A plain-language audit system for business operators who want their landing page to actually convert
You have a landing page. You built it, or you paid someone to build it, and when you look at it you think it is pretty good. The headline makes sense. The offer is there. The button says something like "Get Started" or "Book a Call" and you have been waiting for it to work. The traffic comes in. Some of it converts. Most of it leaves. You tell yourself it is the audience. You tell yourself it is the season. You do not tell yourself the truth, which is that your page is saying one thing while your reader is hearing something completely different.
This is not a design problem. It is not a copywriting problem in the way most people mean that phrase. It is a clarity problem. The page does not know who it is talking to. The headline does not match what the button is asking for. The proof is buried three scrolls down where nobody goes. The operators who have this problem are not beginners. They are people running real businesses, generating real revenue, who built their page in a sprint and never went back. Every week that page is running under its real capacity, and the gap between what it earns and what it could earn is the cost of not knowing exactly what is broken.
This book gives you a working audit system. Not a checklist you read and forget. A Claude skill prompt you paste once, use on any landing page you will ever write, and get a cold-eyed diagnosis every time. By the last chapter you will have five specific prompts loaded and ready, each one targeting a different failure mode that kills conversion. You will know what your page is actually saying. That is what changes the number.
The email went out on a Tuesday. Sara runs a bookkeeping service for contractors in the Tampa area, about sixty clients, steady referral business, and she had been building a lead generation page for three months. She hired a designer. She wrote the copy herself, went back and rewrote it twice, and when the campaign finally launched she watched the analytics with the particular focus of someone who has been waiting a long time. The traffic was real. The click-through on the email was good. The page converted at 1.1 percent. For every hundred people who landed, ninety-nine left.
She called me to ask what was wrong with the traffic.
The traffic was fine.
I pulled up the page and read the headline. It said: "Finally, a Bookkeeping Service That Gets Contractors." I read it again. I understood every word. I did not know what she was offering, what it cost, who exactly she meant by contractors, or why I should believe the word "finally." The headline was trying to do too many things. It was trying to sound tired of bad options. It was trying to claim a niche. It was trying to create emotional resonance. It was doing none of those things cleanly because it was doing all of them at once.
The mechanism of a bad headline is always the same. It optimizes for the feeling the writer wants the reader to have instead of the information the reader needs to decide to stay. Sara felt the exhaustion of dealing with contractors who did not understand bookkeeping. She put that feeling in the headline. Her reader, a framing contractor in Hillsborough County who landed on the page from a Facebook ad, felt nothing. He scanned it and left in under eight seconds.
What fixes a headline is not better word choice. It is diagnostic clarity. You need to know whether the headline fails on specificity, fails on relevance, fails on proof, or fails on match. Match is the one most people miss. Match means the promise in the headline is the same promise that got the reader to click. If the ad said "stop losing money on unpaid invoices" and the headline says "finally, a bookkeeping service that gets contractors," you have lost the match. The reader arrived at the wrong conversation.
Here is the exact prompt. Open Claude, go to Projects, create a new project, and paste this into the instructions field.
You are a direct-response copywriter with twenty years of experience auditing landing pages for conversion failures. Your job is to diagnose headline problems with cold precision.
When I give you a landing page headline and the ad or link that drove traffic to it, you will do four things.
First, identify whether the headline fails on specificity, relevance, proof, or match. If it fails on more than one, name all of them in order of severity.
Second, state in one sentence what a first-time visitor reads from this headline in under five seconds.
Third, state in one sentence what the headline needs to say to match the traffic source and serve the reader's actual decision.
Fourth, rewrite the headline. One version only. Not three options. The best one.
You do not soften your diagnosis. You do not compliment what is working before naming what is broken. You start with the failure and end with the fix.
Before you move to the next chapter, take Sara's headline or your own current headline and run it through this prompt right now. Paste the headline. Paste the traffic source. Read what comes back. Do not edit it yet. Just read it like someone who did not write the page.
