How to Write a Status Update Customers Read

A status update earns its read in the first line, or it does not get read at all. Often the first line is the only one a reader sees before deciding whether to open the rest. Every rule that follows exists because of that one mechanical fact, not because good manners have gone out of style.

The message that gets read puts the change and the action first: what happened, and what the customer should do next. A greeting and an apology usually cost more characters than they earn back, so the strongest updates drop them entirely. Where a time is needed, a range beats a single number, and where there is no news yet, saying so plainly works better than staying silent.

None of this is a matter of taste. To write a status update customers read, follow rules the NHS, the rail industry and the National Cyber Security Centre already publish.

Key Takeaways

• The NHS tells the people who write to patients to start with the most important information: its own app messages can run to 5,000 characters, but the NHS says research shows some patients only read the first part. Android notifications are truncated to one line by default. (NHS digital service manual, August 2026; Android Developers, August 2026)
• A range beats a single number. Across eight preregistered experiments with 5,323 participants, people rated an app more positively when it gave a range instead of a point estimate. The range even beat a cautious estimate set at the later end of it. (Management Science, 2025)
• The National Cyber Security Centre’s advice to businesses is to avoid links where possible and never use a URL shortener. The NHS and GOV.UK Notify give the same warning. (NCSC, updated October 2023)

The first line is the whole message

Front-loading is not a style preference; it is the NHS’s own instruction to anyone writing a message to a patient. Its digital service manual tells writers to start with the most important information: most of a message may never get read.

NHS App messages can run to 5,000 characters including spaces. The NHS itself says research shows some patients only read the first part of the message. That is the NHS’s own claim about its own channel, not a general finding about text messages, and the guidance does not name the underlying study.

Android builds the same assumption into the software. By default, a notification’s text content is truncated to fit one line (Android Developers, August 2026). That is a documented default for Android phones, not a measurement of every handset on the market. Nothing here claims to know how an iPhone’s lock screen behaves.

None of this depends on which channel carries the message. A push notification, a text message and an app alert all get the same glance at whatever shows first. Then comes a decision about whether the rest is worth opening. That is a different discipline from writing an ordinary sentence.

What changed and what the customer needs to do about it belongs in those opening words. The methods for telling customers their order is ready cover which channel to use.

A greeting is the most expensive word in a message

The NHS’s own worked example of a patient text has no greeting, no apology and no sign-off. It is name, fact, action, link: every word is doing something.

((firstName)) ((lastName)), you are due for your NHS blood test. Book your appointment or find out more at http://www.nhs.uk/service-name-invite

Source: NHS digital service manual.

The NHS puts the length cost one way. Its own words: ‘if a text message is longer than 160 characters including spaces, it will be charged as more than 1 message’ (NHS digital service manual). GOV.UK Notify states the same threshold differently: ‘if a text message is longer than 160 characters (including spaces), it counts as more than one message’ (GOV.UK Notify). What that costs is answered in what SMS customer notifications cost.

Message length	Charged as
Up to 160 characters	1 text message
Up to 306 characters	2 text messages
Up to 459 characters	3 text messages

Source: GOV.UK Notify, retrieved 2026-08-29.

The characters that count twice

Certain symbols cost double. GOV.UK Notify lists brackets, curly braces and the pipe character among signs that count as 2 characters each. Its pricing table for accented characters shows the allowance dropping to 70. Twilio’s documentation puts the wider case. Including a UCS-2 character, such as an emoji, switches the encoding, and that ‘reduces the character length limit of the message to 70 characters’.

A common trap catches people who never touch a symbol on purpose. A word processor often swaps a straight quotation mark for a curly one while someone types. That single character is enough to trigger the same drop.

The words that carry nothing

Greeting, apology, brand name, a marketing line: none of it changes what the customer needs to know or do next. All of it spends characters that a name, a fact and an action could use instead. The NHS makes the greeting case itself, noting that a text does not always need a formal one.

The sign-off is the exception that same guidance draws, advising a recognisable one so the reader can tell who sent the message. It also says to avoid URL shortening services, because patients do not trust them, and to avoid attachments. Cutting the rest is not rudeness; on a message capped at 160 characters, it is the only way the fact and the action both fit.

Should you give a time, or a range?

Give a range, not a single number. Across eight preregistered experiments totalling 5,323 participants, people rated an app more positively when it gave a time as a range rather than a single point estimate. That held across different situations and time lengths (Hu, Gaertig and Dietvorst, Management Science, 2025).

The experiments tested the wording an operator would choose between:

• Point estimate: ‘Your food will arrive in 45 minutes.’
• Range: ‘Your food will arrive in 40–50 minutes.’

The obvious objection is that a business could just quote the pessimistic end of the range instead, and call it a single number. The researchers tested that too, and it does not hold up. Participants also rated ranges more positively than a conservative point estimate corresponding to the later, safer end of the same range. A range does not just protect against being wrong; people preferred it even against a number that was already the cautious choice.

