Finding the Story Buyers Want to Hear

You sit down to write a story about your business and nothing comes.

You have a list of facts instead. You launched a product. You got some customers. You’ve been at this a while. None of it feels like something a stranger would stop and read.

That blank feeling usually gets blamed on having a boring life. The real cause is different. You’re searching your memory for something dramatic, and drama is rare. The search comes back empty, so you decide you have nothing worth telling.

The stories that sell are almost never dramatic.

They’re small and they’re specific. A customer who stopped dreading her inbox on Sunday nights. A buyer who cancelled a piece of software he’d been paying for since the day he started.

Small stories don’t announce themselves, and that’s the trouble with them. They sit inside ordinary weeks, buried under the parts of your work you think of as routine. You walk past them every day because they don’t look like material to you.

Sit with that for a second, because it changes where you look.

The story isn’t out ahead of you waiting to be thought up. It’s behind you, in a week you’ve already lived, and it has been sitting there the whole time.

Nothing has to be invented here. You already own every story you’ll ever need for your marketing. What’s missing is a dependable way to dig those stories out, look at them with clear eyes, and decide which pieces deserve to go in front of a buyer.

Digging beats inventing for a plain reason.

Invented stories sound invented. Your readers have seen a thousand fabricated marketing tales and they know the shape of one. The moment they suspect you made it up, they stop trusting the offer attached to it.

A dug-up story has texture that a made-up one can’t fake. It has the wrong weather, the odd detail, the part that doesn’t quite fit the moral. Readers feel that texture even when they can’t name it, and it’s what makes them believe you.

The digging has three sources, and you’ll use all of them.

Your own history is the first. Your buyers are the second. The daily work you do without thinking about it’s the third, and that one turns out to be the richest of the three.

Start with the mistake almost everybody makes on the first try.

Asked for a story, they describe the product. They explain what it includes, who it’s for, and what it does. That’s a description, and a description isn’t a story.

A story requires change. Something has to be one way at the start and a different way at the end. If nothing is different by the last line, you wrote a brochure paragraph with a friendly tone on top of it.

Here’s why that difference costs you money.

A reader can hold your description up against a competitor’s in about four seconds, and the cheaper one wins that comparison. Nobody can comparison-shop a change that happened to a real person, because there’s nothing to hold it up against.

So the question you ask yourself is never “what does my product do?” The question is “what was different for somebody afterward?” Those sound similar. They pull completely different material out of your memory.

Try it on something concrete.

Say you sell a course on email marketing. The description is that it teaches subject lines, segmentation, and send times. Nobody’s pulse changes reading that sentence, and if you’re honest with yourself, yours doesn’t either.

Now ask what was different afterward. A woman who used to write one email a month because each one took her an entire Saturday now writes four in an hour. She got her Saturdays back. That’s the change, and Saturdays are a thing people want.

Notice that the change had almost nothing to do with the features.

Segmentation didn’t appear in that sentence. The reader doesn’t need to hear about segmentation to want what happened to her, and hearing about it too early gives them something to argue with.

This is also why feature lists feel exhausting to write. You’re trying to make a spreadsheet emotional, and spreadsheets refuse. Once you’re looking for changes instead of features, the writing gets easier, because you’re describing something that happened instead of arguing for something you sell.

There’s a second reason changes work better than features, and it has to do with belief. When you claim your product is good, you’re an interested party and the reader knows it. When you describe what happened to somebody, you’re a witness.

Witnesses get believed more than salespeople. That isn’t a trick of psychology so much as ordinary common sense. A witness has no obvious reason to lie about a detail like Saturdays, so the detail gets to pass through the reader’s defences.

You can hold onto your claims.

You just make them after the change has already done its work. The reader who wants her Saturdays back will happily read about segmentation once she understands what segmentation buys her.

So the order is the change first, the mechanism second, and the offer third. Most writing does that backwards, which is why most writing gets skimmed and abandoned somewhere near the middle of the second paragraph.

You just make them after the change has already done its work.

AI is useful at this exact step, and it isn’t useful in the way people expect. You’re not asking it to write anything yet. You’re asking it to look at your product and pull the buried changes out of it.

Here is my product and who it’s for: [paste your product description and your audience]. Ignore the features entirely. List fifteen specific things that would be different in this buyer’s week after they use it. Be concrete and small. Don’t write benefits like “saves time” or “makes more money.” Write changes like “stops re-writing the same email every Monday.” For each one, name the day, the task, or the feeling that changed.

