SEO BANDWAGON

Interactive · updated with every weekly sweep

The Periodic Table of SEO Ranking Factors

The famous table every agency has on its wall was editorial: experts weighting factors by consensus. This one is empirical. We test ranking factors against real search data — 162,244 businesses across 54 US metros, re-measured weekly — and stamp each element with what we found, including the nulls. 8 of 46 factors measured, 2 tracked live, 1 in pilot. The rest is folklore until tested.

ContentHTMLArchitectureLinksTrustSocialPersonalLocal / Map PackAI / GEOViolations

Select an element to see what we measured — the stat, the sample size, and the method — or that nobody has measured it at all.

Every measured finding

The full text of each measured element, in plain HTML — no clicking required.

UGC platforms in SERPs (Su) — OBSERVED
reddit.com net −73 positions across 53 metros in one week, p = 0.0026. The "Reddit is taking over search" era is receding in local results: reddit.com was the only major UGC domain moving with a statistically significant direction in our sweep — and the direction is down. It declined in 28 of the 37 metros where it moved at all (76 top-10 positions lost vs 40 gained). Yelp, Facebook, Tripadvisor and Instagram were indistinguishable from coin flips. (SEO Bandwagon weekly local-SERP sweep, Aug 2026; n = 53)
Review count (Rc) — NULL
r = −0.021 (n = 8,384); review leader averages position 10.06. Review count does not predict map-pack rank. The most-reviewed business in a market ranks 10th on average, and top-3 businesses out-review their nearest rivals in only 47.5% of head-to-head matchups — worse than a coin flip. Reviews behave as a visibility threshold (~75 reviews), not a ladder. (SEO Bandwagon Map-Pack Ranking Study, Jul 2026; n = 8,384)
Star rating (Rr) — NULL
r = −0.033; Bayesian-adjusted r = −0.036; top-3 average 4.55 vs 4.54 for near-misses. Star rating does not separate winners from losers in the map pack. Top-3 businesses average 4.55 stars against 4.54 for those ranked 4–10. Even a Bayesian-adjusted rating that blends stars with review volume stays null. (SEO Bandwagon Map-Pack Ranking Study, Jul 2026; n = 8,384)
Keyword in business name (Kn) — HOLDS
28.8% vs 13.7% top-3 rate (n = 59 vs 729, z = 2.5, p ≈ 0.01) — specialist terms only. A business with the search term in its name reaches the top 3 at twice the rate of one without — but only for specialist services. "Leak detection" in the name is worth +6.5 positions; on broad terms like "plumber" the same test gives 16.0% vs 14.4%, nothing. That is the signature of relevance matching, not a name bonus. (SEO Bandwagon Surface-Relevance Study, Aug 2026; n = 788)
Surface coverage (Sc) — HOLDS
top-3 rate by surfaces matched: 0 → 13.8%, 1 → 18.4%, 2 → 31.6% (monotonic). The more indexable surfaces a keyword appears on (business name, category, domain), the more likely the business ranks top-3 for it — monotonic, with two surfaces already 2.3× zero. Partial result: only 3 of 8 surfaces are collected; review text and page content are pending, and self-selection is the confound the full study is designed to separate. (SEO Bandwagon Surface-Relevance Study, Aug 2026; n = 788)
Website presence (Wp) — NULL
r = −0.020; 10.4% of top-3 businesses have no website vs 11.4% of near-misses. Merely having a website does not predict map-pack rank. The no-website rate among top-3 businesses is statistically identical to the group ranked just below them. (SEO Bandwagon Map-Pack Ranking Study, Jul 2026; n = 8,384)
Proximity (Px) — NULL
r = 0.008 (distance from market center). Distance from the market center showed no correlation with rank — but read this one narrowly: we collected from a single point per metro, so this measures coarse centrality, not the searcher-position proximity effect, which is real and which our method cannot see. (SEO Bandwagon Map-Pack Ranking Study, Jul 2026; n = 8,384)
Category breadth (Cb) — NULL
r = −0.075 — the strongest signal in a 13-signal battery, still noise. How many categories a business lists under was the strongest of thirteen candidate signals we tested against rank — and at −0.075 it is still far too weak to act on. Every visible listing attribute we tested came back null. (SEO Bandwagon Map-Pack Ranking Study, Jul 2026; n = 8,384)
Rank stability (Rs) — OBSERVED
tracked live — see the weekly numbers on this element. Map-pack rank is not a stable property of a business. We re-measure the same listings every week; the retention and movement numbers on this element update with each sweep. (SEO Bandwagon weekly local-SERP sweep, Aug 2026; n = 67,607) Live this week: 78.4% of top-3 businesses held their spot this week (67,607 paired listings) (Aug 16, 2026 → Aug 23, 2026).
Reviews by rank band (Rd) — OBSERVED
tracked live — see the weekly numbers on this element. Businesses that rank higher do have more reviews on median — a selection effect worth seeing next to the null correlation on the Review Count element: the bands differ, yet within a market, review count still does not pick the winners. (SEO Bandwagon weekly local-SERP sweep, Aug 2026; n = 73,284) Live this week: median 318 reviews in the top 3 vs 176 at rank 21+ (sweep of Aug 23, 2026).
Entity presence (Ae) — Pilot
pilot: knowledge panel present for 17/21; owning the #1 result for their own name: 3/21. In a 21-business pilot of branded SERPs, most businesses had a knowledge panel, but only 3 of 21 owned the #1 result for a search of their own name. Pilot scale — a direction to test, not a finding. (SEO Bandwagon weekly local-SERP sweep, Aug 2026; n = 21)

Where the numbers come from

The measurements behind this table come from our own collection programs: a 30-metro, 42,599-listing map-pack study (July 2026), a per-service-term rank study across 5 metros and 8 service keywords, and a weekly sweep of 54 metros that re-measures the same businesses every week — the live elements on this page refresh from it automatically. Full distributions are published in our Google Business Profile Benchmarks. We publish aggregates only, never listing-level data. A verdict of Null means we tested the factor and found no association — which is a result, not an absence of one. Not measured means exactly that; we do not weight factors by opinion.

Cite or link this table

SEO Bandwagon, "The Periodic Table of SEO Ranking Factors", week of Aug 23, 2026. https://seobandwagon.com/periodic-table-of-seo

Free to reference with attribution. The numbers change as new studies land and the weekly sweep runs — the URL stays stable.

Embed this table on your site

The full interactive table, free to embed. It stays current automatically — the live elements refresh from our weekly sweep. Please keep the source line under it.

<iframe src="https://seobandwagon.com/embed/periodic-table-of-seo"
  title="The Periodic Table of SEO Ranking Factors" width="100%" height="980" loading="lazy"
  style="border:0;border-radius:8px"></iframe>
<p>Source: <a href="https://seobandwagon.com/periodic-table-of-seo">The Periodic Table of SEO Ranking Factors</a> by SEO Bandwagon</p>