Scope this carefully. The 5,323 participants took part across eight online experiments, and the abstract does not state their country. What was measured was how positively people rated an app after seeing its past estimates, not whether anyone actually waited. Whether an estimate helps at all in the first place is worked out in whether telling people how long they will wait helps.

A second study, on food delivery, adds what happens once a time has been quoted and missed. Its authors found that arriving later than promised and earlier than promised both change how soon a customer orders again. A late delivery costs a business more than an early one gains it. This is drawn from transaction data at a Western European delivery service and a matching lab experiment (Harter, Stich and Spann, Journal of Service Research, 2024).

No effect size from that paper appears here; its abstract states the direction only, not a number. What happens once a business has quoted its own time and missed it is worked out in what happens when a business misses the time it quoted.

Rail companies put ‘we don’t know’ in writing

The rail industry has gone further than most and put a rule about admitting uncertainty into writing. The Rail Delivery Group’s customer information pledges commit operators to four qualities when they explain a problem to passengers. One says outright that not knowing something is acceptable, so long as the operator says so.

The four-part test, printed on the Rail Delivery Group’s own site, reads:

• Clear: easy to understand and jargon-free
• Consistent: you will get the same information whatever channel you use
• Correct: ‘we will give you factual information. If we don’t know something, we will tell you that we don’t know it.’
• Concise: ‘we will keep things simple’

Source: Rail Delivery Group, Customer Information Pledges, dated September 2025.

The Correct pledge is the one worth building a message around. An operator committing, in writing, to telling customers when it does not know something has already solved the hardest part of a delay message. Staying silent until there is good news leaves the customer checking a phone that never rings.

Two further pledges are bullets running on from their own introductory lines, so they are quoted from the verb onward:

• At a station: ‘…keep announcements to a minimum, so you can easily pick out important information’
• As soon as a problem is known: ‘…give our best estimate for how long the disruption will last’

What to send when there is no news

A holding message earns its place by doing what the Correct pledge above describes. It says plainly that the outcome is not known yet, rather than staying silent until it is, and that still tells the customer someone is watching. Keeping garage customers updated on a repair works through the same principle against a specific set of moments in one trade.

A link is the riskiest thing you can put in an update

The National Cyber Security Centre’s advice to businesses is blunt: don’t include links, if possible, and don’t ask for personal details. That runs against instinct for a business wanting to send a tracking link, but the reasoning is about what a link teaches a customer to do.

It also tells businesses not to use URL shortening services and to avoid language that implies urgency, calling those ‘tactics often used by phishing attacks’. Businesses should also let customers check a message independently.

GOV.UK Notify agrees, and explains why: ‘your users cannot see where the link will take them’. Three UK public bodies, the NHS, the NCSC and GOV.UK Notify, independently warn against link shorteners.

There is a reason customers are primed to distrust an unfamiliar link. Four in ten (40%) UK mobile users reported having received at least one suspicious message in the past three months. That comes from a 2026 Ofcom survey of a representative UK sample aged 16 and over. An unrecognised sender’s link competes directly with that expectation.

None of this means a business should never collect a number or send a link. What a service message may contain, and what turns it into something closer to marketing, is covered in what happens when you collect customer phone numbers.

One way to sidestep the question is not to send a link, or collect a number, at all. Upd8All has genuine limits worth naming. Messages are capped at 40 characters, channels run strictly one way, and the customer needs a smartphone with a working camera and connection. The scan that sets the channel up happens at the customer’s end.

That setup step changes what the update itself has to carry. The channel already exists before the first message goes out, so the update needs no link and asks for nothing back. No phone numbers or email addresses are exchanged between the business and the customer. Pairing happens by scanning a QR code.

This site is published by Upd8All Limited, the maker of the app.

How Upd8All’s QR pairing works covers the mechanism in full.

The most-quoted texting statistic has an author

Two figures about how often text messages get read circulate constantly, most often printed with no attribution at all, and they do have one. A digital marketing firm called Mobilesquared put both numbers into print, in its own words, on the same page:

• ‘Back in 2010 we released a white paper called “Conversational Advertising” in which our research revealed that 98% of all SMS messages were read, and that 90% of all SMS messages were read within 3 minutes.’ (Mobilesquared, retrieved 2026-08-29)
• ‘Our latest data suggests 55% of all SMSes are read, however, 100% of SMSes are viewed.’ (Mobilesquared, retrieved 2026-08-29)

Neither figure carries a sample size, a country or a stated method, on the page that prints them today. The firm does not stand behind either one. It says that ‘neither datapoints represent an accurate reflection of today’s world of mobile messaging’ (Mobilesquared).

That matters more than the number does. Whatever a business decides about how often its own updates get read, a figure its own publisher has withdrawn should not be deciding it.