Run that prompt and you’ll get fifteen candidates, and roughly ten of them will be useless. That’s the expected result and not a sign anything went wrong. You’re mining, and mining produces mostly rock.

The five that survive are the ones where you thought “oh, that’s true, I’ve seen that happen.” That flicker of recognition is your signal. A change you recognise is a change you can write about truthfully, because you have somewhere to go for the details.

Save the discarded ten anyway, in the same file you save everything else in.

A change that looks generic today can turn specific later. A customer says the same thing in an email, and now you have two sources pointing at one idea.

Your own past is the easiest source to reach and the hardest one to see clearly.

You’ve told yourself the events so many times that they’ve worn smooth. The parts that would grip a stranger are the parts you find boring, because you were there.

The fix is to stop searching for stories and start searching for moments. A story feels like a big thing you have to have. A moment is small enough that you’ll admit you have several.

The fix is to stop searching for stories and start searching for moments.

Four places hold almost every moment you’ll ever use, and it helps to go through them one at a time instead of staring at your whole life at once. Each one produces a different flavor of story, and each has a different job in your marketing.

The first is the failure that came before the fix. Every method you teach exists because something else didn’t work first. That failure is the most valuable thing you own, and it’s the thing you’re most tempted to leave out.

You leave it out because it makes you look bad, and looking bad feels like a poor sales strategy.

It’s the opposite. A reader who is currently failing at the same thing can’t relate to your success. They can relate to your failure immediately.

The second place is the day you nearly quit. Not a metaphor for quitting, but the specific afternoon you sat down and did the arithmetic and thought about going back to a job. Everybody in your audience has had that afternoon or fears having it.

The third place is your first customer.

First-customer stories work because they contain surprise, and surprise is the ingredient that keeps somebody reading. You didn’t expect it to work. Then it worked, and you weren’t sure why.

The fourth place is something you believed and stopped believing. Belief changes are the most useful stories in marketing, and almost nobody tells them. They give your reader permission to change their own mind without feeling stupid about it.

That last one needs a little unpacking.

When you tell somebody they’re wrong, they dig in and defend themselves. When you tell them that you were wrong about the same thing, they lean forward instead, because you’ve just made being wrong survivable in front of you.

Once you’ve picked a moment, the next task is recovering the detail. Detail is what separates a story a reader believes from a story a reader skims. It’s also the first thing your memory throws away, because you never expected to need it again.

Detail earns belief because invented stories rarely have any. A person making something up says “I was struggling.” A person remembering says something else. “I had a spreadsheet open with the same four numbers in it. I’d been staring at them for two hours.”

Once you’ve picked a moment, the next task is recovering the detail.

The second version isn’t better writing. It’s better evidence. The reader’s mind accepts the spreadsheet because nobody bothers to invent a spreadsheet, and once the spreadsheet is accepted, everything around it gets accepted too.

You recover detail by answering questions, not by trying harder to remember. Sit with the moment and ask what was on the screen, what time it was, who else was in the house, and what you did right after. The answers come back faster than you’d expect.

This is another place where AI does something you can’t easily do alone.

It can interview you. An interviewer asks the follow-up question that you’d never think to ask yourself, because you already know the answer and forgot it was interesting.

Be careful about one thing while you dig. Don’t clean the memory up as it comes back to you. The embarrassing part, the petty thought, the thing you did that you wouldn’t repeat, those are the pieces that make a reader trust the rest of it.

I want to recover the details of a moment from my own history so I can write about it truthfully. Here’s the rough memory: [describe the moment in two or three sentences]. Don’t write anything for me. Interview me instead. Ask me one question at a time, twenty questions total, about the physical details, the time of day, what I could see, what I’d tried before that, what I told myself, and what I did immediately afterward. Wait for my answer before asking the next one.

That prompt works because of the one-question-at-a-time instruction. Given twenty questions at once, you’ll skim them and answer four. Given one question, you’ll answer it properly, and the answer will remind you of something you’d forgotten.

By question fourteen or so, you’ll typically hit something you didn’t plan to say. That’s the part of the story that will do the work when you publish it. You’d never have reached it by staring at a blank page and trying to be interesting.

The instinct with AI is to ask it for a story.

