The Disturb Report comes in two parts, and they work differently. The AI Site Score grades your website with the same automated checks every time. The AI Operations Audit looks at how your business runs after a lead reaches out, and a person reviews every one. This page explains both, plus the standard I build to and the research behind every number I publish.
- Two checks, one report. Part 1 grades your website. Part 2 grades what happens after a lead reaches out.
- Part 1 is automated.Same checks every run, scored 0 to 100. Conversion counts 40%, Design 30%, Function 30%, plus Google’s own Lighthouse numbers.
- Findings cite their sources. Every conversion gap names the Hormozi principle behind it, so you can check my work.
- Part 2 is human. Ten yes-or-no questions. A person writes the report and sends it back within 2 business days.
- I build to the same rubric I grade with. The checklist I score your site against is the checklist my builds have to pass.
- Every number I publish shows its work. Source, year, study size, and the caveats, all listed at the bottom of this page.
Part 1: The AI Site Score
The AI Site Score is a website audit. I run your site through the same checks every time and turn the result into one number from 0 to 100. The score has three parts: three pillars I grade myself, the technical scores Google’s own tools report, and a separate look at how your offer reads to a cold visitor.
How I read the site
Before I score anything, I pull a clean copy of your site. I use Firecrawl to grab the homepage at desktop and phone widths, including the screenshot and the underlying text. That gives me the same view a real visitor gets, plus a structured copy my detectors can read. Everything else on this page runs against that snapshot.
The three pillars
Every site is graded on three things. Each pillar is scored from 0 to 100, and your overall score is a weighted blend. Conversion carries the most weight because it is the most direct lever on revenue.
- Conversion. Does your homepage actually sell? I check whether your value proposition is clear, your headline does work, your call-to-action is specific, you show social proof above the fold, and you handle risk reversal, urgency, and reason-why. A site that looks fine and works fine but does not convince visitors to act loses the lead at the moment of decision. This is where the biggest losses for small businesses sit.
- Design. Does the site look like a business worth trusting? I check layout, typography, images, mobile fit, brand consistency, and the trust signals a visitor reads before a single word. People decide whether to trust a business in seconds, and most of that judgment is visual.
- Function. Does the site actually work? I check that the phone number taps to call, the email is a mailto link, forms submit somewhere, links are not broken, pages load fast, SSL is valid, and the layout holds up on a phone. A site that looks fine but does not work still loses the lead.
Hormozi-cited findings
Most audit tools flag a problem and leave you to take their word for it. I do not. Every gap I flag on the Conversion pillar cites a specific principle from Alex Hormozi’s $100M Offers and $100M Leads: the dream outcome, the perceived likelihood, the value equation, the reveal-the-problem lead magnet, and so on. So the rubric is not my opinion. You can read the source, see the citation, and judge for yourself whether the gap is real.
Offer strength
Alongside the three pillars, I score your offer separately. Two lenses:
- Value equation (Hormozi). Dream outcome × perceived likelihood, divided by time delay × effort and sacrifice. I grade each of the four levers from your homepage and call out the weakest one: the lever most likely costing you conversions right now.
- Awareness level (Eugene Schwartz). Five stages, from a visitor who has never heard of the problem to a visitor who is ready to buy from you specifically. Copy written for the wrong stage misses on purpose. I tell you the stage your homepage is targeting so you know whether it matches the visitors you actually get.
AI originality check
I run your homepage copy through an originality detector. If the writing reads like it was generated by ChatGPT or another large language model and pasted in unedited, I flag the specific lines verbatim in the report. Generic AI-written copy converts poorly, reads cold to visitors, and is starting to get penalized by search engines. Seeing the flagged lines tells you exactly what to rewrite in your own voice.
What Google sees
Alongside the three pillars, I show the scores from Google Lighthouse, the same tool Google uses to measure websites. These are technical, and most visitors never see them, but your search ranking depends on them.
- Speed. How fast the site loads. Slow sites lose visitors before they read a word, and Google ranks them lower.
- Accessibility. Whether everyone can use the site, including people on older phones or with disabilities.
- Technical health. Whether the site follows current web standards for security and code quality.
- SEO. Whether Google can read and understand your pages well enough to index them correctly.
I also report Core Web Vitals, the set of numbers Google uses to measure how fast and stable the site feels to a real visitor on a real device.
How the grade works
The overall score maps to a letter grade, so the result is easy to read at a glance.
