Calling Names vs Digital Alerts: Which Works?
Calling names vs digital alerts doesn’t have a winner. Each method works for a different customer and fails for a different one, and both failures are silent. The person who missed it is standing in the room, and never says a word.
One method means a member of staff calling a name or number out loud, across a counter or a waiting room. The other means a digital alert: a screen, a pager, or a message sent straight to a customer’s own phone. Both are meant to do the same job, and both quietly miss people the venue never notices.
Key Takeaways
- A called name only reaches someone within earshot who can pick it out of the room.
- A digital alert only reaches someone whose phone is charged, connected and being looked at.
- The worst-measured indoor speech intelligibility in a 1,600-bed Indian hospital was recorded at the registration desk, exactly where names are called (Indian Journal of Public Health, 2026).
- Around 1 in 3 people in the UK are deaf, have hearing loss or tinnitus: over 18 million people (RNID, 2026).
- The median smartphone notification gets swiped away or opened in about 21 seconds, from data collected between October 2018 and October 2020 (Sensors, 2024).
- Where a loudspeaker call and an on-screen display ran side by side, both staff and patients rated the screen higher (Indian Journal of Radiology & Imaging, 2025).
- No search of the recent literature turned up a trial that isolated the calling method itself and measured missed calls at a counter.
What does each one ask the customer to do?
Calling names vs digital alerts is really a question about preconditions, not features. A called name needs a customer close enough to hear it and alert enough to notice it. They also have to pick it out of whatever else is happening in the room. A digital alert needs a different set: a phone that’s charged, connected and being watched, or a screen the customer can see.
The table below sets out what has to be true for each method to reach someone, before a single word or message is even sent.
| Precondition | Called name | Digital alert |
|---|---|---|
| Within earshot of the counter | Required | Not required |
| Able to pick the message out of room noise | Required | Not required |
| Able to see a screen from where you’re standing | Not required | Required, for a screen or pager display |
| Phone charged and switched on | Not required | Required, for a phone alert |
| Signal or a working connection | Not required | Required, for a phone alert |
| Actually looking at the right moment | Not required, sound reaches you | Required, for a screen or a phone |
Neither precondition is exotic. Both fail constantly, in perfectly ordinary rooms, for perfectly ordinary reasons.
None of this is about cost or hardware. What each notification method costs a restaurant covers the price side of pagers, screens and messaging platforms. The full set of ways to tell someone their order is ready lays out the whole menu a venue can choose from. What matters here is narrower: whether the message physically reaches the person it’s meant for, and whether anyone would know if it didn’t.
Can a name called across a room be heard?
A called name has to beat whatever noise the room is making, and that noise climbs through the day. In one United States emergency department, researchers recorded mean sound levels of 69.6 dB(A) in the afternoon against 65.9 dB(A) in the morning. The measurements came from 47 sampling sessions across a single site (The American Journal of Emergency Medicine, 2026). The authors call it a pilot study, and both figures sit well above a quiet office or a home living room.
Part of the reason a room like that gets louder is a feedback loop driven by the Lombard effect, the reflex that lifts voices in noise. A study of 142 people across 12 restaurants in Berlin proposes that reverberation feeds this through sound strength, pushing patrons to speak up to stay intelligible. That, in turn, pushes the background noise higher still (Forum Acusticum 2023, 2023). It’s the authors’ own modelled interpretation, not a measured cause, but it explains why a counter’s noise floor never settles.
The clearest evidence of what that means for a service desk comes from a 1,600-bed hospital in North India. Of the 30 sites measured, 23 were indoors, and their speech intelligibility ranged from 0.25 at outpatient registration to 0.95 elsewhere in the building. The study itself set 0.75 as the threshold for good speech intelligibility (Indian Journal of Public Health, 2026). Sixteen of the twenty-three came back poor or only fair. The worst-measured indoor location in the hospital was the registration desk, precisely where names are called forward.
Workplace noise law covers the staff doing the calling, not the customers waiting to be called. The Health and Safety Executive sets a lower exposure action value of 80 dB(A) as a daily or weekly average. Its regulations explicitly do not apply to “members of the public exposed to noise from their non-work activities” (Health and Safety Executive, 2025). A five-minute snapshot can’t be compared to a daily average anyway. Neither number tells you whether a customer at the desk heard their name.
Who does not hear it, and how would you know?
