An approval isn’t a slow reply. It’s a decision being taken inside somebody else’s organization, usually by a person who has never seen your thread. And it’s usually waiting on something nobody has thought to tell you about. Chase it like an email that got buried and the fourth message does no better than the third. None of them touches what’s actually in the way.
You sent the phase two note on the ninth. Your contact read it, said it looked right, and said she’d get it signed off. It’s the twenty-third. You’ve written twice, and both times she’s said she’s chasing it internally. Nothing in your mailbox tells you whether that means Tuesday or never.
Your two messages were fine. The thread is stuck somewhere inside their building, in a place you can’t see, and that’s the ordinary case.
- Silence here covers six conditions. An approval stalls for about six reasons, and every one of them is something other than forgetting.
- The person you’re writing to is usually somebody other than the person who decides. The work stopped where their organization begins, and you can’t see in.
- Most of the outcome is fixed before you send. Three things make an approval easy to grant: it fits somebody’s authority, it arrives before the meeting that could grant it, and one word answers it. Those are the ones that come back.
- Repeating yourself is the one move with no upside in it. Each attempt has to change something. With a decision, that means the decision itself, the forum, or who’s doing the asking.
Three things hiding under one word
Approval is one word covering three obstacles. A chase so often achieves nothing because it’s aimed at whichever one you happened to assume.
A judgment. Somebody has to read what you sent and form a view. The report, the design, the model, the recommendation. This one needs their attention and an uninterrupted half hour. That’s the scarcest thing they own.
A permission. Somebody with authority has to commit money, risk or other people’s time. This one needs a person at a particular level. There’s nearly always a threshold attached, and nobody has mentioned it to you.
A procedure. A system has to receive a number, a form or a signature. A purchase order, a supplier record, a legal review, a line in a board minute. This one runs on paperwork alone. It has a route, a queue and somebody administering it. It moves at the speed of that queue, however much your sponsor wants the work.
Most approvals in a live engagement stack two or three of these on each other. The phase two note needs your sponsor’s judgment. Then a permission from whoever holds the budget at that size. Then a purchase order raised before you can invoice anything. It stalls at whichever of the three you didn’t know was there. The last one is the one nobody warns you about, because inside the client’s building it’s so ordinary that it seems too obvious to say.
This is why “have you had a chance to look at it” is such a weak question. It addresses the first obstacle only. Say your contact read it two weeks ago, liked it, and has been waiting for the November investment meeting. Your question leaves her nothing to say beyond yes, and the two weeks repeat.
So before you write anything, work out which of the three you’re standing at. The last thing you were told usually tells you. It looks good and I’ll get it signed off is a permission problem. I want to go through it properly is a judgment problem. I’ve sent it to procurement names a queue rather than a status, and it’s worth one short question of its own.
Why it has gone quiet
Break the silence into its parts and it gets much easier to think about. An approval nobody has granted yet sits in one of six conditions. Each one wants a different response from you.
It’s still with your contact. She’s holding it. She’s waiting for a better moment, or for a meeting, or for the version she thinks she can defend. From your side it looks exactly like indifference. It’s usually care: she’s being careful with your proposal because she means to win with it.
It reached them, and something small is missing. A number, a comparison, an answer about risk, a view from a person you’ve never heard of. Nobody asks you for it, because asking would mean admitting the thing has been sitting there.
It’s queued behind a gate with a calendar. Investment committee monthly, board quarterly, procurement in batches on a Wednesday. Everything here is working as designed. You’re four weeks from the next date on which a yes is possible, and the person who knows that hasn’t thought to tell you, because to them it’s the weather.
The decision sits above your contact, and she hasn’t said so. She meant to sign it, then found she couldn’t. It’s usually an honest surprise to her, and it’s very common. It happens most when the amount has crept over a threshold since the last piece of work.
It’s already approved and nobody told you. Somebody said yes in a meeting you weren’t in. The order exists in a system you can’t see. The person who would have emailed you assumed somebody else had. This happens far more than it sounds. It’s the one condition where another chase costs you something, because it lands on a client who believes the matter is settled.
It’s a no that nobody wants to write. Priorities moved. The budget went somewhere else. Your sponsor lost an argument he’d rather keep to himself. A no is unpleasant to send and silence is free, so silence is what you get.