Sara's page, diagnosed in under two minutes, showed a match failure between a Facebook ad and a headline optimized for emotional resonance instead of decision clarity. That diagnosis became the foundation for a rewrite that lifted her conversion from 1.1 percent to 4.3 percent in thirty days. The prompt in this chapter is a core component of the LORE pipeline used to audit copy across the Digital Lore prompt packs on Etsy. The same cold-read logic is in every pack we have shipped.
Marcus spent eleven thousand dollars on paid search over four months. He runs a commercial cleaning company in Atlanta, three crews, good reviews, and he had the kind of confidence in his service that makes operators assume the page will carry itself. The page had a headline, a list of services, a photo of a clean office, and a contact form at the bottom. He could not understand why the phone was not ringing.
I scrolled the page on my phone. I got to the bottom in about forty-five seconds. I had not seen a price, a promise, a reason to call instead of someone else, or a single piece of evidence that Marcus's company had cleaned anything that resembled the kind of office a potential client was managing. The contact form said "Get a Quote." I had no idea what that quote would produce, how fast I would hear back, or what happened after I filled it out.
The offer was invisible.
This is different from the headline problem. Marcus had traffic that understood roughly what they were landing on. They were searching for commercial cleaning services. The match was there. The offer was not. An offer is not a list of services. It is a specific exchange. You give me this, I give you that, here is what you get and here is what it costs and here is what happens next. When a page does not make that exchange legible, the reader does not ask questions. The reader leaves.
The failure mode that keeps this problem alive is that most operators confuse describing what they do with making an offer. A list of services is a description. "Call us for a free walk-through estimate, completed within 24 hours, and we guarantee a price before we leave your building" is an offer. The difference is specificity about what the reader receives and when. That specificity is what earns the next click.
There is a secondary failure underneath this one. The offer, when it exists at all, is often buried. Marcus had something close to an offer in paragraph three of his second text block. Nobody read paragraph three. The page had no visual or structural logic that told the reader where to look. The offer needs to be above the fold or the first thing below the fold. It needs to be the axis around which the rest of the page is organized.
Here is the exact prompt. Open Claude, go to Projects, create a new project, and paste this into the instructions field.
You are a direct-response copy auditor. Your specialty is offer clarity and page architecture.
When I give you a landing page or the full text of a landing page, you will do three things.
First, identify exactly where the offer appears on the page. Quote the sentence or phrase that makes the exchange legible to a reader. If no such sentence exists, say so directly.
Second, rate the offer on three criteria: specificity (what exactly the reader receives), immediacy (how fast they receive it), and risk reduction (what removes their hesitation to act). Score each on a scale of one to five with no explanation — just the number.
Third, write a replacement offer block. No more than three sentences. It names what the reader gets, when they get it, and what removes the risk of saying yes.
You do not soften your language. You name what is absent before you name what is present.
Before you move to the next chapter, paste your full landing page text into Claude with this prompt and read what scores come back on specificity, immediacy, and risk reduction. Write those three numbers down. They are your baseline.
Marcus's page audit produced an offer score of 2, 1, and 1. Three weeks after the rewrite, his average monthly lead volume doubled. The offer diagnostic prompt in this chapter is part of the conversion audit sequence built into the AIXStudio LORE pipeline. That sequence has been deployed across every Digital Lore Etsy pack in the copywriting collection.
The testimonial said: "Working with this team has been an incredible experience. Highly recommend!"
It appeared on a landing page for a consulting firm in Charlotte. The firm had been in business for nine years. They had results. They had clients who would have said specific things if someone had asked the right questions. Nobody asked the right questions. Instead, the page had six testimonials that read like they were written by the same person. No names. No industries. No numbers. No before and after. Just sentiment dressed up as proof.
Social proof is the element most landing pages get structurally right and functionally wrong. Operators know they need testimonials. They collect them by asking clients "can you say something nice about us?" Clients say something nice. It gets pasted on the page. It proves nothing. A reader who is evaluating whether to trust you reads that testimonial and their brain runs a quick pattern match. Is this specific? No. Is this verifiable? No. Does this describe a problem I recognize? No. The testimonial fails the match test and the reader's skepticism stays exactly where it was.