Hearing difficulty is common enough that a called name can’t assume it will land. Around 1 in 3 people in the UK are deaf, have hearing loss or tinnitus: more than 18 million people in total (RNID, 2026). That headline count covers three different things, and not everyone in it would miss a name.
RNID builds its hearing-loss figures from average hearing levels of 25dB or worse in one or both ears, for adults. It combines that with population data from the Office for National Statistics and the MRC National Study of Hearing. That threshold describes hearing sensitivity, not how well someone picks a name out of a busy room. The wider one-in-three count also takes in tinnitus, which that threshold doesn’t define.
The hospitality experience compounds this. One 2024 survey of deaf and hard-of-hearing people asked about their treatment by hospitality staff. It found that 44% of people with hearing loss had experienced negative attitudes and behaviours from staff in the last 12 months. The figures were 57% for deaf people and 77% for BSL users (RNID, 2026). That figure measures how staff treated people, not how many missed an announcement. It shows a counter is hard for many of these customers before a name is called.
There’s also a mechanism that makes a missed name easy to miss twice over. In a laboratory study of Cantonese-speaking students, hearing your own name stopped capturing attention once it was spoken quickly. At the shortest compression tested, only 4 of 23 participants noticed it had been said at all (Cognitive Research: Principles and Implications, 2024).
An earlier study, published in the Journal of Neuroscience in 2023, looked at attention in a different way. Hearing your own name in an unattended stream of speech did not make listeners turn towards the speaker. The authors suggest that detection may stay below conscious awareness. Either way, a name can register without producing the turn of the head a member of staff is watching for.
That’s the operational problem in one sentence. The customer who missed it doesn’t know a message was sent, so nobody complains, and the failure never reaches the person running the counter.
Does a digital alert get looked at?
The number behind this question depends entirely on what’s being counted, so the definition has to come first. One study defines Interaction Delay as the gap between a notification appearing and a person swiping it away or opening it (Sensors, 2024). A swipe counts exactly the same as an open. That matters, because dismissing a message isn’t the same as reading it.
Across nearly ten million notifications from 922 Android users, the median Interaction Delay was 0.35 minutes, about 21 seconds. The median for messaging apps specifically was slightly faster, at 0.33 minutes. The mean was far higher, 16.52 minutes, because a small number of notifications sit unopened for a long time and drag the average up.
That figure has a limit, though. Interaction Delay runs from the moment a notification appears to the moment it leaves the bar, so one nobody ever touches produces no figure at all. The study therefore can’t say what share of alerts go unopened. It describes only the ones somebody eventually dealt with.
Age changes the picture too, in the smaller subset of 537 users who gave their age. The median rose to 0.55 minutes for users aged 45 to 62, and to 0.83 minutes for users over 62. The mean for that older group climbed to 31.43 minutes, though it is much the smallest age band in the data. For a venue relying on a phone alert, older customers are the ones most likely still to be waiting when the food or the car is ready.
A slow response might mean someone didn’t see the message, or that they saw it and chose to deal with something else first. The authors are candid about that limit. They write that they “cannot differentiate whether a high IDL is caused by technical issues or the app user’s individual prioritization of apps without the collection of more data”.
None of this says what the alert should say, only whether it gets looked at. Getting the wording of an update right covers what to put in the message itself.
What has to be working for the alert to arrive?
Coverage figures make a phone alert sound more dependable than it might be at the counter, simply because of where they’re measured. Ofcom reports: “All MNOs maintain coverage of at least 99% outside of UK premises” (Ofcom, 2026). That figure is a prediction supplied by the operators’ own models, not a measurement taken inside any particular building. A counter at the back of a building is not the same place as a pavement outside it. The national figure says nothing about which one a customer is standing in.
Even a phone with good signal doesn’t always alert straight away. Android’s own documentation describes a low-power Doze mode. It triggers when a device is left “unplugged and stationary for a period of time, with the screen off” (Android Developers). Once triggered, normal-priority messages wait: they’re “delivered during the periodic Doze maintenance windows or as soon as the user wakes the device.”
A phone a customer keeps picking up and checking rarely meets all three conditions at once. Android also lets a sender mark a message high priority, which wakes the app straight through Doze. So this mainly bites an ordinary message sent to a phone left alone in a bag or a pocket.
One app illustrates this precondition from the sender’s side. Upd8All is strictly one-way, so a customer who misses an alert still has to come back and ask, and each message is capped at 40 characters. It also depends on the same thing this section has been about: a phone that’s charged, connected, has a working camera and is being checked.