Look at what’s absent from that list. Forgetting. Every one of those six is a live reason held by somebody who remembers, and that’s the whole difference between this and chasing a document out of a client, where forgetting genuinely is on the list and making the task smaller genuinely does move it. Here, a reminder treats one condition in six, and a rare one at that.
Notice also what five of the six have in common. They’re facts you could have had by asking a different question, and mostly by asking it earlier.
Ask for the route before you need it
The cheapest place to spend effort on an approval is weeks before you request one. Spend it on finding out how decisions actually get made in this organization.
Ask at kickoff. It’s a scheduling question, so ask it as one. When we get to the phase two decision, who signs that off, and is there a committee or a purchase order in the way? I’d rather build the timeline around their calendar than around mine. Nobody takes offense at that, and the answer is worth more than anything else you learn that day.
Four things are worth having written down.
Who signs, by name. The person, rather than the role. You can watch for a person, telephone them, and put them on an agenda.
The threshold. Every organization has authority limits, and almost nobody volunteers them. Your contact can approve to one level and her director to another. Knowing where those lines fall changes the shape of every proposal you write for that client from then on, and it’s the single most useful number you can hold about them.
The calendar. Miss a monthly committee by a day and a two week wait becomes a seven week wait. Ask for the dates once. Put them on your own calendar. Work backwards from them when you decide when to send.
The form a yes has to take. Some organizations need a portal, a signed order, a countersigned change note or a minuted decision, and email alone won’t carry it. Ask for approval in a shape the client can’t give and you get exactly the silence you were trying to avoid. From outside, it’s invisible.
All of this lives in conversation rather than in the contract. What a statement of work does hold, and which of its clauses actually do any work once you’re in a mailbox, is a subject this collection takes separately.
One more thing belongs in the same conversation, which is what happens between approval and payment. If a purchase order has to exist before you can invoice, find that out on the day you agree the work. The alternative is finding out when your first invoice comes back from accounts payable with a question on it.
Making the yes cheap
An easy approval comes back on its own. Most of what makes one easy sits in the message you send, and that part is yours.
One decision per message. Three approvals in the engagement means three emails. A message carrying two decisions gets read, gets filed under think about properly, and stays there.
Recommend one thing. Three options is three decisions plus a comparison, and all four of those are jobs you’re better placed to do than they are. Give the recommendation. Then name the alternative you rejected and why, one line each. Somebody can approve a recommendation while walking between meetings. A menu gets discussed instead.
Put the decision in the first line. What you’re asking for, and what it costs, before anybody has to scroll. Context, recap and thanks go underneath.
Say what happens on a yes, with dates in it. If this is approved by the eighteenth, the workshop runs on the fourth of November. That sentence does more persuading than a page of rationale, because it turns an abstract yes into a thing on a calendar.
Write it to survive a forward. This is the part most people get wrong, and it follows from the whole shape of the problem. Your contact is your advocate rather than the decision maker. What she’ll do with your email is forward it, or paste a paragraph of it into an internal thread you’ll never see. So the message has to answer her director’s question rather than justify your work to her. What is this, what does it cost, what do we get, what happens if we do nothing, who has agreed to it so far. The craft of sending a document so that it holds up after it leaves the person you sent it to is worked through properly elsewhere in this collection, on proposals, and everything in it applies here.
Accept the smallest possible yes. Reply approved to this and I’ll start on Monday is an instruction somebody can follow from a phone in a corridor. If the client’s process later needs a form or an order number, that’s a procedure and it can catch up. What you want first is a decision you can rely on.
Land it in time for their meeting. If the committee sits on the second Tuesday, your note goes out early enough for your contact to read it, ask you something, and get it into the pack. Send it on the Monday evening and it goes into a four week wait.
The general anatomy of a request that gets answered applies to anything you ask a client for, decisions included, and it’s set out in this collection’s guide to the weekly update. What’s particular about an approval is the forwarding, and the threshold, and the meeting date.
The second message
One rule about repeated chasing holds for every kind of waiting, and managing client document requests derives it in full: sending the same message twice is sending it once. It’s already been read and already been left. The second copy teaches its reader that your messages can safely be left again.
So the follow-up has to carry something the first one didn’t. Three shapes do.
The smaller question. Ask what would move it, rather than whether they’ve decided. Is there anything you need from me to get this over the line is good. Who else needs to see this before it can be a yes is better, because somebody who has done nothing at all can answer it honestly, and that’s precisely the person you’re writing to. It spares them any embarrassment, and the answer tells you which of the six conditions you’re in.