How long an audit takes
A full audit takes about thirty to sixty seconds. The slowest step is Google Lighthouse, which the report streams in once it lands. You can start reading the Conversion and Design sections before the technical scores finish.
A snapshot, not a live feed
I score your site as it looked on the day I ran the audit. Websites change, so I cap the report at seven days. After that, the site may have moved on, and the scores should be treated as history, not a current read. A fresh audit takes only a few minutes.
What the score is not
The AI Site Score is an honest read, not a promise. It is not a Google ranking, and it does not predict where you will land in search results. It does not guarantee leads or sales. What it does is measure whether your website follows the practices that help a small business win work, and point to the gaps most likely costing you. What you do with that is the real opportunity.
Want to see where your own site lands? The score takes about a minute.
Part 2: The AI Operations Audit
A website can be graded by a machine in a minute. How you handle leads, follow-up, booking, and billing cannot. So Part 2 works differently: ten yes-or-no questions about what happens after someone calls, books, pays, or goes quiet. Missed-call follow-up, response speed, online booking, review requests, invoicing, after-hours capture.
There is no instant AI dump on the other side. A person reads your answers against how your kind of business runs, writes the report by hand, and sends it back within 2 business days. Gaps come back named, worst first, in plain English, with what fixing each one looks like. Honest answers get an honest report, and if your operations check out, the report says that too.
How I build
I build to the same standard I grade by. The rubric above is not just how I judge your old site, it is the checklist your new one has to pass. Every site I ship is built to:
- Load fast and work on a phone first. Most visitors arrive on a phone, and a slow page loses them before they read a word.
- Say one clear thing and ask for one clear action. A single value proposition and one primary call to action above the fold, so a cold visitor knows what you do and what to do next in seconds.
- Use real photos and visible trust signals. Real work, real people, and proof a stranger can believe over stock imagery.
- Make contact effortless. One-tap call and email, and a short, low-friction form that asks only what it must, behind a spam firewall so the real leads reach you.
- Read clearly to search engines. Clean structure, LocalBusiness markup, and the technical fundamentals Google measures, so the foundation is set right.
- Sell, not just describe.Copy written with conversion principles, not filler, aimed at the visitor’s actual decision.
What I do not do is promise a Google ranking or a number of leads. I build the foundation that gives you the best shot at both. The rest is the work you put in on top of it.
This is the same checklist I run against your current site. See how it holds up.
The numbers behind my claims
When I put a statistic in front of you, I tell you where it came from. Every figure below names its source, its year, and how big the study was, with a link so you can read it yourself. Where a number is older, narrow, or a widely-repeated industry estimate rather than a single controlled study, I say so. That candor is the point. It is also the answer to a fair question about my own Site Score: a number is only worth trusting if it shows its work.
First impressions form in about 50 milliseconds
The claim, as I word it:“People judge your entire business in just 50 milliseconds.”
The figure: visitors form a lasting first impression of a web page in roughly 50 milliseconds, and that snap judgment colors how they rate everything else on the page (the halo effect).
Source: Lindgaard, Fernandes, Dudek & Brown, Behaviour & Information Technology, 2006. Participants rated the visual appeal of pages shown for just 50ms.
How it applies: the first thing a visitor reacts to is your design, so I build the top of the page to win that half-second.
Worth noting: a 2006 study, foundational and widely replicated since.
Fixing only the top of a page took one client from 1% to 5%
The claim, as I word it:“I have watched one page go from 1% to 5% conversions by fixing nothing but the top of it.”
The figure: a roughly five-fold lift in conversion from reworking the above-the-fold section alone.
Source: a landing-page teardown drawing on more than 1,500 pages audited (practitioner case study).
How it applies: the first screen is the highest-leverage real estate on the page, which is why I spend the most effort there.
Worth noting: a single practitioner case, not a controlled study. Treat it as a directional example, not a guarantee.
The typical page converts about 2%, the best about 11%
The claim, as I word it:“The average website turns 2 of every 100 visitors into a customer. The best turn 11.”
The figure: a median conversion rate of 2.35%, with the top 10% of pages converting at 11.45% or higher (top 25% at 5.31%). I round to 2 and 11 for plain reading.
Source: WordStream, analysis of thousands of advertiser accounts representing about $3 billion in annual ad spend.
How it applies: the gap between a typical page and a top one is the room a better-built page can capture.
Worth noting: these are paid-traffic landing pages, and the dataset is older. Use it as a directional benchmark, not a promise for every site or traffic source.
