined exposure zone of your screen — the area in which a person of standard height, facing the display direction, could physically see the screen.
For a UK high street retailer, OTS is calculated by combining:
| Input | How It's Measured | What It Represents |
|---|---|---|
| Pedestrian footfall | AI body-counting camera or manual count | Total population passing the screen |
| Exposure zone width | Screen size + viewing angle calculation | Area in which the screen is visible |
| Screen display status | CMS uptime monitoring | Whether the screen was actually showing content |
| Viewing distance | Screen size + content legibility testing | Maximum distance at which content is readable |
| OTS impressions | Footfall × exposure zone fraction × uptime % | Standardised impression count for your screen |
OTS is the data that justifies the investment in your window screen — it translates an anecdotal "lots of people walk past" into a measured, defensible number that you can benchmark against the cost of the screen and the content. A screen generating 50,000 OTS impressions per week at a total weekly cost (screen amortised + CMS + content) of £20 is delivering impressions at £0.0004 each — which is competitive with or better than most digital advertising channels.
4. Dwell Time — The Signal That Content Is Working
Dwell Time is the metric that separates passive exposure from genuine engagement. You can have high OTS — many people walk past — but if nobody stops or slows down, the content is not performing its primary function: arresting attention at the buying decision moment.
Research consistently shows that interactive and animated displays increase dwell time by up to 40% compared to static signage. More time on your content means stronger recall, higher conversion probability and measurable ROI improvement. Dwell Time is the single best proxy metric for content effectiveness available to window screen operators.
What Dwell Time Benchmarks Look Like for UK Retail Windows
| Dwell Time | What It Suggests | Content Implication |
|---|---|---|
| Under 1 second | Glance only — no engagement | Content not capturing attention — review visual hierarchy and motion |
| 1–3 seconds | Brief notice — standard walk-past | Screen is visible and functional; content being registered but not arrested |
| 3–8 seconds | Active viewing — content working | Good performance; content driving short engagement — typical for well-designed screens |
| 8–20 seconds | Extended engagement — high interest | Excellent; viewer processing content in detail — strong conversion probability |
| 20+ seconds | Deliberate stop — peak engagement | Exceptional; viewer has stopped specifically for your screen — near-certain walk-in consideration |
5. How AI Anonymous Video Analytics Actually Works
The technology behind OTS, dwell time and gaze measurement is called Anonymous Video Analytics (AVA) — computer vision systems that use a camera to detect and count people, measure their proximity and presence duration, and optionally estimate head pose and gaze direction. Here is exactly what happens:
A small camera captures the field of view in front of the screen
The camera can be integrated into the screen frame or positioned separately. It covers the exposure zone — the area in which passers-by could realistically see the screen content.
AI model runs locally on the device — no images sent to the cloud
This is the privacy-critical step. All processing happens at the edge — on the device itself. The system detects body silhouettes, estimates proximity and measures dwell duration. No image frames are transmitted. No faces are identified. No biometric data is created.
The system outputs statistical data — never individual records
The output is aggregate data: "47 people passed between 10:00 and 10:15, average dwell time 3.2 seconds, 28% had head pose directed at screen." No individual is tracked, named or identifiable. The data is mathematically equivalent to a person standing at the window and counting with a clicker.
Head pose estimation identifies whether viewers were looking at the screen
Advanced systems add head pose estimation — detecting whether a person's face was directed towards the screen (within a defined angle threshold). This produces the "attention" or "watcher" metric — the most qualified engagement measure. Still entirely anonymous: the system detects direction, not identity.
Aggregated data populates a CMS analytics view or connected dashboard
The statistical output feeds into your CMS analytics — showing OTS impressions by hour, dwell time by content type, stop rate over time, and demographic estimates (age range, group size) where supported. This data drives content optimisation decisions: what runs at what time of day, which creatives drive higher dwell time, and whether seasonal changes affect engagement.
6. Privacy & GDPR — What This Is and What It Isn't
This is the question every UK business owner immediately asks — and the answer is unambiguous when the system is properly implemented.
