Keeping a client informed and keeping a client written to are two different jobs. Past a certain rate, the second one crowds out the first. Willpower rarely fixes it. What fixes it is deciding three things: what has to arrive today, what waits for a scheduled slot, and what was only ever yours.
- Over-emailing trains your clients more than it irritates them. Once your mail reads as background, your questions stop working first, ahead of your updates.
- Count arrivals rather than messages. Three things in one email cost your reader one interruption. The same three spread across a Tuesday cost three. Much of what looks like over-emailing is a batching problem.
- The cadence belongs to your client’s decision rhythm. You want to write when you finish something. They need to hear when something changes for them. That’s a different clock, and usually a slower one.
- Sending a great deal is one of the more reliable ways of avoiding the thing you owe them. A full inbox and a client who’d struggle to say where the work stands is the ordinary combination.
What the volume actually costs you
Nobody tells you. A client who finds your correspondence heavy keeps that to themselves, because saying somebody is too communicative is awkward and sounds ungrateful. What happens instead is a quiet reclassification. Your messages become a category that gets dealt with, once a day at best, in a pass whose purpose is to reach the bottom.
Four costs follow from that, roughly in the order they arrive.
Your requests are the first casualty. A question sitting inside a stream is a question that can wait until the stream has been handled, and the stream stays unhandled. This degrades faster than anything else in client email, and it costs you the most, because a request is the one kind of message with your own week hanging off it. Then you chase. The chase is another arrival, and that arrival makes the next request slightly less likely to be answered. An unanswered thread looks exactly like a finished one, which is what makes this loop so hard to see from inside.
Everything you send becomes the same size. Twelve messages a week leaves your reader unable to tell which of the twelve mattered. The one that mattered arrived dressed identically to the other eleven. What you lose is the ability to make one message land, and you’ll want that on a specific Thursday you can’t predict yet.
You get handed downward. A senior contact spending twenty minutes a day on your correspondence will eventually, and quite reasonably, route it to somebody more junior. Inside the client that gets described as giving you a proper day to day contact. It’s a volume decision, and writing better emails afterwards leaves it where it is.
It reads as an absence of judgment rather than as diligence. Part of what a client buys is a view about what deserves their attention. Passing everything along undivided says you haven’t formed one. This is the cost people find hardest to accept, because the sending felt conscientious at the time and every single message was fine.
One qualification matters more than it looks. The budget is per person rather than per engagement, and it runs wildly different at different heights. Four messages a day to a counterpart you work alongside can be exactly right. One a week to their director may already be at the limit. So am I sending too much waits on too much to whom, and the routing half of that question is taken separately in this collection.
Informed is a state rather than a rate
A client is informed when they could answer three questions without asking you.
Where does the work stand, and has that changed since the last time I thought about it. What do you need from me, and by when. Is anything about to happen that I’d rather hear about beforehand than afterwards.
That’s the whole of it. Anything that moves none of those three may still be worth sending, and courtesy is a real reason to write. Just count it separately, because it spends from the same account as the messages that carry the work.
Two things follow, and they’re the useful part.
The first is that being informed is a condition you maintain rather than a quantity you emit. So the question on a Wednesday afternoon is whether your contact, stopped in a corridor, would get those three answers right. If they would, you owe them nothing today, and the email you were about to send is for you rather than for them.
The second is that the condition decays, and the decay rate is what sets your cadence. Through a stable stretch of delivery, what a client knew on Friday is still true the following Thursday. An email in between has nothing to add. In the week of a cutover it’s stale by lunchtime. Most engagements run at a rhythm chosen in week one and left alone after that, which guarantees it’s wrong at both ends.
Email is only one instrument for holding that condition, and it has the worst interruption profile of the lot. A ten minute call holds it. A page somebody can open when the question occurs to them holds it. That observation does more work later in this guide than it looks like here.
Why the extra messages get sent
Four mechanisms account for nearly all of it, and every one of them makes sense from the inside.
To stop carrying it. By volume this is the largest by a distance. A thought arrives at four in the afternoon, and sending it is how you put it down. The instinct is sound. The destination is the problem, because somebody else’s inbox is doing the work of your memory. It’s also the most fixable of the four, and the section below is about that.