You describe your business, you ask for a compelling narrative, and you get back something smooth, plausible, and completely hollow. Then you decide AI can’t do storytelling and you go back to struggling alone.

The output was hollow because you gave it nothing to work with. You asked a stranger to write your memoir. It had no memories, so it produced the average of every business story it has ever seen, which is exactly what you got.

Then you decide AI can’t do storytelling and you go back to struggling alone.

Turn the relationship around. AI is a poor storyteller and an excellent interviewer, and interviewing is the part of the process you’re worst at. You can’t interview yourself, because you never ask yourself the obvious question.

A good interviewer has three qualities you can request directly. Persistence, which means following up instead of accepting your first vague answer. Curiosity about the physical world. And a total lack of embarrassment about asking small questions.

That third one carries more weight than it seems to.

A friend won’t ask you what you were wearing or what you ate that day, because it feels intrusive and pointless. The small questions are where the usable material sits, and AI will ask them without hesitation.

There’s a practical setup that makes this work far better, and it takes one extra sentence. Tell the model who your audience is and what you sell before you start, so it knows which of your answers to chase.

Without that context, it will ask evenly about everything.

With it, the model hears you mention something offhand and stops to dig there. It understands that a detail about your reader’s exact frustration is worth more than a detail about your childhood.

After the interview, you have a transcript. A transcript isn’t a story yet, and this is where people quit. The raw material looks like a mess on the screen, so they assume the hour was wasted and they go back to staring at a blank page.

The transcript will run long and feel useless on the first read. That’s what a transcript is supposed to look like, and judging it by how it reads is a mistake, because nobody was writing while it was being made.

The mess is fine. A story is made of maybe two hundred words pulled out of two thousand, and the two thousand had to exist first so you’d have something to choose from. Nobody writes a good story straight through on the first try.

So the next task is pulling the good parts out. You hand the transcript back and ask for the pieces that would land with your specific reader. Ask for them ranked in order, with a short explanation of why each one would work on somebody like them.

Nobody writes a good story straight through on the first try.

Ask for the explanation, not the ranking alone. The explanation tells you whether the reasoning is sound or whether the model liked a piece because it sounded literary. You keep the judgement. It does the sorting.

Expect to reject about half of what comes back. The model will over-value anything that sounds like a lesson, because lessons are what business writing usually contains. The details worth keeping are the ones that teach nothing and simply happened.

Here’s the transcript of an interview about a moment from my working life: [paste the full transcript]. My readers are [describe them] and I sell [describe your offer]. Pull out the eight most specific, concrete details in this transcript. Ignore anything general or philosophical. For each detail, tell me in one sentence why a reader in my audience would recognise themselves in it. Then rank the eight by how strongly a reader would feel it.

You’ll disagree with the ranking, and disagreeing is useful. The one you keep defending in your head is the one you should write, because your defence of it is the story trying to come out.

Don’t ask the model to write the story once the interview ends. You’ll get something competent and slightly false, and the falseness comes from the parts it filled in where your answers were thin. Write it yourself from the details it pulled out.

Keep every transcript, even the ones that go nowhere.

A detail that’s wrong for today’s email is often exactly right for a sales page you write three months from now. Re-reading an old transcript is faster than running a fresh interview.

Your own history has a limit.

You’ve lived one life, and if you publish often you’ll run through the good parts of it faster than you expect. Your buyers have lived thousands of lives, and they’ve been handing you their stories for years.

The handing-over happens in places you don’t think of as story sources. Support emails. Refund requests. The reply somebody sent to a broadcast that you read and never answered. Reviews. The question that keeps arriving in different words from different people.

These are better than your own stories in one specific way.

They come from somebody the reader can see themselves in. When you tell your story, you’re the person who already succeeded, and that creates distance. When you tell a buyer’s story, the distance disappears.

Start with the support inbox, because it’s the richest and the most ignored. A support email is somebody describing a problem in their own words, under pressure, with no reason to perform. That’s the most honest writing your audience will ever do.

Refund requests deserve special attention, and this is the part people resist.

A refund request tells you exactly where the promise broke. That’s uncomfortable to read, and it’s also the raw material for the most trustworthy story you can tell.

Here’s why. A story where everything worked is a story your reader half-believes. A story where something went wrong, and you name what went wrong, buys enormous credibility for everything else you say in the same piece.

