Digital Signage
Part of Digital signage in retail
Measuring whether a signage placement is noticed
Use defined viewing areas, playback checks and consistent observation to assess whether a retail screen is noticed.
To test whether shoppers notice a signage placement, observe people with a realistic chance to see the target message while it is visible. Count that opportunity separately from an apparent glance or a later action. A playback log alone proves neither attention nor understanding.
Define who had an opportunity
Mark a viewing area from which the message can reasonably be read. Walk normal routes through it and note glare, shelving, queues and other obstructions. Specify the message and observation periods.
If the target message appears in a rotation, a person passing while another item plays has not had an opportunity to see that target message. Exclude periods when the screen is off or the view is blocked under the rule you set in advance.
For a small manual check, record separate counts using written definitions:
- Eligible passes:passages through the defined area while the target message is visible and a usable view exists.
- Observed glances:eligible passes during which the person appears to look towards the screen under a consistently applied rule.
- Visible actions:eligible passes followed by a defined nearby action, such as approaching the indicated range.
Report observed glances divided by eligible passes for the sampled periods, with both counts and the rules. Call this a local observed-glance rate; it does not show what anyone read or remembered.
Count passes rather than unique shoppers unless the method can reliably distinguish repeat visitors. A glance towards the screen also cannot prove that the person saw the target frame rather than another item or something beside it.
Key metrics for measuring signage visibility
- Eligible passesTotal number of shopper passages through the defined viewing area with clear line of sight and message visible.
- Observed glancesNumber of eligible passes where shopper appeared to look at the screen, per agreed rule.
- Local observed-glance rateObserved glances ÷ Eligible passes, reported with counts and methodology.
- Uncertainty countNumber of ambiguous or blocked observations recorded but not classified as 'noticed'.
Observe consistently
Choose periods that represent the trading conditions the placement is meant to serve. Record date, time, staffing, promotions, screen status, competing signs and unusual events. Use the same area boundary and glance rule each time. A second observer can classify a short shared period independently; disagreement shows where the definition needs work.
Keep observers out of shoppers’ way. Record blocked views and ambiguous gaze movements as uncertainty rather than forcing them into a noticed count. Do not infer understanding from a head turn.
Pre-observation checklist for signage placement
- Use representative trading periodsInclude typical staffing levels, promotions, and store activity.
- Fix observation boundaries and rulesApply the same area definition and glance criteria across all observations.
- Avoid observer interferenceKeep observers out of shoppers’ paths to prevent influence.
- Document uncertaintyRecord blocked views or unclear gaze movements as uncertain—not counted as 'noticed'.
- Do not infer understandingA head turn does not prove recognition or comprehension of the message.
Separate delivery, exposure and effect
Treat target-message playback, eligible passes and observed glances as separate observations. A playback record or eligible pass does not establish that the message was seen. The manual glance rate above is a proposed local measure.
Keep separate fields for screen status and target-message playback, eligible passes, observed glances and any relevant action. A later purchase is another outcome. Stock, price, staffing and other promotions can affect sales, so a placement observation alone cannot attribute a sales change to the screen.
Key distinctions in signage measurement
- Target-message playbackScreen shows the message – no evidence of viewer attention.
- Eligible passesShopper passes through viewing area while message is visible and view is clear.
- Observed glancesShopper appears to look at the screen during an eligible pass, per consistent criteria.
- Visible actionsShopper approaches the indicated product range or engages with a service point after passing the screen.
Compare a defined change
If the observed-glance rate is weak, change one controllable feature where practical: position, angle, obstruction, headline or timing. Observe comparable periods using the same rules before and after. An unchanged comparison placement can add context, but it will not remove every difference between periods.
Few eligible passes point towards the location. Many passes with few apparent glances call for a sightline or message review. Glances followed by the same unresolved shopper question call for a content or service-path review. These are possibilities to check, not findings about a particular store.
Check privacy before adding sensors
Manual aggregate counts may answer the placement question without storing images. Before using a camera or other sensor, establish what it captures, whether people can be identified, access, retention and applicable privacy and surveillance requirements.
The OAIC says personal information collected through surveillance devices must comply with the Australian Privacy Principles where the organisation is covered by the Privacy Act; state and territory surveillance laws may also apply. A vendor’s ‘anonymous analytics’ description does not settle how the proposed system handles data.
Report the sampled locations, periods, counts, exclusions and uncertainty. Use the result to guide a local placement decision, not as a universal attention rate or proof that signage caused a purchase.
Manual observation vs. sensor-based tracking
- Pros of manual observationNo personal data storage; simpler compliance with Australian Privacy Principles (APPs); suitable for small-scale testing.
- Cons of manual observationTime-intensive; limited scalability; potential for observer bias without double-checking.
- Pros of sensor-based trackingEfficient for large volumes; enables real-time analytics and pattern detection.
- Cons of sensor-based trackingMay collect identifiable personal information; requires strict adherence to APPs and state surveillance laws; vendor claims of 'anonymous' data must be verified.
