{"id":2094,"slug":"research-brief-smb-widget-market-difficulty-june-2026","title":"Research brief: SMB widget market difficulty — six ranked factors (June 2026)","kind":"reference","scope":"business","status":"current","audiences":["kevin","candid-team"],"topics":["incumbent-authority-accumulation","proximity-as-dominant-local-factor","serp-feature-crowding","content-saturation-per-query","review-revenue-causal-evidence","keyword-difficulty-tool-limits","smb-difficulty-widget"],"reference_body":"**Status:** Synthesised June 2026. Sister briefs: [[research-brief-smb-widget-capture-layer-june-2026]], [[research-brief-smb-widget-spend-benchmarks-june-2026]], [[research-brief-smb-widget-difficulty-to-work-june-2026]], [[research-brief-smb-widget-presenting-tiers-june-2026]], [[research-brief-smb-widget-vertical-difficulty-june-2026]]. Cluster entry point: [[research-cluster-smb-difficulty-widget-june-2026]].\n\n## TL;DR\n\n- The single most load-bearing factor in **national/organic** digital competitive difficulty is **incumbent entrenchment expressed through accumulated authority** (referring domains + brand search demand + age-of-page), because it is the slowest signal for a newcomer to replicate; for **local** businesses, **physical proximity to the searcher plus review prominence** dominate instead, and these are partly outside any operator's control.\n- Difficulty has risen structurally for everyone regardless of competitor strength because **SERP features (AI Overviews, ads, map packs) now consume most of the clickable real estate** — 58.5% of US Google searches (59.7% in the EU) ended without any click in 2024, and when an AI Overview appears, users click a traditional result in only 8% of visits versus 15% without one ([Pew Research Center](https://www.pewresearch.org/)).\n- \"Keyword difficulty\" scores from SEO vendors are useful **directional triage tools, not reliable measurements** — the same keyword can score wildly differently across tools because each uses a proprietary, commercially-incentivized formula, so they should never be the sole basis for a difficulty verdict.\n\n## Ranked: The Factors That Most Determine Digital Competitive Difficulty\n\nOrdered by how load-bearing each is — i.e., how much it determines whether a newcomer can realistically compete, and how hard it is to overcome.\n\n1. **Incumbent authority accumulation (links + brand demand + content history).** [Verified] This is the slowest-moving moat and the hardest to close. Referring-domain counts and brand search demand correlate strongly with rankings, and they compound over years. See [[incumbent-authority-the-deepest-moat]].\n2. **Proximity and review prominence (local markets only).** [Verified / Industry-consensus] In local search, physical distance from the searcher is the dominant factor and is largely fixed; review volume, recency, and rating are the strongest factors a business can actually move. See [[proximity-and-reviews-dominate-local]].\n3. **SERP feature crowding (AI Overviews, ads, packs).** [Verified] A market-wide difficulty multiplier independent of competitors: it shrinks the organic real estate available to win. See [[serp-features-difficulty-multiplier]].\n4. **Content saturation and query intent.** [Industry-consensus] The volume and quality of existing content for a query determines how much better a newcomer must be to displace incumbents. See [[content-saturation-per-query-not-niche]].\n5. **Trust/review moats (conversion, not just ranking).** [Verified] Even when a newcomer ranks, weak trust signals suppress the click and the conversion. See [[reviews-as-conversion-moat]].\n6. **Keyword/topic difficulty as measured by tools.** [Single-source / Directional] A triage signal, not a true difficulty measure. See [[keyword-difficulty-as-triage-only]].\n\n---\n\n## Detailed Findings\n\n### 1. Incumbent strength: how established competitors build barriers\n\n**(a) What the factors are and how they work.**\n\n- **Domain age is not itself a ranking factor; it is a proxy for accumulated signals.** [Verified] Google's John Mueller stated flatly in 2019 that \"domain age helps nothing.\" Multiple independent analyses agree that older domains rank better only because they have had more time to accumulate backlinks, content history, and brand recognition — not because of registration date. A clean aged domain confers a head start; a dormant or spam-laden aged domain does not (and Google's expired-domain-abuse spam policy actively penalizes the latter).\n- **Backlinks/referring domains are the most-studied and most durable barrier.** [Verified] Large correlation studies consistently find referring-domain count correlates with rankings more cleanly than raw backlink totals. **66.31% of all pages on the web have zero backlinks** ([Ahrefs](https://ahrefs.com/) analysis of ~1 billion pages in Content Explorer, which also found ~94% of pages get no Google search traffic — see [[ahrefs-backlink-distribution-tail]]) — meaning an incumbent with a diversified link profile sits in a small, hard-to-join club. In some verticals (legal, finance) a high domain-authority threshold functions as a near pay-to-play entry requirement.