Reviews are useful in a narrower way.

Most reviews are too short to contain a story, but they contain the sentence. The one line where somebody says what changed for them is often better phrased than anything you’d write, because they lived it.

The repeated question is the last source and possibly the best one. When the same question arrives thirty times in slightly different language, you’ve found something your entire audience is stuck on, and stuck is where stories start.

Once permission is handled, collecting is mechanical work. Export the emails, paste in the reviews, gather up the replies, and put all of it in one place. Don’t read any of it carefully yet. Get the volume together first and judge it afterward.

Reading carefully too early is what stops people. You read three emails, feel something, get distracted by one of them, and never build the pile. The pile is the point, because patterns only appear when there’s enough material for a pattern to show up in.

Once the pile exists, you’re looking for the sentence that repeats.

Not the same words, but the same complaint wearing different clothes. Six people describing the same Sunday-night feeling in six different ways is worth more than one dramatic email.

The repeated complaint is a story that already has an audience, which is the rarest thing in marketing. You don’t have to convince anybody it’s important. They’ve told you it’s important by bringing it up unprompted, over and over.

They’ve told you it’s important by bringing it up unprompted, over and over.

There’s a version of this that goes wrong, and you should watch for it. If you only collect from people who bought and stayed, you end up with a pile of stories about people who were already halfway convinced. The buyer you want to reach hasn’t been convinced of anything yet.

So keep the emails from people who bought and refunded, the ones who never bought at all, and the ones who wrote to say they were thinking about it. Those describe the state of mind you’re writing to, and they’re the hardest ones to keep because they sting.

Here are [number] pieces of writing from my customers, including support emails, replies, and reviews: [paste them all]. Find the complaints, frustrations, and small wins that repeat across more than three people. Group them. For each group, quote the two or three sentences where somebody said it most plainly in their own words. Then tell me which single group would make the strongest story for someone who has this problem right now but hasn’t bought anything from me yet.

That last instruction is the one doing the heavy work. Left alone, the model will hand you the most emotional group. The most emotional group is often a problem only your existing customers have, which makes it useless for reaching anybody new.

You want the complaint that a stranger would recognise before they’ve ever bought from you. That’s the story that pulls people in from outside, and the difference between the two is the difference between growing and talking to the same people forever.

By now you have more raw material than you can use, which is a better problem than the one you started with.

The next task is throwing most of it away, and doing that on purpose instead of by accident.

Three tests will sort a pile of moments in about twenty minutes. Run them in order rather than all at once. The first test eliminates more candidates than the other two combined, and it takes the least time to apply, so it saves you the most work.

The first test is the before-and-after.

Can you say, in one plain sentence, what was true at the start and what was true at the end? If you can’t, you don’t have a story. You have an anecdote, and anecdotes don’t sell anything.

People fail this test constantly and don’t notice. They have a vivid, interesting event with no change in it. It was memorable to live through and it goes nowhere on the page, because the reader finishes it and thinks “and?”

The second test is the specific detail.

Is there at least one concrete thing in this story that a person could photograph? A time on a clock, an object on a desk, a number in a spreadsheet, the place somebody was standing.

If the whole thing is made of feelings and conclusions, it will read as invented even when it’s true. That’s an unfair outcome, and it’s still what happens, because your reader has been lied to by marketers before and defends themselves accordingly.

The third test is the quiet connection to your offer.

Does this story point at what you sell without being about what you sell? A story that never touches your offer is entertainment. A story that’s obviously an advertisement gets skipped.

The sweet spot is a story where the reader arrives at your offer on their own. They read what happened, they feel the problem, and they’re already wondering what the answer is by the time you name it.

Running the three tests by hand gets slow once the pile is large. It’s a sorting job with clear rules attached to it. That’s exactly the kind of work you should hand off rather than grinding through it yourself on a Tuesday night.

Give the model the tests and the pile, and demand a verdict with reasoning on each one. Reasoning is what lets you catch a bad call, and there will be bad calls, because judgement about what your audience feels is still yours.

What you get back is a short list of stories that are ready to write.

Under it sits a maybe pile you’ll return to and a discard pile you can look at once and forget. That short list saves you the hours you’d have spent agonising over which idea to pick.

One more thing is worth checking before you commit to a story. Ask yourself whether you’d tell this to a friend at a table without being asked. If the honest answer is no, the reader will feel that same reluctance coming off the page.