No phone numbers or email addresses are exchanged between the business and the customer. Pairing happens by scanning a QR code. That single step replaces any sign-up process. There are no accounts and no sign-up, for either side. Nothing further needs installing. The customer does not need the app: scanning the QR code opens the updates in their browser through Upd8 Go.
That’s what a one-way channel can carry: a short alert, not a conversation. A venue weighing whether to collect numbers at all can read about the obligations that arrive with a customer’s mobile number first. What a coaster pager can and cannot reach runs into the same charged-battery constraint, on different hardware.
Where has this been tested side by side?
One facility in Nigeria ran a loudspeaker call and an on-screen ticket display side by side, three months after installing a queue-management system (Indian Journal of Radiology & Imaging, 2025). The sample covered roughly 200 radiographers, radiology assistants and patients. It’s the only setting found where both ran in the same room at the same time, rather than being studied separately. Both staff and patients rated the screen above the voice call.
| Statement rated (scale of 5) | Group | Mean |
|---|---|---|
| Speakers sound whenever a ticket is called | Radiology assistants (n = 50) | 3.96 |
| A corresponding display of called tickets appears on screen | Radiology assistants (n = 50) | 4.56 |
| Applicants are attentive to tickets being called or displayed | Radiology assistants (n = 50) | 2.32 |
| Staff give personal guidance so applicants know they’ve been called | Radiology assistants (n = 50) | 4.96 |
| Ease of hearing when tickets are called | Patients (n = 100) | 3.71 |
| Ease of seeing tickets displayed on screen | Patients (n = 100) | 4.05 |
The statements above are condensed from the paper’s own wording. It also reports these results as percentages, but each one is the five-point mean rescaled to 100, not a share of people who agreed. The means are the figures printed for each item.
Three things stand out. Both groups scored the screen higher than the loudspeaker call. The lowest-rated statement of the whole set was whether anyone in the waiting area was paying attention to either method, at 2.32 out of 5. The highest-rated statement in the table wasn’t about the technology at all. It was that staff usually go over and guide someone in person, at 4.96 out of 5.
This is opinion on a five-point scale, gathered in one facility three months after installation. It counted no missed calls and ran no control condition, so it can’t say how often either method failed. What it does show, in the one place both ran together, is that the screen scored higher than the speakers on every item that asked about them. Staff still gave personal guidance as routine.
What these studies do not settle
Every source here measures something real, in one place, at one time, using a method built for that setting. None of them measures the same thing as any of the others. Stretching one to answer a question it wasn’t designed for is exactly the mistake worth avoiding. Table 3 sets out where each one stops.
| Source | Setting | When | What it measured | What it did not measure |
|---|---|---|---|---|
| Indian Journal of Public Health | One 1,600-bed hospital, North India | Day and night, weekdays and weekends | Speech intelligibility at 30 sites | Whether any individual heard their name |
| The American Journal of Emergency Medicine | One United States emergency department | May to September 2025, 47 sampling sessions | Ambient noise levels, plus internal staff safety reports | Whether any patient heard a call |
| Cognitive Research: Principles and Implications | Laboratory, Lingnan University | Published 2024 | Attention to a compressed own name | A real service counter |
| Sensors | Android smartphone users | October 2018 to October 2020 | Time to interact with a notification | Whether a notification was ever opened at all |
| Indian Journal of Radiology & Imaging | One hospital, Nigeria | Three months after installation | Agreement with voice vs screen calls | Missed calls or no-shows |
| Ofcom | UK-wide | January 2026 data | Predicted coverage outside premises | Signal strength inside any specific building |
Repeated searches of the biomedical literature from the last three years, run in different framings, did not surface a trial that isolated the calling method itself. None measured missed calls or no-shows at a real counter. A further search, on number boards and displays specifically, returned results but none that measured whether people watching a board noticed their number. Both are statements about what the searches found, not claims that no such study exists anywhere.
That gap is exactly why the answer to calling names vs digital alerts has to come from preconditions and mechanisms, not from a single trial that settled it.
The failure you can live with
Every method here fails someone, and the decision comes down to which failure a venue can live with. A missed name and an ignored alert can look identical from behind the counter, but they aren’t the same problem, and they don’t share a fix.
One version means the message never arrived at all: out of earshot, or a phone with no signal or a dead battery. The other means it arrived and still didn’t land: spoken too fast to register, or a screen nobody happened to be watching. Chasing the wrong one wastes effort. Fixing a delivery gap means extending reach. Fixing an attention gap means making the alert harder to miss once it’s already there.