To be visible. From the inside, a week in which the client hears nothing from you feels like a week in which they start wondering what they’re paying for. So you write, and the writing does reassurance rather than information. The trouble is the direction it travels. Clients form their view of value from things landing on the dates they were promised, and heavy correspondence from a supplier sits outside that.
To put it on the record. Legitimate, and in small doses correct. It turns corrosive when it becomes the reason a message exists rather than a side effect of one. A run of confirming notes has a texture, and clients read that texture accurately, which is that a file is being built. If you’re far enough into an engagement to be doing this on purpose, the volume is a symptom of something an email can’t settle.
Reporting at your own team’s pace. Internal work runs on a much faster loop than a client relationship does. Somebody used to a channel where five messages an hour is normal will hand a client roughly four times the traffic they can absorb before either party notices. It’s at its worst in the first month, when everything is new and everything feels worth mentioning.
Three of those four are solved by changing where the message goes. That’s a relief, because resolving to write fewer of them has a poor record.
Count arrivals rather than messages
Two emails at 9:10 and 9:45 cost your reader twice what one email containing both would have cost. The information is identical. What differs is the number of times a person has to stop, look, decide whether it needs them, and put it down again. That’s the thing actually in short supply.
So the number worth knowing is how many separate occasions you land on, rather than how many messages you send in a week. Counted that way, a fair share of what looks like a communication style turns out to be a batching failure, and batching failures are much easier to fix than a style.
The audit takes about fifteen minutes and is worth doing once, properly, on whichever client you’re least sure about.
Open your sent mail, pick that client, and take the last three weeks. Count arrivals rather than messages, so two on the same morning count as two, because that’s how they landed. Then sort each one into three piles: it needed them to do something, it changed their picture of the work, or neither.
The third pile is your answer. In most practices it runs between a third and a half of everything sent. Every item in it was written carefully and sent in good faith. All of it was optional, and all of it was charged to the same account as the messages that carried the work.
Then make one more pass over the first pile and count two numbers. How many separate things you asked for, and how many you got. That ratio is the most honest measure of your standing in a client’s inbox that exists, and you can produce it on your own. Above three quarters, your rate is fine whatever it is. Under a half, the rate is a problem however good the individual emails are.
One correction to the count, and it’s the one people miss. If a colleague is also on the engagement, the client’s experience is the sum of both of you, and that total is the number neither of you sees. Two people each sending what feels like hardly anything is the ordinary way to arrive at nine arrivals a week that nobody feels responsible for. Two fixes work and both are cheap. One of you writes to the client and the other writes to that person, or you divide by subject so each of you knows what the other sent this week.
The clock changes the felt volume more than the count does, too. Six messages between nine and six read as a busy week. The same six with two at eleven at night and one on a Sunday read as somebody who can’t stop, and they quietly set an expectation about your availability that you’ll be living with in month five. If you work late, write late and send in the morning.
Whose clock the cadence runs on
Left alone, cadence follows your work. You finish a piece, you report it. The trouble is that finishing is an event in your week and an ordinary Tuesday in theirs.
Their clock has three hands.
When they actually decide things. Every client organization has a rhythm. An operations meeting on Monday morning, a steering group every second Thursday, a leadership session on the first Tuesday of the month. An update landing the day after the meeting where your work gets discussed is a week less useful than the same update landing the day before it. This is the cheapest single improvement available in the whole subject, and it costs one question at kickoff that almost nobody asks.
How fast their picture goes stale. Steady delivery decays slowly. A migration weekend decays hourly. Cadence should follow that curve, which means it’s supposed to change two or three times across a long engagement. Fixed at kickoff and left there, a weekly note manages to be too much through the quiet middle and nowhere near enough in the week that decides everything.
How exposed the person is. A sponsor who put their name on this in front of their own board needs to stay ahead of it. That rarely means more email. The right answer is usually a short call rather than a longer note, because what they need is a chance to ask the question they’d keep out of writing.
With those three read off, the shape is usually obvious.
For most engagements, one scheduled written update a week, plus the one or two things in a week that genuinely can’t wait, is right, and it’s a lower total than most consultants are running. Stretch the scheduled note past two weeks and the client starts reconstructing your work from memory, which is when they begin asking where things stand. Push it beyond twice a week and it stops being an event and becomes traffic.