Here are [number] raw story candidates from my own history and from my customers: [paste them]. Score each one against three tests. Test one, before and after: can you state in one sentence what was true at the start and what was true at the end? Test two, concrete detail: is there at least one physical, photographable detail in it? Test three, offer connection: does it point toward [describe what you sell] without being about it? Give each story a pass or fail on each test, with one sentence of reasoning. Then sort them into ready, maybe, and discard.

Don’t run these tests while you’re attached to a particular story. Attachment bends every verdict you make, and you’ll pass something that fails all three because you spent an hour thinking about it. Sort first, get attached afterward.

Read the discard pile once before you accept it. Somewhere in there will be a story you know is good and the tests rejected, and when that happens you’re usually right and the tests are usually missing context.

Don’t run these tests while you’re attached to a particular story.

Tests are a filter, not a judge. They exist to save you from spending three hours writing a story that was never going to work. They’re also wrong often enough that you should never obey them without reading the results yourself.

A story that passes all three tests can still land flat, and the reason is almost always the same.

The stakes are missing. You told what happened and left out why it was frightening, and without the fear there’s nothing for the reader to hold onto.

Stakes are simply what stood to be lost. Not in a dramatic sense, but in an ordinary one. The rent. The last shred of confidence you had. The belief that this whole thing was ever going to work.

Readers don’t lean in because something happened.

They lean in because something was at risk while it was happening. Take the risk out and the same events become a report, and reports get skimmed regardless of how well they’re written.

You leave the stakes out because you know the ending. You survived, the problem got solved, and from where you’re sitting the fear looks a little silly in hindsight. Your reader is standing where you were standing, and to them it doesn’t look silly at all.

So put the fear back in, and put it in early.

The reader needs to feel what was hanging in the balance before they hear how it turned out. Once they know the ending, they have no reason left to worry about any of it.

The second missing piece is the turn. Somewhere in the story there was a specific instant where things changed, and it’s almost never where you think it was. It’s rarely the moment of success. It’s usually the moment of realising.

That instant is what the reader remembers.

Everything before it’s setup and everything after it’s proof. If you can name the exact sentence where the thing turned, you have a story people will finish and repeat to somebody else.

Here’s a way to find it. Go through your material and look for the point where you stopped doing what you’d been doing. Not when it started working, but when you gave up on the old approach. That’s the turn, and that’s your centre.

There’s a last question worth asking about any story, and it’s the one that separates content that pulls buyers from content that gets polite compliments. Does the reader see themselves in this, or are they watching you?

Watching is pleasant and it sells nothing. A reader who admires your story goes away impressed with you. A reader who sees themselves in your story goes away worried about themselves, and worried people buy solutions.

The way you make them see themselves is by leaving space.

Describe the problem in enough detail that it’s clearly real, and describe the solution in a way that leaves the door open for them to walk through it.

Over-explaining the ending closes that door. When you spell out every step of how you fixed it, the reader files the whole thing as your story. They stop applying it to their own life, because those specifics belong to you and not to them.

They stop applying it to their own life, because those specifics belong to you and not to them.

The correction is to describe the change and hold back the recipe. Say what became different and let the reader arrive at the obvious question, which is how. That question is the reason they keep reading, and answering it too early throws it away.

You can see this happen in your own reading. The pieces you finish are the ones that made you want something before they explained anything. The pieces you abandon are the ones that handed you the full method in paragraph two, before you cared.

Here’s a story I want to use in my marketing: [paste your draft]. My readers are [describe them]. Answer four questions about it, one at a time. First, what was at risk in this story, and did I say it clearly enough for a reader to feel it? Second, where exactly does the story turn, and is that turn on the page or missing? Third, name three places where a reader might stop seeing themselves and start merely watching me. Fourth, tell me the one sentence I should cut because it explains too much.

That last question is the one you’ll want to argue with, and you should read the answer twice before you dismiss it. The sentence that explains too much is nearly always the sentence you were proudest of writing.

Everything up to here solves the problem once.

You dug, you sorted, you found a story, and you wrote it. Three weeks later you’re staring at the same blank page with the same empty feeling, and you have to start the whole process again.

The reason is that you dug and never stored anything. Excavation is expensive and it should only happen once for any given piece of material. What you need is a place to put the pieces so they’re waiting for you next time.