The either or. Is this waiting on the November committee, or on something at my end? A choice between two is far easier to answer than an open question, and both answers are useful. If it’s the committee, you now have a date. If it’s something at your end, you have work to do this afternoon.
The consequence, once it’s true. A fact with a date attached, stated once. The team I had lined up for the first half of November is being asked about other work next week. This is the last rung on the ladder and the only one carrying any weight, so it has to be true and you have to mean it. Say it, let the date go by, and carry on as though nothing happened, and every message you send that client afterwards is worth less.
Two things to leave out. Skip the apology. Sorry to chase tells the reader that the request was an imposition, which is the frame you least want. You’re removing an obstacle from work they’re paying for. And keep a message that adds nothing down to once a week at most. When you have nothing new to say, the line belongs in the standing weekly update, where it costs the reader nothing and accumulates a date each time, rather than in a separate email that reads as pressure.
How long a quiet thread has to stay quiet before it’s worth acting on, and why the careful message you took trouble over is the slowest of all to come back, is measured properly in unanswered emails.
Proceeding on an assumption
One sentence shifts more stuck approvals than any follow-up ever written, and it’s worth understanding before you use it.
Unless I hear otherwise by Thursday, I’ll proceed on option B.
It works by moving the cost of silence. Up to that point, saying nothing was free for the client and expensive for you. Afterwards it’s the other way round. A decision that had no deadline suddenly has one, invented by you, which is the only kind you were ever going to get.
It’s legitimate when four things are true. The choice is reversible, or cheap enough to reverse. The cost of being wrong falls on you rather than on them. What you’re proposing sits inside work already agreed, so you’re choosing a route rather than expanding the job. And the client can genuinely stop you with one word, which means you’ve given a real date, a plainly stated default, and an easy way to refuse. Say stop and I’ll hold is better than let me know if you have any concerns, because it names the reply.
Keep it away from three places. Anywhere the answer costs money that nobody has approved. Anywhere a third party will rely on the outcome. Anywhere with legal, regulatory or safety weight. It’s also wrong in the first weeks of a relationship, where it reads as pressure rather than as efficiency, and wrong on anything you’d be uncomfortable being shown eleven months later in a thread where nobody replied.
Where you do use it, write it in the thread the decision belongs to rather than in a fresh message, and use words a search will reach for in a year: the choice, the default, the date. If it turns out badly, the sentence that protects you is the one nobody answered, and it protects you once somebody can find it.
All of that covers choosing a route. Starting extra work on the theory that nobody said no, and invoicing for it later, is a scope change you granted yourself. Recognizing a scope change and pricing it rather than absorbing it is its own guide in this collection.
Changing the decision rather than the message
Past the second attempt, the useful moves are about the ask rather than the wording. Four of them change the thing you’re asking for, or where you’re asking, and they’re in rough order of how often they work.
Make it smaller. This is the move that does most, and it’s honest work rather than a trick. An approval stuck at an authority level can often be split so that the first part fits comfortably under the level below. A phase two that nobody your contact can reach is able to sign becomes a short discovery piece she can sign this afternoon. The larger decision comes later, taken by people who have by then watched something work. You get started, they get evidence, and the big number gets easier rather than harder. Half the reason discovery phases exist is this.
Make it reversible. People approve reversible things quickly and irreversible things slowly, and the difference is enormous. A pilot in one region, a two week trial, a draft issued for comment rather than for adoption. Where a decision genuinely can be made reversible, say so out loud. Most of the delay was somebody’s private caution, and that sentence answers it.
Change the forum. If a committee has to grant this, your question is how an item gets onto that agenda and who puts it there. Somebody owns that agenda. That person has a name, and they can usually add a line to it the same day. Writing to your contact rarely surfaces this, because she thinks of it as furniture rather than as information.
Change who is asking, with your contact rather than around her. Would it help if I wrote a short paragraph you can send on, or joined for five minutes to answer questions? That hands your advocate a better weapon and costs her nothing. Going round her, straight to her director, occasionally buys a decision and reliably costs you the relationship that produces the next three pieces of work, and it’s almost never necessary. If it truly is time to escalate, ask her to escalate.