In a phase with daily consequences, invert it. Three or four lines each afternoon and no weekly at all, on the stated understanding that it ends when the phase does. Saying when it ends is what makes people willing to accept it in the first place.
On a slow advisory retainer, a monthly note and a call is honest work. The alternative is the padded weekly email that teaches a client your mail can be skipped.
Whatever interval you land on, the slot does more work than its contents. Something that arrives predictably lets a client stop wondering, and it gives every non-urgent thought in your head a destination, which is what makes the rest of this possible.
Somewhere for a thought to wait
Things get sent that didn’t need sending, and almost nobody decides to send them. They get sent because at four in the afternoon there was nowhere else to put them.
So make somewhere. Almost anything works. A draft message per client that you keep open and add to. A note file. A page in a notebook. An item in whatever list you already keep. There are only two requirements. One per client, and adding to it takes about ten seconds, because a holding place with any friction in it loses to the send button every time.
Then one question, at the moment your hand is on the keyboard. Does this have to travel today?
Yes, because waiting costs them something. Send it now, on its own, and keep it short.
No, but it belongs in the next scheduled note. Into the holding place. This is most of it.
No, and on reflection it’s yours rather than theirs. This pile surprises people the first time they look. Working notes, a thought you haven’t finished having, something you wanted acknowledged rather than acted on. Keep all of it. Just keep it out of their inbox.
Consolidating what has been waiting has a craft of its own, and getting it wrong produces the worst of both arrangements.
Three asks is the ceiling for one message, and even three want to be on separate lines with a name and a date beside each. Past three, the answer rate collapses rather than degrading gently, and what comes back is a partial reply covering the first one or two. That leaves you chasing a subset, which spends another arrival to recover ground you had.
Order them so the thing blocking you is first, because partial replies come from the top.
And if you honestly have nine questions, that’s a call rather than an email. Twenty minutes with the list in front of you, plus the two lines you write afterwards recording what was decided.
Four things go straight out whatever the holding place says, because a rhythm with no exceptions is a rule you’ll break at the worst available moment.
Anything with a clock on their side. Their approval window, their board pack deadline, the sign-off that has to be in before Friday. Holding that spends their time rather than tidying yours.
Anything that gets more expensive while it waits. A number you’ve discovered to be wrong in a document they’re about to circulate is worth an interruption at any hour of the day.
Bad news. The same day, and by voice first when it’s real, with the written version following as the record. Why that order matters is set out in this collection’s piece on status updates, and it applies here word for word.
Anything you got wrong. The gap between your discovering a mistake and their hearing about it is the whole of what gets remembered afterwards, and it’s the one number in this guide worth driving to zero.
One hazard is quieter than those. A holding place works as long as the scheduled note actually goes out. Miss the slot and you’ve converted a week of small items into a week of silence with all of them still in your possession, which is the failure the next section is about.
The failure that costs more
Everything so far pushes in one direction, so here’s the other one plainly. Under-informing is rarer and considerably more expensive.
A gap reads as trouble rather than as calm. A client who has heard nothing for two weeks builds an explanation out of the last thing they heard, and that explanation is never the true one and rarely the generous one. When you do write, you’re arguing with a version of the work that got there first rather than describing the work.
Longer gaps become agenda items. Three weeks of nothing converts how is it going into we should talk about how this is being run, and that’s a different meeting with different people in the room.
Then there’s the disguise, which is the part worth watching in yourself. You can be in a client’s inbox nine times in a week and still leave the risk to the date unsaid. That combination is common rather than paradoxical, because sending is how discomfort gets discharged and the message you actually owe is the uncomfortable one. So on any week where you’ve written a great deal, the useful question is whether the hardest sentence you know is anywhere in it.
The quiet week deserves its own note. When there’s genuinely nothing, the temptation is to skip the slot. Skipping is the most expensive thing available to do with a slow week, because a slow week is exactly when a client starts wondering what they’re buying. Keep the appointment and shrink the message instead. What a thin update should contain, and why padding it out costs more than sending three lines, is the subject of another piece here.
