You Have Read That Email Before

Both of the loud stories about this technology reached you as somebody’s pitch. That puts you in an unusual position, because reading a pitch is your trade.

The first one arrived at twenty to seven on a Tuesday morning, from a rep you have never met. Lower-case subject line. A first sentence about how manual prospecting is finished, a second about an AI that books qualified meetings while you sleep, and a P.S. offering fifteen minutes Thursday or Friday, whichever is easier. You could see the frame under the paint. You knew what the platform costs, that the same words went to four hundred people before breakfast, and that this was step three of five. You have sent that email yourself. Possibly last quarter.

The second story never reaches you directly. It goes into rooms you are not in, and it says that if a machine produces outreach at ten times the volume, some share of the people producing it is a number that can come down. You get it second-hand: a question in a one-on-one about coverage per rep, a territory that quietly got wider without any headcount.

The two agree on nothing except their size. Both are set in a year nobody has closed yet, and neither moves anything on your board this week. Believe whichever you like. Thursday’s forecast call still wants a number you do not entirely have. Two commits have gone quiet, which is not the same as dead and is worse. Six discovery calls this week produced six recaps, and you sent two. The CRM holds a version of the last three weeks that you will rebuild on Sunday night out of your own calendar. That is not a record. It is archaeology.

This book has no view about the year nobody has closed. It is about the four recaps.

Start with the first story, because it falls over on the parts of the job that happen between two people. Not one of them is futuristic.

It cannot read the room. Nineteen minutes into a first call the VP stops taking notes. That is the whole event: a pen goes down. The transcript records a sentence about integration timelines and nothing else. You saw the room change, dropped the next two slides, and asked what happens to her team if this slips past March. Deals turn on that, and no recording explains it.

It cannot tell you the quiet one was the decision-maker. Somebody said almost nothing for forty minutes and asked two narrow questions about how the data gets in. The loud one deferred to them without appearing to notice he was deferring. You noticed. A transcript gives you word counts, and word counts get that exactly backwards.

It cannot build the thing that makes somebody take your call in year three. That asset is built out of small unpaid moments: the time you said this one is not for you, buy the other one; four years of never wasting a Tuesday of theirs. Its whole value is that a person did it.

It cannot negotiate. At ten to five on the last day of a quarter, procurement says a number and then says nothing at all. What happens in that silence is nerve, judgment, and the floor you settled on before you dialed. A model has nothing at stake and no way to hear that the pause was theater.

And it cannot tell you why the deal actually stalled. It will offer reasons, and they will sound sensible. The real one is usually that your champion’s budget moved sideways in a reorganization she has been told not to discuss, and it surfaces only in a phone call where you ask a question you would never put in writing.

One more, and this one can cost you money: it does not know your product. It will sound as though it does. It will describe a capability that got cut two releases ago in a clean, confident sentence, and it will price your product for you if you let it. Chapter 10 covers what happens then, which is that whatever you put in writing becomes your company’s claim.

That is the first story answered. The number still lands on a person, and no vendor has offered to take that part. It is also the only reassurance in this book: reassurance is cheap and you have a call at ten.

The unease underneath it is earned, though. Your buyer does most of the research before you appear now, and turns up with a shortlist and half an answer already formed. The volume tools made outreach nearly free, and nearly free ends up worthless: an inbox that once rewarded a good cold email now punishes all of them equally. So the thing to keep an eye on is not a machine closing a deal. It is whether you get the meeting at all. Chapters 4, 5 and 9 are about being the rep worth answering.

Now the concession, because you can smell an oversell as well as anybody alive, and a book claiming nothing had changed would lose you somewhere around Chapter 4.

Something is changing. Not the selling. Everything written around the selling.

Nobody warns you about that when you take a bag. You signed up for conversations, and you got them. You also got the second job that starts when the conversations stop. The research that makes an opener worth opening. The three follow-ups nobody answers. The same-hour recap: what they said, what was agreed, who owes what by when. The proposal that hands their problem back in their own vocabulary, because your champion has to forward it to two people you have never met. The CRM fields. The forecast narrative that separates what is real from what is hope. The referral ask you never send, because making it sound natural takes twenty minutes.

None of that is selling. None of it is comped. It gets done after six, or on a Sunday, or it does not get done.

All of it is words. Words are the one thing the machine is reliably good at. Not the read on the call — the six sentences that carry your read to the four people who were not on it.

How they fail never comes up in the demo. They make things up, and they do it in the voice they use for everything else. Your version of that problem is worse than most, because the sentence this machine most wants to write is the persuasive one. Ask it for proof and fifteen seconds later you have a case study: a plausible customer, a percentage, a quote with a job title attached. Well made, and not real. It has not read your price book.

