# The Quiet Weeks

The month after you put it up and nothing happens: what to check, in what order, and what to stop doing.

Price: $39

## What the quiet is, and what it is not

You put the thing up. You told the few people you could tell. Since then, nothing.

Not a no. Nothing. A handful of views that went nowhere, somebody who opened the page for nine seconds on a Tuesday, and a number at the bottom of the dashboard that has not moved since the day you published.

Almost nobody describes this stretch honestly, because the people writing about it are selling the launch. The launch is the exciting part and it photographs well. The month after it is the part everybody actually lives in, and it is quiet in a way that feels personal.

It is not personal, and here is the arithmetic that shows why.

Take the number of people who have actually opened your page since you published. Not views on a post. Not followers. The number of human beings who landed on the page where the thing is sold. For most people in their first month that number is somewhere between ten and sixty.

Now pick any rate you like for how many strangers buy a thing the first time they see it. Pick a generous one. Multiply. In almost every case the answer you get is well under one person.

So zero was the single most likely result, and it was the most likely result before you wrote a word. The quiet is not a verdict on what you made. At this size it is barely a measurement at all.

That cuts both ways, and this is the part people skip. If zero tells you almost nothing, then so does one. A single sale in week two is not proof either. You are working with numbers too small to carry an opinion, which means the useful work in the next month is not persuading anyone. It is getting enough people in front of the thing that the number starts to mean something, and knowing which of the four silences you are actually in while you do it.

**Do this now.** Open the page where your thing is sold. Find the count of people who have opened it since the day you published. Write that number down with today's date. If you cannot find that number anywhere, that is your first job and section three is about it.

## The four silences, and which one you have

Silence is not one thing. There are four of them, they have different causes, and the moves that fix one do nothing for the others. Most of the wasted effort in a first month comes from treating silence two as though it were silence one, and posting harder.

| The silence | What you see | The one number to look at |
|---|---|---|
| 1. Nobody arrived | Views on posts, almost nothing on the page | People who opened the page since you published |
| 2. They arrived and left | Page opens happening, nobody scrolls, seconds on page | How many reached the bottom of the page |
| 3. They read it and stopped at the price | People reach the price and go no further | How many started a checkout |
| 4. They bought and asked for it back | A sale, then a refund request | Refunds as a share of sales |

Work down the table in order. You cannot diagnose silence three while silence one is true, because you do not have enough people reaching the price for their behaviour to mean anything.

There is a fifth case and it is the most common of all: **you cannot tell**, because nothing is being counted. If your page has no way of knowing how many people opened it, you are not in a silence, you are in the dark. Fixing that comes before everything else in this book, and it takes about twenty minutes.

Two traps at this stage.

The first is the percentage. One person out of twelve is not eight per cent of anything. At these sizes a single extra person swings your rate by ten points. Count people, in whole numbers, and leave the percentages alone until the bottom number is in the hundreds.

The second is the week-to-week comparison. A Tuesday is not a Sunday, and a week you posted five times is not a week you posted once. Comparing them produces a feeling, not a finding.

**Do this now.** Write four lines in a plain text file, dated today: page opens, reached the bottom, started a checkout, refunds. Fill in every number you can actually see. Put the word "unknown" next to the ones you cannot. That file is the only record you need for the next month, and the unknowns on it are your work queue.

## Silence one: nobody arrived

This is where almost everybody is, and it is the least interesting problem, which is why it gets skipped in favour of rewriting the page for the fourth time.

First, separate two numbers that people mash together. Views on a post are not visits to a page. The distance between them is enormous and it is where nearly everything goes. A post can be seen by two thousand people and send four of them anywhere. That is normal. It is not a sign that the post failed.

So the number to move is page opens, and there are only three honest ways to move it in a first month.

**Put the link where you already exist.** Profile bio, email signature, pinned post, the footer of whatever you already write. This takes an hour and it is the only compounding thing on the list.

**Go where the question is already being asked.** Find one place where the exact people you wrote this for are already describing the problem in their own words. Read thirty threads without writing anything. Then answer ten of them properly, with no link anywhere in the answer. The link comes later, and only when someone asks. This feels unbearably slow. It is also the only version of this that does not get you removed.

**Tell people by hand, one at a time.** Twenty messages, written individually, to people who would genuinely want it. Section seven has the wording.

Now the rule that decides whether any of this teaches you anything: **one place, for two weeks.** If you start a second channel before the first one has a number attached to it, you end up with two channels and no information. Two weeks is enough to know whether anyone arrives from it. It is not enough to know whether it works, and nothing you do in month one will tell you that.