The last one is invisible from your side and worth naming for that reason. A client can feel under-informed while being perfectly well informed, if what they needed went to somebody else or landed in a thread they have no way to see. That’s a routing failure rather than a volume one, and this collection takes routing separately. When the complaint is coming from somebody senior while your day to day contact seems content, it’s nearly always that.
Agreeing it at the start and correcting it later
All of this is far easier to set than to change, and setting it takes about four minutes at kickoff, where it reads as professionalism rather than as a negotiation about your workload.
Four questions, in plain words.
How often would you like to hear from me, and would you rather that was written or a call. When does your own team meet about this, so I can land before it rather than after. What would you want to hear about on the day it happened, however small. Who else needs this, and how much of it.
The third question is the one that earns its place. It converts a vague preference into a rule you can apply on a Tuesday afternoon without thinking about it, and the answers are more specific than people expect. One client cares only about anything that touches their own customers. Another only about money. Another about anything a named colleague might hear before they do.
Then say the answer back once, in a sentence, so it becomes an agreement rather than a habit. A note on Thursday afternoon ahead of your Monday meeting, a call if a date moves, Priya copied on the note but not on the working thread. Written down like that, it can be revised later in one line. Left as a habit, revising it turns into a conversation about whether something has gone wrong.
The signals that you’re over the line arrive sideways rather than in words. Replies get shorter and then stop. Some of your questions are answered and others in the same message sit there. Direct questions come back the same afternoon while anything asked inside an update waits a week. Somebody says, kindly, that they only need the headlines. Your contact starts replying to your second message rather than your first, which is a small and precise piece of information about how your mail is being processed.
Being under the line looks different. They ask where things stand. A question arrives from somebody you rarely hear from. You find out that a meeting about your work happened and you heard about it afterwards.
Changing the rate needs no ceremony, and ceremony is the mistake most people make here. Nobody wants to receive an email about your emailing policy. Change it, and if it wants a sentence at all, one clause in the next note does the job. From here I will put the smaller things in the Thursday note rather than sending them as they come up.
Two client instructions are worth translating rather than obeying.
Keep me posted, copy me on everything. This is a request for reassurance far more often than a request for mail. It says they want to stay ahead of surprises and doubt they’d hear in time. Answering it with volume makes it worse, because now they’re surprised inside a stream they’ve stopped reading. The right response is a question followed by a rule. What would count as a surprise, and then, I will send you anything that touches the date or the number directly, and everything else will be in the weekly. That’s met with relief far more often than people expect. A client who genuinely does want all of it, having been asked, is telling you something about trust, which is a real problem, and copying them leaves it where it is.
Only come to me if there is a problem. Take it, mostly, and pin one thing down first. Keep a short scheduled note anyway, even three lines. Otherwise you’ve built a relationship in which the first message they get from you in a month is bad news, and the sight of your name in their inbox becomes an event in itself. That’s a strange thing to construct on purpose and remarkably easy to construct by accident.
What email is the wrong tool for
A fair amount of over-emailing is email being asked to do jobs it’s bad at, rather than a discipline problem. The cure is to move the job somewhere better suited to it.
Current position. Email carries events. Holding state is beyond it, and every attempt to make a thread do that produces a thread nobody can read. One page you keep current, which the client can open whenever the question occurs to them, removes more traffic than any other single move in this guide, because it absorbs the entire category of message whose real content is where are we. The mail then carries only what changed. If the work is big enough that the page wants dependencies and a critical path in it, that’s project software, and buying some is the correct answer rather than an admission of anything.
A subject that has taken four messages in two days. That’s a conversation being held in the wrong medium, and each of the four cost an arrival. Ten minutes on a call, then two lines in the thread recording where it landed.
Arranging times. The most reliable source of arrivals that carry no information whatsoever. A link that shows your real availability removes most of it, and Point will run the back and forth itself if you’d rather it stayed out of your week altogether.
Work that lives in the client’s own system. If they run the project in their tool, a parallel email stream becomes a rival record to the official one, and the rival loses. Work in theirs, and keep email for what has to be durable and findable in a year.