- Body presence detection — counting silhouettes, not identifying people
- Proximity measurement — how far from the screen
- Dwell time counting — how many seconds presence detected
- Head pose estimation — direction of gaze, not identity
- Demographic approximation — age range bands, group composition
- All processing on-device — no images transmitted to any server
- No biometric data created or stored anywhere
- Output is statistical aggregates — never individual records
- Equivalent legally to counting customers entering a shop
- Facial recognition — not identifying who someone is
- Biometric data collection — no templates, no hashes, no face vectors
- Individual tracking — no person is followed across time or space
- Personal data under GDPR — no identifiable individual is processed
- CCTV — footage is not recorded or stored
- Surveillance — no linking of behaviour to individuals
- Consent-requiring activity — anonymous statistical counting does not require individual consent under UK GDPR
Anonymous aggregate statistical data that cannot be used to identify an individual does not constitute personal data under UK GDPR. AVA systems that process all data at the edge and output only aggregate statistics are therefore not subject to the consent requirements of the GDPR — they are legally equivalent to a human counter standing at the shop door. Always verify the specific implementation with your signage technology supplier — "privacy by design" must be confirmed, not assumed.
7. The ROI Calculation Path — From OTS to Pounds
Measure Impressions
How many people passed within viewing range of your screen this week?
Measure Engagement
Of those, how many stopped? For how long? Is dwell time increasing over time?
Measure Conversion
Are footfall counts inside the store rising proportionally to OTS? Track door count vs screen impressions.
Attribute Sales
Correlate screen content changes with revenue data. Did the promoted item sell more during high-OTS periods?
Calculate CPM
Total weekly screen cost ÷ (OTS ÷ 1000) = cost per thousand impressions. Compare against your other channels.
A typical UK high street window screen generating 50,000 OTS impressions per week at a total annualised cost (screen + CMS + content) of approximately £1,600/year — that is £31/week — delivers impressions at a CPM (cost per thousand) of £0.62. For comparison: Google Display Network CPM averages £1.50–£3.00 in the UK; Facebook/Meta averages £5–£12 CPM. Your window screen is competing on pure impression economics with the most efficient digital ad platforms — while reaching people at the exact physical moment they are near your premises.
8. What UK Independents Can Do Right Now
You do not need specialist hardware or a signage network analytics contract to start measuring your window screen's performance. Here is the practical implementation path, from zero to full analytics:
| Stage | Method | What You Measure | Cost |
|---|---|---|---|
| Stage 1 — Baseline | Manual footfall count before screen installation. Count walk-ins for 2 weeks. Note daily sales. | Baseline walk-in rate and revenue | £0 — staff time only |
| Stage 2 — CMS Analytics | CleverPosters AI analytics tracks content performance. Note which content plays at which times and correlate with sales data. | Content engagement patterns, optimal display times | Included in CleverPosters £120/year |
| Stage 3 — QR Attribution | Add a unique QR code to your screen content linking to a landing page or offer. Track scans as direct attribution events. | Direct walk-in attribution from screen exposure | Free QR code generation + Google Analytics |
| Stage 4 — A/B Content Testing | Run two content variants on alternate weeks. Compare walk-in rate and promoted item sales between periods. | Content effectiveness comparison | No additional cost — CMS scheduling |
| Stage 5 — Full AVA | Add anonymous video analytics camera system to your screen. Dashboard shows OTS, dwell time, stop rate and demographic estimates. | Full OTS, LTS, Dwell Time, Gaze metrics | £200–£800 hardware + analytics subscription |
Most UK independent retailers running Khazina Digital window screens reach Stage 3–4 within their first month of operation, providing solid evidence of ROI without any additional hardware investment. Stage 5 full AVA measurement is increasingly available as an add-on feature from specialist signage analytics providers and is becoming the standard for any retailer operating multiple window screens or selling advertising space on their screens.
Frequently Asked Questions
Shop High-Brightness Window Displays
From £499+VAT. Free bespoke animated design. Measurable OTS impressions from day one. UK supplier since 2013.