So every workflow here divides the same way. Facts are yours: claims, prices, numbers, what was said, the read on the room. Sentences are its side of the line. Where that line is easy to step over, a WATCH OUT box writes out the check. It runs about a minute. The book calls it the 60-second check, and it is the one ritual here that is not optional.

One thing about money, said once. Buyers talk to several vendors in the same two weeks and mostly go with whoever is easiest to say yes to internally: the recap that lands before their next meeting, the security questionnaire answered Tuesday instead of the following Monday. Nothing new in that. The change is what speed costs you, which is minutes now instead of an evening.

There is a particular kind of Thursday in the third week of a quarter, and anybody carrying a number knows it.

Six meetings, back to back, nine minutes between them. The nine minutes go on coffee and your SE asking which version of the deck to use. So nothing gets written all day, which is fine, because at four o’clock it is all still in your head. Most of it is.

At twenty to seven you write one recap properly, for the biggest deal, and it is very good. It uses the CFO’s own phrase for the problem. It states the three things that were agreed and the one that was not. Your champion forwards it to two people you have never met, and it makes you easy to argue for in a room you are not in.

The second recap gets four lines at ten past eight, because it is Thursday and there is one of you. Four lines is not the thing that gets forwarded.

The other four never get written. By Friday afternoon each has thinned to a sentence. By Sunday night, when you open the CRM to make the week presentable before the forecast call, two have become “good call, sending pricing”. That is not a note. It is a shrug in a text field.

Three weeks later one of those four goes quiet, and nobody can say when it turned, including you, because there is no record of what was said to measure the silence against. It closes a quarter late and four points thinner, or it does not close and the reason logged is “went with competitor”, which is what gets typed when the honest answer is that you lost the thread.

None of that is a discipline problem, whatever the slide at the QBR says. It is arithmetic. A recap worth forwarding takes twenty-five minutes. Thursday brought six, and there is one evening and one of you.

What this is, before Chapter 2 asks for your thirty minutes.

A working manual for using a plain AI chat tool on the written half of selling. ChatGPT, Claude, Gemini, Copilot, whichever your company already pays for. Research digests and openers that are relevant because you did the research. Pre-call briefs built in the nine minutes you have. The same-hour recap. Proposals, objection prep, negotiation memos, mutual action plans, the stalled-deal ladder, the loss letter. CRM notes out of a two-minute ramble in the car. Forecast narratives, handoffs, win-loss. Battlecards built from official materials. There are 150 prompts, numbered, and the library at the back turns one up in ten seconds while you walk from the car to reception.

All 150 were tried in ChatGPT, Claude and Gemini, and any prompt that only behaved in one of the three was dropped. None of it is a guess about what AI may eventually do for sales teams. It is the wording that survived all three.

What it is not: an AI SDR, a lead-gen platform, a data provider, a research source, or a substitute for discovery. It does not know things; it organizes what you give it, and Chapter 8 is strict about the difference. It is also not your company’s AI policy, which outranks every page in here. Most people this series is written for work for themselves. You almost certainly do not, so the first job in Chapter 2 is a five-minute check on what you are allowed to paste. Any time the tool gets near something expensive (claims, prices, confidential information, outreach law, call recording) a RULES CORNER box tells you where the line sits. Chapter 10 puts all of them in one place, and it wants reading before a prospect’s name gets typed into a chat window.

There is nothing technical in any of it. You need a phone, half an hour, and the work email you already have. The one real skill lives in Chapter 3, and you sell it already: briefing this thing is briefing your SE before a demo. The situation, what you want, what to leave out, what it should look like coming back. Brief it vaguely and it hands you the standard deck.

Last thing before you start, and it is about size.

Nothing here changes your life or closes a deal by itself. What it changes is the block between six and eight, and inside a month, most of your Sundays. The page opposite adds it up chapter by chapter and arrives at roughly eleven hours a week. That number is real, and it costs something up front: week one takes time instead of giving any back. Around week three the arithmetic starts running the other way.

Most reps finish with two habits, usually the research loop and the same-hour recap. Two is the target, not the consolation prize. Run them until they stop feeling like procedure, and take back four or five hours a week. Reaching Chapter 12 earns nothing on its own.

Nobody is coming for your number. Somebody still has to hear the pen go down, sit still in the silence after procurement names a figure, and ask the question that has no polite written form. What you get here is smaller than both stories and worth more than either: the recap in their inbox before their next meeting starts, the proposal out the day it was promised, and a Sunday evening not spent excavating your own week.

The next page puts the hours beside the chapters that produce them. After that it is thirty minutes of setup, and you can do it on a plane.