Daily chatter with an embedded team. If you’re alongside their people every day, a channel is the right instrument and email at that pace is the wrong one. The scheduled written note still does the thing a channel can’t, which is leaving a durable position somebody can forward. Where both exist they at least want to arrive in one place rather than two, which is what mail and chat in a single feed is for.
And the honest one to end on. If you’re writing a great deal because you’re unsure the client believes in the work, cadence design leaves that where it is. That’s a conversation, and it’s usually a shorter one than you fear.
What Point changes about the volume
Every decision above is yours. What deserves an interruption, how much a particular client can carry, whether the awkward sentence goes in this week’s note or into a phone call today: that’s the work, and it stays yours whatever software you run. What software can change is the pressure that produces the extra messages, because most of them get sent for mechanical reasons rather than considered ones.
The holding place is the first and the biggest. A thought gets emailed at four in the afternoon because the alternative is carrying it until Thursday, and carrying things is what people are worst at. When a commitment made inside a message turns into a dated item instead of depending on your remembering it, the thought has somewhere to go besides the client’s inbox. Whatever you asked of the client sits in the same register, stamped with the day you asked for it. That register is also the working version of the audit earlier in this guide. A set of requests that produced nothing week after week is the measurable end of a client who has stopped reading you, and having an open loop come back before it goes cold keeps one chase from turning into three more arrivals.
Knowing what you already sent is the second. A useful share of over-emailing is repetition you can’t see, because checking what went out on Monday costs more than writing it again. In Point a conversation arrives with a plain account of itself attached. A week with one client becomes a couple of minutes of skimming rather than an evening, and a long chain can be picked up at the point it has reached rather than from the top. Writing from the record instead of from memory is most of what keeps this week’s note from repeating last week’s.
The return leg matters as much as the outgoing one. Sending less is half of a calm engagement. The other half is that a client’s reply gets seen the same day instead of sitting among ninety other things, and marking who actually matters is the cheap version of that. Answering a client’s own question quickly is the one kind of speed that costs you nothing from the budget this guide is about, because it was their arrival rather than yours.
Scheduling deserves a line of its own. It generates more messages than any other subject in an engagement, and not one of them carries information. Point runs that exchange itself, to the limit you set for it.
Drafting is available, with one caution specific to this subject. Point writes new messages as well as replies, and learns how you write, so a scrappy note about the week comes back as a draft you can cut down instead of a blank window. The thing to watch here is length. A model handed thin notes produces a full page, and a full page is precisely what this guide is trying to talk you out of, so cut the draft back to what changed.
Three limits belong in plain sight rather than in a footnote.
Point doesn’t know how much mail your client can take. Whether four messages a week to a particular operations lead reads as attentive or as oppressive sits outside what Point can see, and that judgment is the entire subject of this guide.
Point can’t tell you whether anything you sent was read, and nothing honest can. The behavioral signals set out above are what you actually have.
And a message that has arrived on somebody else’s server stays there, whatever any product suggests. Point keeps a timestamped record of what it did and most of what is in it can be reversed, though a delivered email sits outside that set. That’s an argument for the pause before sending rather than a way of recovering from having skipped it.
How much Point does without you is a separate setting for each type of action, and review is where all of them begin. Anything addressed to a paying client is a fair candidate for staying there permanently, however far you raise the rest, and what that setting actually controls is worth reading properly before you touch it.
Client correspondence is confidential by default, and deciding where it’s allowed to go is better done on purpose than by whichever window happened to be open. The questions worth putting to a vendor have a guide of their own.
All of this runs on the address you already have. Point sits on top of the Google or Microsoft mailbox you use now, and your client keeps writing to whichever address they’ve always used. Every capability has a line on benefits, the case as it applies to an advisory practice sits at Point for consultants, and the product lives at Point.
Common questions
How often should I email a client?
One scheduled written update a week, plus the one or two things in a week that genuinely can’t wait, fits most engagements, and it’s a lower total than most consultants are running. Set the interval from the client’s rhythm rather than your own. Land the update before the meeting where your work gets discussed, tighten it during a phase where the picture goes stale in a day, and loosen it through a stretch where nothing moves. Below one note every two weeks the client starts reconstructing your work from memory, and above two a week the update stops being an event and becomes traffic. Whatever you settle on, keep the slot even in a week with nothing in it, because a skipped week in a quiet stretch is exactly when somebody begins wondering what they’re paying for.