\n- **Brand/branded-search demand is an increasingly important — and very hard to fake — signal.** [Industry-consensus] Branded search volume is hard to manipulate (it reflects real awareness), which is precisely why search systems appear to weight it. Google rolled out a Branded Queries filter in Search Console in late 2025, signaling investment in measuring brand entities. Note: much of the strongest \"brand is now the top signal\" framing comes from vendor/SEO-agency correlation studies and should be treated as directional.\n- **Entrenchment via user-behavior signals (NavBoost).** [Verified] The 2023 [US DOJ v. Google](https://www.justice.gov/atr/case/us-and-plaintiff-states-v-google-llc) antitrust trial confirmed, under oath from Google VP Pandu Nayak, that a click-based re-ranking system called NavBoost is one of Google's important ranking signals, drawing on a rolling 13-month window of user-interaction data. This entrenches incumbents: established brands accumulate the \"good clicks\" that reinforce their position, and the system was deemed so central that the antitrust remedy ordered Google to share the data with qualified competitors. See [[navboost-doj-trial-confirmation]]. **Corroborated by the 2024 Google API documentation leak** ([[google-api-docs-leak-2024]]), independently reported by Rand Fishkin and Mike King.\n\n**(b) Evidence it matters.**\n\n- Ahrefs (vendor — see quarantine) found **72.9% of pages in Google's top 10 are more than three years old** and the average #1 page is five years old. Independent corroboration: SE Ranking (vendor) reports a majority of top-10 domains are 15+ years old. These are correlational, but consistent across datasets. See [[ahrefs-page-age-top-10]].\n- The trial testimony and the 2024 API-documentation leak (independently reported) are the strongest non-vendor evidence that accumulated user-behavior and link signals are load-bearing.\n\n**(c) Magnitude and how fast a newcomer can close the gap.**\n\n- **Realistic timeframe: 4-12 months to see meaningful traction, longer for competitive terms.** [Verified] This range is corroborated by Google's own former spokesperson Maile Ohye (\"four months to a year\"), Google's John Mueller, and multiple practitioner surveys. New domains commonly take 6-12+ months in competitive niches.\n- **The odds for new pages are low and have worsened.** [Single-source, vendor] Ahrefs' study found only **1.74% of newly published pages reach the top 10 within a year** (down from 5.7% in 2017); of pages that do reach the top 10, **~40.8% did so within one month** — implying early momentum matters (see [[ahrefs-40-percent-reach-top-10-within-month]]). Treat exact percentages cautiously (single vendor), but the direction (harder for new pages now) is consistent with other sources.\n- **Concession:** Incumbency is not destiny. Independent (non-vendor) analysis of **2,000 keywords by Adilo** found that in **66% of cases a lower-DA site outranked higher-DA competitors, and in over 30% of keywords a site 10+ DA points lower ranked #1**, driven by superior intent match, page-level relevance, and engagement. The lesson: **authority gaps are surmountable on narrow, well-matched, lower-competition queries, not on head terms.** See [[adilo-2000-keyword-low-da-outranks]].\n\n### 2. Content saturation\n\n**(a) What it is.** [Industry-consensus] Content saturation occurs when the supply of content for a query exceeds demand — Google has an abundance of good answers and does not need another generic guide. It is best understood as a per-query, per-intent condition rather than a whole-\"niche\" condition. See [[content-saturation-per-query-not-niche]].\n\n**(b) How to tell saturated from open.** There is no authoritative measurement, but practitioner-consensus diagnostics include:\n\n- **SERP composition.** [Industry-consensus] If the first page is dominated by large, entrenched brands (e.g., Investopedia/NerdWallet for finance) and the results all answer the query thoroughly, the space is saturated. If results are thin, off-intent, or dominated by forums/low-quality pages, there is an opening.\n- **Long-tail difficulty.** [Single-source] One practitioner reported even long-tail keywords in \"home workout equipment\" scoring difficulty above 65 — a directional sign of saturation. Treat the specific number as illustrative only.\n- **The \"content saturation index\" heuristic** (raw result counts: under 10,000 = room; over 1 million = saturated) is an old, weak heuristic and should not be relied upon. [Directional-Speculative]\n\n**(c) The honest caveat.** [Industry-consensus] Several credible practitioners argue saturation is overstated: semantic search means a newcomer can win on unaddressed sub-intents, question-based queries, and specific audience slices even in \"crowded\" spaces. The real test is whether profitable, winnable sub-queries remain — which is a query-level, not market-level, judgment.\n\n### 3. Local pack / map competition\n\n**(a) Google's stated framework.** [Verified] Google states local results rest on three pillars: **relevance, distance, and prominence.