How this site sources, grades and corrects its numbers
Every statistic here is traced to where it actually came from, graded on what can be verified, and dated. This page says what that means in practice.
Most local SEO and local business statistics circulating on the internet trace back to a small number of original sources, and everything downstream is a paraphrase of a paraphrase. The point of this site's research isn't to add one more link to that chain. It's to find where the chain actually starts, say so plainly, and refuse to repeat a number that doesn't survive the trace.
That standard applies to every statistic, study and claim published here, including ones that would be more convincing left unexamined.
The grading system
Every statistic on this site carries a letter grade describing what a reader can actually verify about it — not whether it sounds plausible.
Cite freely
Primary source, dated, with sample and method published. Quote it, with its scope attached.
Cite with scope
Real research with a published sample, but narrower than the way it is usually quoted — self-reported, or one country, or one kind of business.
Cite with caution
Named source and plausible method, but the sample size or methodology is not published where the number is. Attribute it, do not lean on it.
Do not cite
Untraceable, undated, or superseded by later changes to the thing it measured. These are the numbers keeping the roundups alive.
Grading is deliberately unforgiving of missing information rather than generous toward reputable organisations. A statistic from a well-regarded research firm still earns a C if its sample size isn't published alongside the number — because the grade describes what's checkable, not who said it.
Source hierarchy
Not all evidence carries the same weight. This is the order used to decide what supports a claim.
- Primary sources — the organisation that actually collected the data.
- Official documentation — Google Search Central, Google Business Profile Help, Meta for Business, and equivalent platform documentation.
- Government and census data.
- Original research produced for this site — once the studies described in our research programme are running.
- Academic research.
- Reputable industry research — organisations like BrightLocal, Whitespark, Sterling Sky, Ahrefs, Semrush, Unbounce, LocaliQ, Pew Research and comparable named, dated, sampled sources.
- Named expert commentary — practitioners with a public, checkable track record on the specific claim.
- Community discussion — Reddit, forums and similar sources, used only for the language people use and the problems they describe, never as evidence for a statistic.
A claim supported only by level 8 doesn't appear on this site as a fact. It might appear as a description of what business owners say, clearly framed as exactly that.
What 'traced to source' actually means
For every published statistic, four things are checked before it appears on this site.
- Origin. Not the page that repeats the number — the organisation that produced it. Where a figure only traces to a roundup citing another roundup, it's excluded or listed as untraceable, never presented as fact.
- Date of the underlying data, not the publication date of whatever article currently repeats it. A statistic from a 2016 study is labelled as 2016, even when it's quoted in a "2026 statistics" article.
- Population, geography, sample size and method, stated as precisely as the original source discloses them — and stated as "not published" when it doesn't, rather than left implicit.
- Whether the claim has drifted. Where the same statistic is quoted with materially different numbers or framing across multiple sites — a sign it has been paraphrased away from its origin — it gets checked against the primary source directly rather than the most commonly repeated version.
On claiming something hasn't been studied
This is the single most misused phrase in content about original research, and this site handles it deliberately.
This site never states that "nobody has studied" a topic. What it states instead is: "no substantial study was identified in the sources reviewed on [date]," with the sources actually checked named or described.
The difference matters. The first claim is a statement about the entire world's research output, which almost nobody is in a position to verify. The second is a specific, falsifiable, dated claim about a real search — one a reader can check, dispute, or correct by pointing to something that was missed. Several gap claims on this site will turn out to be wrong as new research is found or published, and that is a feature of stating them this way, not a flaw.
Correlation, causation, and client data
Three specific rules that get enforced in the writing, not just stated here.
Correlation is never presented as causation. Findings from observational research — most of what exists in local SEO — are described with language like "associated with" or "tend to," and studies that are correlational say so in their own title where that's practical.
Client data is never presented as an industry benchmark without disclosed composition. When this site's own research draws on consented client data, the sample size, industry mix and any aggregation thresholds are stated plainly, and a small or narrow sample is labelled as exactly that rather than generalised.
Estimates are never presented as measurements. A modelled or estimated figure is labelled as an estimate. A survey finding is labelled as a survey finding, not "studies show."
Updates and corrections
Every research and statistics page states a last-verified date. This is what that date actually means.
Statistics pages are reviewed quarterly at minimum, and re-checked in full whenever the underlying source publishes a new edition. Fast-moving categories — anything involving AI search or AI adoption — are held to that same minimum but treated as needing it more urgently, because the underlying reality changes faster than a quarterly cycle can track.
When a figure is found to be wrong, outdated, or newly traceable to a better or worse source than previously known, the page is corrected and its verified date updated. Pages do not silently accumulate stale numbers between reviews.
If you've found an error, a better source, or a study this site missed, that correction is genuinely wanted — get in touch.
These grades apply to every page in the statistics archive. The same evidence standard governs how client work is written up, which is set out in the case study standards, and how the findings are turned into practical sequence across the guides.
The standards above, applied.
The statistics hub is where this methodology gets used on real numbers — including the ones that didn't survive the trace.
Replies usually go out within one working day.