The place doesn’t need to be sophisticated.

A single document with a date and a few sentences per entry outperforms every elaborate system anybody has ever built, because you’ll keep opening it week after week.

Elaborate systems fail for a boring reason. They add friction to the moment of capture, and the moment of capture lasts about fifteen seconds. If storing a story takes longer than remembering it, you’ll stop storing them within two weeks.

So the rule is speed.

When something happens that made you feel anything, write four sentences about it before you lose the detail. What happened, what you felt, what was at risk, and one physical thing you can still see.

Those four sentences are enough to rebuild the whole story months later. Without them, you’ll remember that something interesting happened and be unable to recall a single specific about it, which is the same as not having it at all.

Add your customers’ material to the same document as it arrives.

A support email that made you wince goes in. A review with one perfect sentence in it goes in. A question that keeps repeating goes in, with a count next to it.

Within a couple of months, you’ll have more usable stories than you have places to put them, and the blank page stops being a problem you have. You’ll stop asking what to write about and start choosing which one to write first.

There’s one more habit worth building, and it costs almost nothing. Once a month, read the document from the top. You’re not looking for a story to write. You’re looking for two entries that turn out to be about the same thing.

Those pairs are where the strongest material comes from. A moment from your own history and a customer email from four months later, both circling the same problem, make a story neither one could carry alone.

That connection only becomes visible when the entries sit next to each other on a page.

It will never occur to you while the pieces are scattered across your memory, your inbox, and a notes app you stopped opening.

Give the file a plain name and put it where you already go every day. A story file living in a folder you open twice a year is the same as having no file. The point is that it sits there waiting on the day you’re in a hurry.

Here’s my story file, which contains moments from my own working life and material from my customers: [paste the whole file]. Read all of it. Find every pair or group of entries that are secretly about the same underlying problem, even when they look unrelated on the surface. For each group you find, name the shared problem in one plain sentence, and tell me which entry should open the story and which should close it.

The blank page was never a shortage of stories. It was a shortage of a place to look. The search kept coming back empty because you were hunting for drama, and your working life runs on small events instead.

Small events were always the right target. A cancelled subscription, a returned Saturday, a support email written at eleven at night by somebody who’d given up. Those are the things your reader is living through this week.

Those are the things your reader is living through this week.

Big events don’t connect the way you’d expect. A story about the year you tripled your income sounds impressive, and it lands on the reader as distance. They aren’t tripling anything this month, and they know it while they read you.

The small story closes that distance. It says you were sitting where they’re sitting, doing the unremarkable thing they’re doing, and something in that ordinary week turned. That’s a claim they can believe on a Tuesday.

When you tell them one of those and tell it with the fear left in and one real detail intact, they stop scrolling.

Not because you wrote something clever, but because you described their own life back to them more accurately than they’d have managed themselves.

That recognition is what buys you the right to sell them something. A reader who thinks you understand their problem will listen to your answer. A reader who thinks you’re performing will keep scrolling, and no amount of technique will stop them.

So the work is never really about writing.

It’s about looking squarely at what you’ve already been through and at what your buyers keep telling you. Then you need the nerve to put the unglamorous parts on the page where somebody can see them.

You have enough material for a year of content sitting in your support folder and your own memory right now. All that was missing was the willingness to go and get it, and a place to put it once you did.

Start today with one entry.

Pick the last thing that happened in your work that made you feel something, good or bad, and write the four sentences. Tomorrow the file won’t be empty, and next month it will be the first place you look.

A change story is the easiest video you’ll ever make, because it already has a beginning, a turn and an ending.

On video you don’t need the whole story. You need the moment just before the change, the turn itself, and the moment after. That’s three shots, and none of them needs your face.

Show the “before” as a calm generated scene: the cluttered desk, the unopened spreadsheet, the drawer that won’t shut. Say the turn in one line of voice-over. Then show the “after” using your real material, so nothing you claim depends on generated footage.

Here’s a story from my story file: [paste it]. Turn it into a 20-second video plan with three shots: the moment before the change, the turn, and the moment after. For each shot, give me one simple visual I could generate with an AI video tool (one subject, one action, no readable text) and one line of voice-over under 12 words. Keep every detail true to the story I gave you, and tell me which shot should use my own real photo or screen recording instead.