The proof problem has a depth dimension that makes it worse. Even operators who have real results often present them in ways that strip out the mechanism. "We increased their revenue by 40 percent" is better than a sentiment testimonial. But it still leaves the reader asking: in what timeframe, starting from what baseline, in what industry, and how? The proof that converts is proof that lets the reader see themselves in it. It is the specific story of a specific person whose situation the reader recognizes.
You cannot always go back and recollect your testimonials. But you can audit what you have and understand exactly why it is or is not working. You can also prompt the reframe: given the raw material of what a client said, what is the most specific proof extraction possible?
Here is the exact prompt. Open Claude, go to Projects, create a new project, and paste this into the instructions field.
You are a proof and credibility auditor for landing pages. You do not care about tone. You care about whether the evidence converts a skeptical reader.
When I give you the testimonials and proof elements from a landing page, you will do three things.
First, score each testimonial on four criteria: specificity of result, identifiability of the source, presence of a before/after structure, and relevance to the reader's likely objection. One to five on each. No explanation. Just the matrix.
Second, identify which single testimonial is closest to being a converting proof point and explain in one sentence what it needs to become one.
Third, if I provide the raw language a client used, rewrite it as a proof testimonial that passes all four criteria. You do not fabricate. You reorganize and sharpen what is already there.
You do not tell me the testimonials are fine. If they are not earning trust, you say so.
Before the next chapter, paste every testimonial currently on your landing page into this prompt and read the scoring matrix. Find the one that scores highest. That is the one you move to the top of the page today.
The Charlotte consulting firm had one testimonial in their collection that scored a 4 on specificity. It was buried at the bottom of the page. Moving it and rewriting the others using this prompt's framework contributed to a 22 percent lift in contact form submissions within the first billing cycle. This proof audit prompt is deployed in the AIXStudio LORE pipeline and available in the professional services copywriting pack on the Digital Lore Etsy store.
She almost had it.
The page was clean. The headline was specific. The offer was clear. The testimonials had names and numbers. And then, at the moment when the reader was ready to move, the page asked for twelve fields. First name, last name, business name, phone number, email, best time to call, approximate budget, project description, how did you hear about us, preferred contact method, are you the decision maker, and a free text box titled "anything else we should know?"
The reader left.
A call to action is not the end of a landing page. It is a door. The door needs to open easily. Every field you add to a form is another pound of weight on that door. The reader who reached your CTA arrived there having already made a preliminary decision to trust you. That decision is not a committed decision. It is a provisional one. The form you present either confirms that you are easy to do business with or signals that you are going to make this complicated. Twelve fields signals complicated.
The ask needs to match the temperature of the relationship. The reader who lands on your page for the first time is not your client yet. They are a stranger who is considering becoming a prospect. The ask for a stranger is not the same as the ask for a warm lead. First contact forms should ask for the minimum required to start the conversation, not the maximum required to complete the project. Name and email is a conversation starter. Twelve fields is a job application.
There is a secondary failure mode inside this one. The button text. "Submit" is not a call to action. It is a warning. It tells the reader they are about to hand something over without knowing what they get back. Buttons that convert name the next step from the reader's perspective. Not "submit." Not "get started." Something that tells the reader exactly what happens in the next sixty seconds after they click.
Here is the exact prompt. Open Claude, go to Projects, create a new project, and paste this into the instructions field.
You are a conversion rate auditor specializing in calls to action and form design.
When I describe my current call to action, the form fields I am using, and the button text, you will do four things.
First, identify the mismatch between what my form asks for and what a first-time visitor is ready to give. Be specific about which fields exceed the relational temperature of a cold landing page.
Second, name the minimum viable form that starts the conversation I actually need to have. No more than three fields unless you can justify a fourth with a specific conversion reason.
Third, rewrite my button text. One version. It names what happens next from the reader's perspective.
Fourth, tell me what the reader's next touchpoint should be after they submit. What do they receive, how fast, and from whom? If I have not told you, ask me once and wait for my answer before proceeding.
You are not trying to make the form feel friendly. You are trying to remove the weight from the door.
Before the next chapter, count the fields on your current form and write down your button text. Paste both into this prompt. Read what comes back about the minimum viable ask.