And underneath all four, the phone. Fifteen minutes on a call tells you which of the six conditions you’re in, where email rarely does, because none of those conditions is a thing anybody would type. The general case for changing channel belongs to the documents guide. What’s different here is what the call is for. With a document you’re making an errand smaller. With a decision you’re finding out what the obstacle is, and obstacles live in people’s heads.
The slot you are holding open
The cost of a stalled approval is the capacity. You’re keeping time free for work that may not happen, and you turned something down in order to keep it. That cost is real, and your client can’t see it. They have no idea you’re holding anything.
Which gives you the most useful sentence in this whole subject, and it’s a statement about your diary.
I’ve held the first two weeks of November for this. I can keep them until the twenty-second, after which I’ll need to offer them to somebody else. That’s not a deadline on your decision, it’s just where my calendar is.
It works for three reasons. It’s true, so there’s nothing in it to argue with. It describes your business rather than their behavior, so it lands as information. And it supplies the one thing a decision inside somebody else’s organization lacks, which is a reason to happen this week rather than next. Then you have to honor it. A held slot you quietly keep holding after the date is a promise you’ve taught them to discount.
Two things to steer clear of here. Billing for standby: almost no independent makes that stick, and trying converts a soft yes into a hard argument. And letting the delay change the job in silence. Work that slips six weeks is often different work, because the team has turned over or the data is stale or the window you were designing for has closed. Say so when it happens, rather than absorbing the overrun. The hours that vanish into waiting, and how to stop giving them away, are taken up separately in this collection.
Ending it cleanly
Past the third attempt, what you’re short of is an answer of any kind. The fastest route to one is to offer to close the thing.
I’m going to take this off my open list and assume phase two is parked for now. If that’s wrong, tell me and I’ll put it straight back.
Two sentences, sent warmly, with no edge on them. They work for two reasons. Saying no, don’t do that is much easier than making a decision. And a client who has been quiet out of embarrassment gets a way to say something true, with no failure to report. A fair share of the time this produces the approval. Most of the rest of the time it produces the honest answer, which is worth having.
Take that answer well. A no is worth more than a maybe. A parked engagement you’ve genuinely written off is capacity you can sell, and a parked engagement you’re half holding sits idle and unsold. It’s also worth remembering that a no in September about a decision that was budget shaped is frequently a yes in February. So give the parked thing a date rather than deleting it. Come back after their planning round, or after the reorganization settles, rather than whenever it next crosses your mind.
Then keep the thread. Eleven months on, somebody calls you about the same problem. You know what stalled it last time, who was uneasy, and what has changed since, and nobody bidding against you knows any of it. Catching up on a long chain is how you get that back at speed. Your thread is also the only place that history exists, because nobody at the client is keeping it for you.
What Point holds while you wait
Everything above is a practice rather than a product, and consultants ran it on index cards long before software came near it. What software changes is where the effort falls. Point is a full email client sitting on the Google or Microsoft 365 account you already have, so the request you sent and the answer still to come are in one place. The useful part is that the waiting stops being something you carry in your head.
When you send the approval note, you say on the thread, in ordinary words, that you’re waiting on it. That turns into a dated item carrying the conversation with it, and you stay in one application the whole time. Come the date, if it’s still quiet, the thread comes back to the top of your feed and nudges you, and what you send is yours to write. If the answer lands first, the thread is live again. Point also notices work that looks finished. Point marks the task as looking done, shows the reason, and waits for you to agree.
An approval so often turns on something arriving rather than on somebody replying, so the standing watch is frequently the more useful instrument. Tell me when the signed order arrives. Tell me when anybody from procurement writes. Tell me when Priya replies on the phase two note. One alert reaches you on the morning it fires, and your other mornings stay quiet, which is what setting a watch does in an ordinary week. Waiting stops meaning checking.
Mail is weighed before you reach it, on what it costs to be late as well as on how much it matters, so an approval you’re blocked on sits above a renewal notice on a busy morning. Important versus urgent is that distinction taken apart. Replies come back drafted the way you write, for you to change or send, though the sentence about your November calendar is one only you can write. Where the answer turns out to be that this needs fifteen minutes on a call, Point runs the scheduling back and forth, against calendars Point has already reconciled. And when the substance stays with you but the wording has gone, a search understands the thread where they mentioned the committee meeting monthly.