Three form fields is the sweet spot
The claim, as I word it:“Trimming a form from 4 fields to 3 lifts completions by 50%.”
The figure: across more than 40,000 landing pages, forms with three fields converted best, and conversion fell as fields were added. The often-quoted summary puts the jump from four fields to three at close to 50%.
Source: HubSpot, analysis of 40,000+ customer landing pages.
How it applies: I ask only what I must on a lead form, because every extra field costs you completions.
Worth noting:HubSpot’s own data clearly shows fewer fields convert better with three as the sweet spot. The precise “50% from four to three” is the popular summary figure, directional rather than a single controlled result.
More than half of web traffic is automated
The claim, as I word it:“51% of all web traffic now bots.”
The figure: in 2024, automated traffic reached 51% of all web traffic, passing human traffic for the first time in a decade. The share that is actively malicious (bad bots) was 37%.
Source: Imperva (a Thales company), 2025 Bad Bot Report, based on 13 trillion blocked bot requests across 2024.
How it applies: a contact form with no defense fills with junk that buries the real leads. My 8-layer firewall stops it.
Worth noting: the 51% counts all automated traffic, good and bad. The portion that is malicious is 37%.
Reply within 5 minutes and you are 21x more likely to win
The claim, as I word it:“Answer within 5 minutes and you are 21x more likely to win them than waiting 30.”
The figure: contacting a new lead within 5 minutes, versus 30 minutes, made a business about 21 times more likely to qualify that lead. Most buyers go with whoever responds first.
Source: Prof. James Oldroyd (MIT) and the Lead Response Management study, a dataset of 2.24 million leads, reported in Harvard Business Review.
How it applies: instant auto-reply plus an instant alert make you the first responder every time, even mid-job.
Worth noting: the original study is from the early 2010s, and the response-time effect has been re-confirmed many times since.
About 6 in 10 visitors are on a phone
The claim, as I word it:“6 of every 10 people hit your site from a phone.”
The figure: roughly 59 to 60% of global web traffic comes from mobile phones.
Source: StatCounter and Statista, 2025.
How it applies: I build mobile-first, with one-tap call and email.
Worth noting: the exact share varies by industry and region. About 60% is the global mixed figure.
53% leave a page that takes longer than 3 seconds
The claim, as I word it:“53% of visitors leave a page that takes longer than 3 seconds to load.”
The figure: 53% of mobile site visits are abandoned if a page takes longer than 3 seconds to load.
Source:Google, “The Need for Mobile Speed,” 2016.
How it applies: I build for fast load, because speed is the difference between a visitor who stays and one who is gone before the page paints.
Worth noting: a 2016 figure, specific to mobile. Still the most-cited speed and abandonment benchmark, and directionally unchanged since.
Around 30,000 sites are hacked daily, and 43% of attacks hit small business
The claim, as I word it:“Around 30,000 sites are hacked every day, and 43% of attacks hit small businesses.”
The figure: an estimated 30,000 websites are hacked each day worldwide, and about 43% of cyberattacks target small businesses.
Source:Astra Security (aggregated hacking statistics); the 43%-target-small-business figure traces back to Accenture’s cybercrime research.
How it applies: managed hosting, security, and monitoring keep your site online and protected.
Worth noting:these are widely-cited industry estimates rather than a single controlled study. The “30,000 a day” figure is repeated across many sources without one definitive origin.
8 in 10 people search online for local businesses every week
The claim, as I word it:“8 in 10 people look online for local businesses every week.”
The figure: 80% of US consumers search online for local businesses at least once a week, and 32% search daily.
Source:SOCi Consumer Behavior Index, as compiled in BrightLocal’s local SEO statistics roundup.
How it applies: if your business has no website, the search still happens. It just ends at whatever is out there instead of at you.
Worth noting: a US consumer survey figure. The exact share varies by industry, but every study in this space puts weekly local search well past half of consumers.
62% would avoid a business over wrong info online
The claim, as I word it:“62% say they’d avoid a business when the info they find is wrong.”
The figure: 62% of consumers would avoid using a business if they found incorrect information about it online.
Source:BrightLocal, Local Business Discovery & Trust Report, 2023.
How it applies: without a website you control, your hours, phone number, and services live in directories and old listings you never see. When those are wrong, most people quietly move on.
Worth noting: the 2021 edition of the same study put this at 63%, so the figure has held steady across years.
Want to see what your own business scores? Two quick questions point you at the right check.