Neither one stopped staff from checking on people directly, which was the highest-rated statement in the one study that asked about all three. A venue doesn’t need a study to work out which group it’s more willing to miss. It needs an honest look at its own counter at the busiest hour of the week. The thing worth watching for is the moment a customer looks lost, rather than waiting for them to say something.
Frequently Asked Questions
Is a screen on its own enough?
It depends on where the venue puts it, not just whether one exists. Unlike a phone alert, a screen doesn’t rely on a customer’s own battery, signal or connection, since the venue owns that hardware. What it does need is a sightline: a customer must be able to see it from wherever they end up standing, which changes with each layout.
Does combining a called name with a number on a screen work better than either alone?
Combining them covers two separate failure points instead of reinforcing one. A called name fails when someone’s out of earshot or can’t pick it from room noise. A screen fails when someone can’t see it from where they’re standing. Pairing the two means a customer only has to clear one of those hurdles, not both, though neither helps if a phone is what’s uncharged or unwatched.
What can a venue do for a customer who can’t use either method?
The groups this article covers are the place to start. Older users were slowest to deal with a phone alert. RNID counts one in three people in the UK as deaf, living with hearing loss, or experiencing tinnitus (RNID, 2026). Some of them will not reliably catch a name called across a room. For a customer likely to fall into either group, checking in directly beats waiting on a response neither channel can confirm arrived.
Does a pager count as a digital alert?
Yes, in the sense used throughout this article. It’s a device that has to be charged, in range and seen or felt, rather than a call that has to be heard. Pagers face the same battery and range constraints described here, just built into different hardware. They don’t need a customer’s own phone or a signal from a mobile network, which is where they differ from a message sent to a personal device.
Does saying a name twice, or louder, fix the problem?
Not reliably, at least not by speed. One study tested how fast a name could be spoken before people stopped noticing it. Attention dropped away as the name was compressed (Cognitive Research: Principles and Implications, 2024). It never tested volume, so calling louder is untested. Calling again can help if the first attempt was too fast to register, but it won’t reach a customer who has stepped out of earshot.
Sources
- Indian Journal of Public Health, Speech Transmissibility in a Tertiary Care Hospital in North India: An Unexplored Consequence of Hospital Noise, retrieved 2026-08-29, https://doi.org/10.4103/ijph.ijph_1070_23
- The American Journal of Emergency Medicine, Acoustic burden in the emergency department: A latent operational Hazard, retrieved 2026-08-29, https://doi.org/10.1016/j.ajem.2026.06.050
- Forum Acusticum 2023, Noise for dinner – Towards a comprehensive statistical model of the acoustical and overall quality of restaurants, retrieved 2026-08-29, https://doi.org/10.61782/fa.2023.0483
- Cognitive Research: Principles and Implications, Attention capture by own name decreases with speech compression, retrieved 2026-08-29, https://doi.org/10.1186/s41235-024-00555-9
- RNID, Prevalence of deafness and hearing loss, retrieved 2026-08-29, https://rnid.org.uk/what-we-do/data-and-evidence/deafness-and-hearing-loss-prevalence/
- RNID, Facts and statements on inclusion, retrieved 2026-08-29, https://rnid.org.uk/what-we-do/data-and-evidence/inclusion-facts/
- Ofcom, Connected Nations update: Spring 2026, retrieved 2026-08-29, https://www.ofcom.org.uk/phones-and-broadband/coverage-and-speeds/connected-nations-update-spring-2026
- Sensors, Call to Action: Investigating Interaction Delay in Smartphone Notifications, retrieved 2026-08-29, https://doi.org/10.3390/s24082612
- Indian Journal of Radiology & Imaging, Evaluation of Queue Management System (QMS) Use in Chest X-Ray for Tuberculosis Screening: A Case Study, retrieved 2026-08-29, https://doi.org/10.1055/s-0045-1805004
- Health and Safety Executive, Noise: Employers’ responsibilities, legal duties, retrieved 2026-08-29, https://www.hse.gov.uk/noise/employers.htm
- Journal of Neuroscience, Neurophysiological Evidence for Semantic Processing of Irrelevant Speech and Own-Name Detection in a Virtual Café, retrieved 2026-08-30, https://doi.org/10.1523/JNEUROSCI.1731-22.2023
- Android Developers, Optimize for Doze and App Standby, retrieved 2026-08-29, https://developer.android.com/training/monitoring-device-state/doze-standby