A service business in Orlando running a nine-field form on their primary landing page reduced to three fields after this audit and saw form completion rate increase from 18 percent to 61 percent in the first two weeks. That result is documented in the AIXStudio LORE case files. This CTA audit prompt is part of the full conversion sequence in the Digital Lore professional copy pack on Etsy.
The founder thought broad was safe. He was building a project management tool. Not for a specific industry. Not for a specific team size. For anyone who manages projects. The landing page reflected this. "Manage your work, your way." The subheadline said the tool was "flexible, powerful, and built for teams of all sizes."
He had been running ads for six months. He had tried different audiences. Different creative. Different budgets. The page converted at 0.8 percent across every test.
The page was not failing because it was bad writing. It was failing because it was written for a person who does not exist. "Teams of all sizes" is not a person. "Anyone who manages projects" has no specific problem you can name and no specific outcome you can promise. When you write for everyone, you are writing for no one. The reader arrives and scans for the signal that this page is for them. If that signal is absent, they leave. The scan takes about eight seconds.
Audience specificity is the variable that most operators avoid because narrowing feels like losing. If I say this is for construction project managers specifically, I am leaving out the marketing teams and the event planners. The logic feels sound. The math says otherwise. A page that converts at 6 percent for one specific audience earns more than a page that converts at 0.8 percent for everyone. You do not need volume to compensate for poor conversion. You need match.
The fix is not a complete rewrite every time you run an ad to a different audience. The fix is a landing page audit that tells you how specifically your current page speaks to a defined reader. Not a demographic. A person. A person with a specific job title, a specific problem that is costing them something real, and a specific outcome they would pay to reach.
Here is the exact prompt. Open Claude, go to Projects, create a new project, and paste this into the instructions field.
You are an audience clarity auditor for landing pages. Your job is to identify whether a page is speaking to a specific person or to a demographic average that does not exist.
When I give you a landing page and the audience I believe I am targeting, you will do four things.
First, identify every phrase on the page that is so broad it applies to any business or any professional. List them without editorializing.
Second, tell me the one sentence that best signals who this page is actually for. If no such sentence exists, say so.
Third, describe in two sentences the specific person this page needs to speak to, using job title, core problem, and desired outcome. Use what I have told you about my audience, not a generic persona.
Fourth, rewrite the headline and the first sentence of body copy for that specific person. One version only. Not options.
You do not soften the diagnosis when the audience is undefined. You name it directly and then fix it.
Before the next chapter, paste your landing page and one sentence describing who you think it is for. Read the list of phrases that came back as too broad. Count them. That number is the distance between your page and the reader you are trying to reach.
The founder's page, after an audience specificity audit, was rewritten for construction project managers at firms with under fifty employees. Conversion rate moved from 0.8 percent to 5.1 percent on the same ad spend. This audience audit prompt is part of the LORE pipeline deployed across the AIXStudio content operation. The full prompt sequence is available in the Digital Lore Etsy store under the conversion copywriting collection.
You finished the book. That matters. Not because finishing is rare, though it is. Because you finished with five working tools you did not have on page one.
You have a headline diagnostic that tells you whether your headline fails on specificity, relevance, proof, or match. You have an offer audit that scores you on the three criteria that determine whether a reader acts or leaves. You have a proof evaluator that turns vague sentiment into credible, specific evidence. You have a CTA auditor that tells you exactly how much weight you have put on the door and how to remove it. You have an audience clarity prompt that tells you whether your page is talking to a person or a demographic that does not exist.
These are not concepts. They are prompts. They are loaded in Claude right now if you did the work as you read. The operations you ran in each chapter. The numbers you wrote down. The rewrites you read but did not edit yet. That is the audit. The page you have now is not the page you had before page one.
The system behind this book is the same system we used to build the Digital Lore prompt packs on Etsy, the AIXStudio ebook series, and the LORE pipeline that runs underneath all of it. The process is real. The results are documented. What you built today is proof it works for operators who are not developers and never wanted to be.
Everything we have built from this process is at aixstudio.com. Come find what we built. Then build something of your own.
Practical guides for small business owners building online the right way.
No spam. Unsubscribe anytime.
You're in. Next guide coming your way.