Sicc Media · The Method

How we got
those numbers.

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.

Every number is one of three things

Nothing on your receipt comes from a single statistic. Each line is built from up to three layers, and they carry very different weight.

Tier 1

What we measured

Facts about your business, observed from the public internet. No citation needed. We looked, and this is what was there.

Tier 2

What we cited

Published research from the primary source, linked below so you can read it yourself. Never a blog quoting a study.

Tier 3

What we assumed

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.

Tier 1 · What we measured

All of this is public, free, and took under ten minutes. You can repeat every one of these yourself right now.

What we checkedHow
Load speed and Core Web Vitalspagespeed.web.dev, mobile tab
What the page looks like on a phoneLoaded it on a real phone, off wifi, and recorded it
Photo count, hours, categoriesYour public Google Business Profile
Review count and owner repliesCounted by hand from your public reviews
Date of your last postYour public social profile
Whether we could reach you at 9pmWe tried to book or request a quote
Map pack positionSearched your main service plus your town
What this means

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.

Tier 2 · What we cited

Three primary sources. Every figure below was re-verified on 18 September 2026 against the original publisher, not against somebody's summary of it.

Google, on site speed

MetricGoogle's "good" thresholdSource
LCP loadingUnder 2.5 secondsweb.dev
INP responsiveness200 milliseconds or lessweb.dev
CLS visual stability0.1 or lessweb.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.

How we use it

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.

BrightLocal, on reviews and profiles

Local Consumer Review Survey 2026, published February 2026, last updated 14 September 2026.

FindingFigure
Consumers who read reviews for local businesses97%
Who "always" read reviews when browsing41%, up from 29% last year
More likely to use a business after positive reviews85%
Put off by negative reviews77%
Who only care about reviews from the last three months74%
Who will not use a business with fewer than 20 reviews47%
Who require 4.5 stars or better31%, up from 17% last year
More likely to use a business that replies to every review80%
Unlikely to use a business that never replies42%
Put off by templated or generic replies50%
Who visit the business website after reading positive reviews54%
Using ChatGPT or similar for local recommendations45%, up from 6% last year
Who trust AI recommendations as much as reviews42%
Google's share of all local reviews71%, down from 83%

brightlocal.com, Local Consumer Review Survey

Baymard Institute, on carts and checkout

Only used when you sell online. Cart abandonment list, last updated 22 September 2025.

FindingFigure
Average documented cart abandonment, averaged across 50 studies70.22%
Abandoned because they were just browsing, not ready to buy42%
Of the remaining reasons, excluding browsers:
Extra costs too high, shipping, tax, fees40%
Delivery too slow20%
Did not trust the site with card details19%
Site required creating an account18%
Checkout too long or complicated17%
Average US checkout form elements, against an ideal of 12 to 1423.48

baymard.com, cart abandonment statistics

The trap we avoid

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.

Tier 3 · What we assumed

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.

Your figures, not ours

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.

What we never do

  • We never access your analytics, your sales, or your ad accounts. We could not if we wanted to. Everything we used is public.
  • We never cite a statistic we have not opened at the source and read ourselves.
  • We never present a projection as a measurement. Measured things and estimated things are marked differently on your receipt.
  • We never quote a figure without the year it was published.
  • We never use a number we cannot link. If we cannot find the primary source, the claim does not go in.
  • We never promise a result, a ranking, or a timeline.

A worked example, start to finish

One line, all the way down. This is a contractor.

Tier 1 · Measured

Their Google Business Profile has 9 owner-uploaded photos. No finished work visible anywhere. Screenshot taken and attached.

Tier 2 · Cited

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.

Tier 3 · Assumed

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.

Their number

They told us an average job runs about $1,400.

9 photos total$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.

What changed, and when

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.

DateChange
18 Sep 2026Baymard 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 2026Google Core Web Vitals thresholds re-verified. Unchanged at 2.5s, 200ms, 0.1.
18 Sep 2026BrightLocal LCRS 2026 re-verified against the published survey. All figures current. AI recommendation and review-recency findings added.
Aug 2026First published.

Found a mistake?

Then tell us and
we will fix it.

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.