And for seven days, make nothing new. No second product, no redesign, no rebrand. Building is the comfortable move and you already know how to do it. The uncomfortable move is the one with people in it.

**Do this now.** Pick one place. Write its name at the top of the file you started in section two, with today's date and the date two weeks from now. Put the link in your bio before you close the laptop.

## Silence two: they arrived and left

People are opening the page and leaving in seconds. That is not a taste problem. In almost every case the first screen has failed to answer three questions, and the reader has gone before the good part.

The three questions, in the order a stranger asks them:

1. What is this thing?
2. Is it for me?
3. What do I actually get?

Five seconds, all three, above the fold, in words the reader already uses.

Here is a first screen that fails, in the shape most of them fail in:

> **Introducing Ledgerly**
> The modern solution for freelance finance.
> Built by freelancers, for freelancers.
> [Get started]

Nothing there is wrong. It is also unreadable to a stranger. "Modern solution" is not a thing. "Freelance finance" could be six different products. The name is doing work that a name cannot do, because the reader has never heard it.

The same screen, answering the three questions:

> **Every invoice you have sent this year, in one spreadsheet that adds itself up**
> For freelancers who dread the end of the tax year and have receipts in four places.
> A spreadsheet, twelve categories already set up, and a two-page sheet on what counts as an expense.
> $39, once. [What is inside]

Six lines that make readers leave, and what to write instead:

| What is there | Why it loses them | Write this |
|---|---|---|
| The product name as the headline | They have never heard it | What they end up with |
| "Everything you need to know" | Says nothing about them | Who it is for, in their words |
| A long story about you | They do not know you yet | One line, later, lower |
| "Transform your workflow" | Means nothing concrete | The specific thing that changes |
| A features list with no result | They cannot picture using it | What they hold at the end |
| No price anywhere | Reads as expensive or as a trick | The price, in numerals, early |

**Do this now.** Screenshot your own first screen on a phone. Cover everything below the fold with your hand. Read what is left out loud, as somebody who has never met you. Write down which of the three questions it answers. If it is fewer than three, rewrite the top four lines before you do anything else this week.

## Silence three: they read it and stopped at the price

People are reaching the bottom and not buying. This is the only silence where the page itself is genuinely the problem, and it is nearly always the same problem: a question the reader has is left unanswered, so they close the tab to think about it and never come back.

There are ten questions. A stranger asks them in roughly this order, and a page that answers all ten in plain words is a page somebody can say yes to.

1. **What is this, in words I already use?** First line, above the fold.
2. **Is it for me, or for somebody further along?** Say who it is not for. Naming who should not buy is the fastest way to convince the right reader that you know them.
3. **What exactly do I get, and in what form?** A file? Pages? How many? Say the format out loud.
4. **How long will this take me?** An honest number. "About two hours" beats "quick and easy".
5. **What do I have to do myself?** The reader is trying to work out how much of the work is still theirs. Tell them plainly.
6. **Who wrote this and why would they know?** Two sentences. What you have actually done, with nothing inflated.
7. **What does it cost, and is that once?** Numerals, early, and say the word once if it is once.
8. **What happens after I pay?** Where does the thing arrive, how fast, in what form. Unanswered, this reads as risk.
9. **What if it turns out to be wrong for me?** Your refund terms, in one sentence, in your own words.
10. **Why now rather than later?** The honest version of this is usually "no reason, it will be here". Saying that costs you less than a fake deadline costs you.

Then run the check that finds the gaps. Print your page. Go through it sentence by sentence and write the number of the question each sentence answers in the margin. Two things fall out. The questions with no number anywhere are the holes, and they are why people leave. The sentences with no number beside them are decoration, and they can come out.

**Do this now.** Do the margin pass on your own page, on paper. Count the unanswered questions. Write the missing answers as plain sentences first, in a text file, before you touch the design.

## The numbers that lie while they are small

Four numbers will mislead you every week of the first month, and one person on a forum put the whole problem in a single line: "Abysmal click rate on FB Reels (1.5M views, 0 sales)."

**Views.** Every platform counts a view differently, and none of the definitions mean a person read anything. A view can be a fraction of a second of a video that autoplayed while somebody scrolled past. Views tell you the platform showed your thing to people. They tell you nothing about whether those people are yours.

**Followers.** A follow is the cheapest thing a stranger can give you. It costs nothing, means nothing, and does not predict anything at this size. Watch it if it entertains you, but keep it out of any decision.

