The difference between writing for email and the web

The most useful difference between these two kinds of writing is not length, formatting or tone. It is how the reader got there. Somebody reading a web page navigated to it, usually from a search or a link they chose to follow, which means they arrived with an intention and a topic in mind. Somebody reading an email did not ask for it at that moment. It appeared in a queue alongside other things competing for the same attention, and the first decision they make is not whether to read it but whether it belongs in the queue at all.

Everything else follows from that. An email has to earn its place before it earns a reading, so its opening does triage work rather than persuasive work. A web page has already been chosen, so its opening can assume a topic and get straight into substance. You will see confident figures quoted for how many seconds each gets — a few for an inbox, rather more for a page — and those numbers are not something a small business can verify or should repeat. The mechanism is checkable without them: a triage decision is faster than a reading decision, because it is a smaller decision.

A subject line competes against a list of other subject lines from senders the reader already has opinions about, and the question it answers is whether this is worth opening now. That makes specificity and relevance more valuable than intrigue, and it makes the sender name part of the copy — the same subject line performs differently depending on whether the reader recognises who sent it. Front-load the useful words, because the visible length is short and shorter on a phone.

A headline on a web page is doing something different: confirming to somebody who has just arrived that this page is about the thing they were looking for. It is also frequently the text that appeared in the search result they clicked, which means it has a double audience — the person deciding whether to click, and the same person a second later checking they were not misled. That is why headline advice emphasises matching the words people actually search with, while subject line advice emphasises the promise of value. The two can be similar sentences, but optimising one for the other's job is a common and avoidable mistake.

An email should ask for one thing. This is close to an absolute rule, and the reason is that an email is a single act of communication with a single reason for existing — if it contains three requests, the reader has to prioritise them, and the most common resolution of that small burden is to do none and leave the message unread-in-spirit. If you have three things to ask, that is usually three emails, sent at different times, each of which can be answered in a moment.

A web page can legitimately layer several actions, because the reader is in a different posture: browsing, comparing, deciding, and able to move around the page at their own pace. A primary action plus a lesser alternative — buy now with an option to ask a question first — serves readers at different stages of the same decision. The constraint is that the actions must be visibly unequal. Two calls to action of identical weight on a page produce the same paralysis as three requests in an email. The difference between the media is not whether choices can be offered; it is that a page has room to rank them and an email does not.

Email formatting has to survive an environment you do not control. Mail clients render differently, images are frequently blocked by default, and many people read in a narrow column on a phone. The practical consequences are conservative: short paragraphs of a line or two, no dependence on images to carry meaning, no layout that breaks if a column collapses, and a link that reads as a link in plain text. A beautifully designed email that arrives with images suppressed and reads as three disconnected fragments has failed for reasons that had nothing to do with its writing.

A web page has more reliable rendering and a different reading pattern. Readers scan for structure before committing, so headings, short lists and visually distinct sections do real work — they let somebody find the part that concerns them without reading the parts that do not. Headings on a page also have a second function that email headings do not: they describe the page's structure to search engines and to screen readers, which is a reason to write them as accurate descriptions of what follows rather than as teasers. "Pricing" is a better heading than "What will it cost you?" for exactly that reason.

An email is addressed to a person, arrives in a personal space, and reads badly in institutional register. Second person suits it — you, your — and so does the sender being an identifiable human rather than a department. A page addresses an unknown reader at an unknown stage, and a slightly steadier, less insistent voice suits it better, because relentless second person on a long page starts to feel like being talked at rather than talked to.

The distinction is easy to overstate, so it is worth being precise about the mechanism: an email knows who it is talking to and a page does not. That is why an email can reference the reader's specific situation and a page has to describe a situation and let the reader recognise themselves in it. The most common error runs one way — institutional language in email, where "we would like to inform our valued customers" is addressed to a category rather than to the person reading it. The opposite error exists too, and it looks like a web page written as though it were a personal letter to somebody it has never met, which reads as presumptuous rather than warm.

Email gives more feedback than a page, and it is worth knowing exactly what that feedback means. Whether a message was delivered is a fact. Whether a link was clicked is close to a fact, and a genuinely useful one because clicking is a deliberate act. Whether an email was opened is not a fact — open tracking depends on loading a tracking image, which many clients block by default and some privacy features load automatically, so the figure can be understated and overstated at once. Making decisions from open rates alone means making decisions from a number whose error runs in both directions.

On the web the reliable signals are also narrower than dashboards imply. That a page was visited is solid; that somebody spent a certain time on it usually is not, because the measurement is often inferred from a subsequent action that may never happen. In both cases the trustworthy signals are the ones tied to a deliberate action — a click, a form completed, a reply, a purchase. Those are what a small business should judge copy by, and the softer engagement figures are best read as weak, directional hints rather than as measurements.