Shorter than polite

A message that gets read is often ruder than the sender’s instinct wants it to be. That trade is what every rule above is asking for. Every character spent on a greeting, an apology or a company name is a character not spent on what the customer needs. What happened, and what to do next.

None of the organisations behind this article’s evidence, the NHS, the rail industry, the National Cyber Security Centre or GOV.UK Notify, write for warmth first. They write for a reader who may see one line and nothing else. Politeness that costs the customer their own time is not politeness. It is a habit borrowed from a letter, applied to a medium that punishes it.

The rail industry’s written commitment to say plainly when it does not know something is the clearest example of the same principle applied to bad news. Admitting uncertainty is shorter, and more useful to the reader, than staying silent until there is something better to say.

Frequently Asked Questions

How long should a message to a customer be?

There is no single right length, but a text over 160 characters is charged as more than one message. That is true on both the NHS’s guidance and GOV.UK Notify’s (NHS digital service manual, August 2026; GOV.UK Notify, 2026). On a phone screen, the first line matters more, since Android notifications are truncated to one line by default (Android Developers, August 2026).

Should you apologise when an order is late?

An apology is rarely wrong, but it is rarely the priority. The NHS’s worked example of a patient text has no apology at all. It has only a name, a fact and an action (NHS digital service manual, August 2026). On a message with limited space, an apology usually spends characters the fact and the action need more.

Is it better to give a customer a time or a time range?

A range. Across eight preregistered experiments with 5,323 participants, an app that gave a range was rated more positively than one giving a single number (Management Science, 2025). The range still won even against a cautious point estimate set at the safer end of it. The study measured how people rated the estimator, not whether anyone actually waited.

What should you say when you do not know when it will be ready?

Say so, plainly, rather than staying silent. The Rail Delivery Group commits its members in writing to telling customers when it does not know something yet. It also commits to keeping messages simple (Rail Delivery Group, 2025). A short message confirming there is nothing new yet still tells a customer someone is watching.

Should you put a link in a text message to a customer?

Avoid it where possible. The National Cyber Security Centre tells businesses not to include links if they can help it, and not to use URL shortening services (NCSC, updated October 2023). GOV.UK Notify and the NHS give customers the same warning about shorteners, because an unfamiliar or shortened link is exactly what a scam message looks like.

Do customers actually read text updates?

Most probably do, but the figure most often quoted does not hold up. The 98% read-rate figure traces to one marketing firm’s 2010 white paper. That same firm’s current page prints a lower 55%, says neither number reflects mobile messaging today, and gives no sample size or method for either (Mobilesquared). Treat any specific percentage with caution, and write the message to earn its read regardless.

Sources

• NHS England, “Writing NHS messages,” NHS digital service manual, retrieved 2026-08-29, https://service-manual.nhs.uk/content/writing-nhs-messages
• Government Digital Service, “Text message pricing,” GOV.UK Notify, retrieved 2026-08-29, https://www.notifications.service.gov.uk/pricing/text-messages
• Government Digital Service, “Links and URLs,” GOV.UK Notify guidance, retrieved 2026-08-29, https://www.notifications.service.gov.uk/using-notify/links-and-urls
• Android Developers, “Create a notification,” retrieved 2026-08-29, https://developer.android.com/develop/ui/views/notifications/build-notification
• Rail Delivery Group, “Customer Information Pledges,” retrieved 2026-08-29, https://www.raildeliverygroup.com/our-services/putting-customers-first/customer-information/all-the-pledges.html
• Hu B, Gaertig C, Dietvorst BJ, “How Should Time Estimates Be Structured to Increase Customer Satisfaction?” Management Science, retrieved 2026-08-29, https://pubsonline.informs.org/doi/10.1287/mnsc.2023.00137
• Harter A, Stich L, Spann M, “The Effect of Delivery Time on Repurchase Behavior in Quick Commerce,” Journal of Service Research, published online 4 March 2024, retrieved 2026-08-29, https://journals.sagepub.com/doi/10.1177/10946705241236961
• National Cyber Security Centre, “Business communications – SMS and telephone best practice,” last updated 31 October 2023, retrieved 2026-08-29, https://www.ncsc.gov.uk/guidance/business-communications-sms-and-telephone-best-practice
• Ofcom, “New rules to thwart text message scammers and protect consumers and businesses,” published 15 July 2026, retrieved 2026-08-29, https://www.ofcom.org.uk/phones-and-broadband/scam-calls-and-messages/new-rules-to-thwart-text-message-scammers-and-protect-consumers-and-businesses
• Mobilesquared, “SMS engagement dropping, still outperforms other channels,” retrieved 2026-08-29, https://www.mobilesquared.co.uk/2022/08/24/sms-engagement-dropping-still-outperforms-other-channels/
• Twilio, “What is the SMS character limit?” last modified 9 December 2025, retrieved 2026-08-29, https://www.twilio.com/docs/glossary/what-sms-character-limit