**Percentages on tiny numbers.** Two of nine is not twenty-two per cent. Until the denominator is in the hundreds, write the two numbers side by side as whole people and let them be what they are.

**The day-one spike.** Whatever happened on launch day happened because you told everyone you know. It will not repeat and it is not a trend. Start your counting on day three.

Three numbers are worth writing down, once a week, on the same weekday:

1. How many people opened the page.
2. How many got as far as starting a checkout.
3. How many human beings sent you words about it.

The third one is the good one, and it is the one nobody records. Ten people opening the page tells you almost nothing. One person writing to say "I was going to buy this but I could not tell whether it works on a Mac" tells you exactly what to do next. At this size the sentences people send you are better evidence than any dashboard you own.

Keep it in one plain text file with dates down the left. Not a spreadsheet with charts. You want to be able to read four weeks of it in fifteen seconds.

**Do this now.** Add the three numbers to the file from section two, with today's date. Set a weekly reminder for the same weekday. Write "0" where it is zero rather than leaving it blank, because a blank in four weeks' time is an argument with yourself that you will lose.

## The five messages you send by hand

Sending things by hand is the part people skip, because it is slow and it can be refused. It is also the only activity in a first month that reliably produces sentences from real people, and section six explains why those are worth more than the numbers.

Five messages. Send them one at a time, personally, with the person's actual situation in the first line. Never paste the same block to a list.

**1. To somebody who asked the question in public.** No link.

> You mentioned you have been doing this in a notebook. I wrote something about the same problem last month. Happy to send it over if it would help, or to just tell you the short version here.

**2. To somebody who opened the page and went quiet.** Only if you can see it and only once.

> Saw you had a look at the sheet. Anything on there that did not make sense? I would rather fix it than have you wonder.

**3. To a person you know who is exactly the buyer.** The hardest one to send, and the most useful.

> I made a thing for people who do the invoicing themselves, which is you. It costs $39. I am not asking you to buy it. I am asking whether the first screen makes sense to you, because I have read it four hundred times and I cannot see it any more.

**4. To somebody who bought.** One question, not four.

> Thanks for buying it. One question, if you have thirty seconds: what were you hoping it would sort out for you?

**5. To somebody who asked for their money back.** Refund first, ask second.

> Refunded, done, nothing else needed from you. If you have a minute I would like to know what you expected that it did not do. You would be doing me a real favour and I will not ask twice.

Rules for all five. Under five lines. One question. Name the thing. No pressure and no second ask. If they do not reply, they do not reply, and that is the end of it.

**Do this now.** Write out twenty names. Send message one or message three to five of them today, one at a time, each with a different first line. Keep every reply in one file and read it again on Sunday.

## The stop rule, and what you keep

Write your stop rule before you need it, because afterwards everything looks like it was nearly working. That feeling is not information. It is what a month of effort feels like from the inside, whatever the thing was.

A stop rule has four parts and fits on one line.

1. **A date.** Six to ten weeks out. Put it in the calendar today.
2. **A number that is not money.** Money is too rare this early to decide anything. Use the third number from section six: how many people sent you words about it.
3. **A thing you will have done by then.** Twenty hand-sent messages. Ten answered threads. Four weeks of the file. This half matters as much as the number, because it stops you judging a plan you never ran.
4. **The decision.** Written down now, while you are calm.

Here is the shape:

> By 30 November I will have sent twenty messages by hand and answered ten threads. If fewer than five people have written to me about it in their own words, I change one thing. If fewer than two have, I stop.

Three outcomes and only three. Keep going. Change exactly one thing, for another four weeks. Or stop.

"Change one thing" means one. The page, or the place you are posting, or who it is for. Changing three at once means the next four weeks teach you nothing, and you will have spent them anyway.

Now the part nobody selling you anything will say out loud. Stopping one thing is not stopping. The writing is still yours. The list of people who replied is still yours, and it is the most valuable object you now own. So are the threads you read, the words you learned that the buyers actually use, and the knowledge of exactly where they gather. All of that carries over to the next thing and none of it was wasted, which is the difference between stopping and failing.

Anyone whose product depends on you continuing will never hand you a rule that lets you quit. That is the whole reason this section exists.

**Do this now.** Write your stop rule as one sentence, with a real date, and put it in the file from section two. Then put the date in your calendar with the sentence in the description, so the version of you that reads it in nine weeks is arguing with the version that was calm.

---

An AI wrote this for you. Change anything you want before you sell it.