How do I know whether I am sending too much?
Count it rather than guessing at it. Take three weeks of sent mail for one client. Count arrivals rather than messages, so that two on the same morning count as two. Then sort them into things that needed the client to do something, things that changed their picture of the work, and neither. The third pile usually runs between a third and a half of everything, and all of it was optional. Then count the requests you made against the requests that came back. Over three quarters and your rate is fine, under a half and the rate is a problem whatever else is true. The behavioral signals agree with that number. Partial replies, a question answered when you ask it directly and ignored when it sits inside an update, and a contact who starts replying to your second message rather than your first.
Is one long email better than several short ones?
For the reader, almost always, because the cost they pay is per arrival rather than per word. Two messages forty minutes apart cost twice what one containing both would have cost, and the information is identical. The limit is requests rather than length. Three asks is the ceiling for a single message, each on its own line with a name and a date, and past three you get a partial reply covering the first two, which leaves you chasing a subset and spending another arrival to do it. Put whatever is blocking you at the top, since partial replies come from the top. If the long email is long because it has been padded, that’s a different problem and shortening it is the answer.
My client told me to copy them on everything. Should I?
Translate it before you obey it. Almost nobody who says this wants more mail. They’re saying they want to stay ahead of surprises and doubt they’d hear in time, and doing it literally makes that worse, because now they’re surprised inside a stream they’ve stopped opening. Ask what would count as a surprise, then offer a rule. Anything touching the date or the number comes to them directly, everything else is in the weekly note. That’s met with relief far more often than people expect. If, having been asked, they genuinely do want all of it, the message is about trust rather than about information, and copying them leaves that where it is.
What goes in the note during a week when nothing moved?
Send the note anyway and make it three lines. The rhythm carries reassurance independently of its contents, and the week you’re tempted to skip is a slow one, which is precisely when a client starts asking themselves what they’re buying. Say the position, say that it has held, give the next date, and stop. The thing to avoid is padding a thin week into a full one, because filler is the specific thing that teaches a reader your mail can be skipped, and this collection’s piece on updates that actually get read is about that at length.
Can I end up under-informing by trying not to over-email?
Yes, and it costs more than the thing you were avoiding. Silence reads as trouble rather than as calm. A client fills it with an explanation built from the last thing they heard, and you end up arguing with that rather than describing the work. The safeguard is simple enough to apply weekly. Cutting volume has to leave the three things that make somebody informed intact. Where the work stands, what you need from them, and anything about to happen that they would rather know in advance. If all three are current, quiet is fine and you should enjoy it. If one of them is stale you’re late rather than restrained, and having sent nine other messages this week does nothing for it.
The short version
Keeping a client informed and keeping a client written to are different jobs, and the second crowds out the first once the rate passes what they can carry. The cost is a rule your reader learns rather than irritation, and the first thing it breaks is your questions rather than your updates, because a request inside a stream can always wait until the stream has been dealt with. Count arrivals rather than messages. Three things in one email cost one interruption and the same three across a morning cost three, so much of what looks like over-emailing is a batching failure. Audit it once, honestly, with three weeks of sent mail sorted into needed an action, changed their picture, and neither, then compare the requests you made against the ones that came back. Run the cadence on their clock rather than yours, landing before the meeting where your work is discussed, tightening it in a phase that goes stale in a day and loosening it in a slow stretch. Give every non-urgent thought a holding place so the client’s inbox stops being your memory. Keep three asks as the ceiling for one message with the blocking one first, and let four things break the rhythm without hesitation. Anything with a clock on their side, anything that gets more expensive while it waits, bad news, and your own mistakes. Then guard the other edge, because a silence gets filled with an explanation you did not write, and a heavy week with the difficult sentence missing is the usual way to be over-emailing and under-informing at the same time. Agree the shape in four minutes at kickoff, ask what they would want to hear about on the day it happened, and correct it later by changing the rate rather than announcing a policy. What goes inside the update, the note after a meeting, who belongs on which thread and how to answer a request you never agreed to are each taken separately elsewhere in this collection.