How far Point goes without you is a setting for each kind of action, from suggest only, up through the position where Point readies something and stops, to handling the whole thing. Every kind starts at review. What gets done is written down in order with a time against it, and undoing something starts from the line that recorded it, with the one honest exception, which is that a sent email stays sent once it’s on another company’s mail server.
Point stops where the deciding starts. Point can’t see inside a client’s procurement system, and Point holds no view on whether a particular approval is worth another week of your attention. If your real problem is that decisions vanish into a portal you have no login for, that’s their system, and a mail tool stays outside it. What Point takes off you is the part that was never anybody’s job: remembering, on the right morning, what you’re owed and by whom. The rest of what Point does is on the benefits page, and the version of this argument written around a consulting practice as a whole is Point for consultants.
Common questions
How long should I wait before chasing an approval?
Work from events rather than from days. What matters is whether anything has happened that could have produced a decision, so the clock to read is the date of the next event that could, whether that’s a committee, a board, a budget round or your contact getting back from leave. If that date has passed and nothing came, you’re genuinely waiting. If it’s still ahead of you, another message mostly signals that you don’t know how their organization works. For a general sense of how quickly replies of any kind arrive, and why the careful ones are slowest, unanswered emails has the measurements.
Should I copy my contact’s boss?
Rarely, and never as a first move. Your contact is the person selling this internally on your behalf, and going round her turns your best advocate into somebody managing an embarrassment. Go with her instead. Offer her a paragraph she can forward. Offer to join a call for five minutes. Or ask her directly whether it would help for you to raise it a level. If she says yes, you’ve escalated with her blessing, which is a different thing entirely and costs nothing.
What do I say when the delay is costing me money?
Say what’s true about your own calendar, once, in a sentence about your business rather than their behavior. That you’ve held two weeks in November. That you can hold it until a particular date. That after that you’ll need to offer it elsewhere. Then do what you said. Leave standby billing alone, because almost no independent can make it stick and it turns a soft yes into an argument. And keep it framed as your calendar rather than as a deadline on their decision, which it isn’t and which they’d resent.
Can I start work without written approval?
Yes, on four conditions, and it’s often the professional move. The thing is reversible. It sits inside scope already agreed. The cost of being wrong falls on you. And you’ve said in writing what you’re about to do, with a date on which they can stop you. No on anything that commits money, that a third party will rely on, or that carries legal or safety weight, and no amount of goodwill in the relationship changes that. The distinction that matters is whether you’re choosing a route inside agreed work or quietly making the job bigger, because the second one is a scope change however it started, and this collection takes scope changes on their own.
Is chasing an approval the same as chasing a document?
They behave differently, which is why they’re separate guides. A document is an errand. Errands respond to being made smaller, to being named the way the client’s own paperwork names them, and to being asked for over the phone. The whole method is in managing client document requests. A decision works another way. Somebody remembers it. Making it smaller means changing what you’re asking for rather than how you ask. And the obstacle is usually a person or a calendar you’ve never been told about.
What if I am promised the approval every week and it never comes?
At that point you’re being told what somebody wishes were true, which is a kinder thing than it sounds and no use to you at all. Status is the question producing the answer you keep getting, so change the question. Ask for a date on which a decision becomes possible. If nobody can give you one, ask to park it. Offering to close it is the reliable way to find out whether there was ever a yes in there, and it’s a good deal cheaper than a fifth chase.
The short version
An approval is a decision inside somebody else’s organization, and it’s stuck at a judgment, a permission or a procedure. It’s gone quiet for one of about six reasons, and every one of them is something other than forgetting, so a reminder treats a condition you probably don’t have. Nearly everything you can influence sits before you send. Know who signs, at what level, and through what route. Know when they meet. Then write the request so it survives a forward, so one word answers it, and so it lands before their meeting rather than after. When it does go quiet, make the second message carry something new. A smaller question. A choice of two. Or a consequence you’ll actually honor. Past that, change the decision rather than the wording. Make it smaller. Make it reversible. Get it onto an agenda. Or arm your contact to escalate for you. Be honest about the calendar slot you’re holding, and when it has run out of life, offer to park it, which produces an answer more often than another chase. The inquiry that began all this, the scope change nobody confirmed in writing, and the several live engagements all wanting the same Tuesday each have a guide to themselves in this collection, which opens on the consulting practice as a whole.