Sicc Media · The Method
You got a receipt with dollar figures on it. Those are estimates. Here is exactly how each one was built, what it rests on, and where we guessed.
We publish this because almost nobody does, and because an estimate you cannot check is just a number somebody made up to make you nervous. If you find a mistake in here, tell us and we will fix it in public.
Nothing on your receipt comes from a single statistic. Each line is built from up to three layers, and they carry very different weight.
Facts about your business, observed from the public internet. No citation needed. We looked, and this is what was there.
Published research from the primary source, linked below so you can read it yourself. Never a blog quoting a study.
The one guess per line, always conservative, always stated out loud. This is the part you should argue with.
A line item with no Tier 1 in it is a line item about the internet, not about you. We do not send those.
All of this is public, free, and took under ten minutes. You can repeat every one of these yourself right now.
| What we checked | How |
|---|---|
| Load speed and Core Web Vitals | pagespeed.web.dev, mobile tab |
| What the page looks like on a phone | Loaded it on a real phone, off wifi, and recorded it |
| Photo count, hours, categories | Your public Google Business Profile |
| Review count and owner replies | Counted by hand from your public reviews |
| Date of your last post | Your public social profile |
| Whether we could reach you at 9pm | We tried to book or request a quote |
| Map pack position | Searched your main service plus your town |
These are not statistics. They are screenshots. If we say your profile has nine photos, that is because we counted nine photos, and we will send you the screenshot.
Three primary sources. Every figure below was re-verified on 18 September 2026 against the original publisher, not against somebody's summary of it.
| Metric | Google's "good" threshold | Source |
|---|---|---|
| LCP loading | Under 2.5 seconds | web.dev |
| INP responsiveness | 200 milliseconds or less | web.dev |
| CLS visual stability | 0.1 or less | web.dev |
Measured at the 75th percentile of real visits, split between mobile and desktop. So "good" means three quarters of your actual visitors got that experience or better.
We are not claiming Google says a slow site costs you a specific dollar amount. Google says nothing of the kind. We are saying: here is the published threshold, here is what yours measured. The comparison is factual. The money attached to it is ours, and it is a Tier 3 assumption.
Local Consumer Review Survey 2026, published February 2026, last updated 14 September 2026.
| Finding | Figure |
|---|---|
| Consumers who read reviews for local businesses | 97% |
| Who "always" read reviews when browsing | 41%, up from 29% last year |
| More likely to use a business after positive reviews | 85% |
| Put off by negative reviews | 77% |
| Who only care about reviews from the last three months | 74% |
| Who will not use a business with fewer than 20 reviews | 47% |
| Who require 4.5 stars or better | 31%, up from 17% last year |
| More likely to use a business that replies to every review | 80% |
| Unlikely to use a business that never replies | 42% |
| Put off by templated or generic replies | 50% |
| Who visit the business website after reading positive reviews | 54% |
| Using ChatGPT or similar for local recommendations | 45%, up from 6% last year |
| Who trust AI recommendations as much as reviews | 42% |
| Google's share of all local reviews | 71%, down from 83% |
brightlocal.com, Local Consumer Review Survey
Only used when you sell online. Cart abandonment list, last updated 22 September 2025.
| Finding | Figure |
|---|---|
| Average documented cart abandonment, averaged across 50 studies | 70.22% |
| Abandoned because they were just browsing, not ready to buy | 42% |
| Of the remaining reasons, excluding browsers: | |
| Extra costs too high, shipping, tax, fees | 40% |
| Delivery too slow | 20% |
| Did not trust the site with card details | 19% |
| Site required creating an account | 18% |
| Checkout too long or complicated | 17% |
| Average US checkout form elements, against an ideal of 12 to 14 | 23.48 |
baymard.com, cart abandonment statistics
These are percentages of people who abandoned for that reason. They are not a promise that fixing your checkout returns 18% more revenue. Anyone who applies them that way is inflating their own numbers. Where we use these, we apply a fraction of the figure and say so on the line.
Every line has exactly one assumption in it, and it is always the same shape: how often the thing we measured costs you a customer.
We keep that number deliberately low. If a gap plausibly costs you two or three jobs a month, we will estimate one. The point of the receipt is not to produce the biggest number we can defend. It is to produce a number you cannot dismiss.
If you told us your average job or ticket value, we used it. If you did not, we used a conservative placeholder and labelled it on the receipt. Tell us the real number and we will redo the whole thing, free, same day.
One line, all the way down. This is a contractor.
Their Google Business Profile has 9 owner-uploaded photos. No finished work visible anywhere. Screenshot taken and attached.
BrightLocal 2026 finds 97% of consumers read reviews and profiles for local businesses, and 85% are more likely to use a business after positive signals. 47% will not use a business with fewer than 20 reviews.
A profile with under ten photos and no proof of finished work converts fewer of the people who find it. We assume that costs roughly one job per month. That is the guess, and it is a low one.
They told us an average job runs about $1,400.
The photo count is a fact. The consumer behaviour is cited and linked. The job value came out of their mouth. The only thing we invented is "one job a month," it is written down, and it is obviously conservative. That is the whole method.
Research gets updated. A methodology page with stale numbers is worse than no methodology page, because it proves nobody has looked at it since it was written. So this is the log.
| Date | Change |
|---|---|
| 18 Sep 2026 | Baymard figures corrected. Cart abandonment over forced account creation moved from 24% to 18%, and abandonment over unexpected extra costs from 39% to 40%, against Baymard's current published list. Any receipt issued before this date used the older figures. |
| 18 Sep 2026 | Google Core Web Vitals thresholds re-verified. Unchanged at 2.5s, 200ms, 0.1. |
| 18 Sep 2026 | BrightLocal LCRS 2026 re-verified against the published survey. All figures current. AI recommendation and review-recency findings added. |
| Aug 2026 | First published. |
Found a mistake?
Seriously. If a number on your receipt is wrong, or you think an assumption is unfair, say so. We will rerun it and send you a corrected one. No call, no pitch, no catch.
Every figure on this page was verified against its primary source on 18 September 2026. All dollar amounts we publish are estimates built the way described above. Sicc Media, South Shore MA and Rhode Island.