** This is confirmed in Google's own documentation.\n\n**(b) What actually drives competitiveness.** The most respected non-academic source is Whitespark's annual Local Search Ranking Factors survey (a Canadian firm, Edmonton, Alberta — relevant for the Canadian scope; note it is a vendor survey of practitioner opinion, not a controlled study — see [[whitespark-2026-local-ranking-factors]]). Its consensus, corroborated by Google's documentation and controlled tests:\n\n- **Primary GBP category** is repeatedly the single most influential controllable factor. [Industry-consensus]\n- **Proximity to the searcher** is dominant and largely fixed — a well-optimized business several miles away routinely loses to a weaker competitor that is closer, especially on mobile (which sends precise location). [Verified] See [[proximity-uncontrollable-local-factor]].\n- **Review signals (count, velocity, recency, rating)** are a top controllable factor and rising in importance year over year. [Industry-consensus] Whitespark practitioners now rank review recency among the most important factors; one documented case study ([Sterling Sky](https://sterlingsky.ca/) / Joy Hawkins) showed **rankings rising and falling in direct correlation with the presence or absence of a steady stream of new reviews** ([[sterling-sky-review-velocity-case-study]]).\n- **Keywords in the business name** confer a real (and somewhat exploitable) advantage — a controlled test (Joy Hawkins) found adding a service term to a business name moved it from unranked to position four ([[hawkins-business-name-keyword-test]]). [Single-source]\n- **GBP completeness and verification** are table stakes; unverified/suspended profiles lose visibility.\n- **Map filtering:** when multiple businesses share a category and location, Google filters out all but the strongest, so a weaker nearby competitor can be invisible at default zoom. [Industry-consensus]\n\nLocal-pack importance is itself worth flagging: SOCi (vendor) reports local-pack visibility drives ~126% more traffic and ~93% more actions than ranking below — [[spp-soci-126-percent-local-pack-traffic-uplift]] — quarantined as vendor, but directionally consistent with why the 3-pack slot is so valuable.\n\n**(c) How local difficulty differs from national/organic.**\n\n- Local difficulty is **geographically bounded** — you compete only with businesses near the searcher, not the whole web. This makes local markets more enterable for a small operator than national organic.\n- **Proximity caps optimization** — no amount of SEO overcomes being far from the searcher.\n- **Debunked tactics:** controlled studies (Sterling Sky's 9-week test of 441 keywords) found GBP posts produced zero ranking movement; geotagging photos and stuffing keywords in the business description have also been debunked. Google has confirmed the GBP description is not used in ranking. [Industry-consensus] See [[sterling-sky-controlled-tests-debunked-tactics]].\n- **Timeframe:** a new GBP in a moderately competitive category typically needs 3-6 months of consistent work; weeks in low-competition/rural areas; longer in contested urban categories (lawyers, dentists, locksmiths). [Industry-consensus]\n\n### 4. SERP feature crowding\n\nThis is the factor that has raised the difficulty bar for everyone, independent of direct competitors. See [[serp-features-difficulty-multiplier]].\n\n- **Zero-click is now the norm.** [Verified] [SparkToro](https://sparktoro.com/)/Datos analysis (Rand Fishkin) found **58.5% of US Google searches and 59.7% of EU searches ended without a click in 2024** — for every 1,000 US searches, only ~360 clicks reach the open web. A 2026 follow-up (Similarweb panel) put US zero-click at 68%. Caveat: the long-run trend stitches together different clickstream panels (Jumpshot, Datos, Similarweb), so cross-year comparisons are directional, but the direction is unambiguous. See [[sparktoro-datos-zero-click-2024]].\n- **AI Overviews sharply reduce clicks when present.** [Verified — multiple independent sources agree on direction, magnitude varies]\n  - **Pew Research Center** (the strongest non-vendor source; KnowledgePanel Digital browsing data of 900 US adults; the dataset contained **68,879 unique Google searches**, of which 12,593 triggered an AI summary, March 2025): when an AI summary appeared, **users clicked a traditional result in just 8% of visits versus 15% without — nearly half.** Clicking a link *inside* the summary occurred in just **1% of visits**. Sessions ended after **26%** of AI-summary pages versus **16%** without. About **18%** of searches produced an AI summary. See [[pew-2025-ai-overview-click-reduction]].\n  - Vendor/agency studies (quarantined) report larger CTR declines: Ahrefs ~58% drop for position-one content; Seer Interactive ~61% organic CTR drop on informational queries. These are directionally consistent with Pew but should not be blended as precise figures.\n- **AI Overviews are layering on top of ads, not replacing them.** [Single-source, vendor] [Semrush](https://www.semrush.com/)/Datos found Google Ads now appear at the bottom of ~25% of AI-Overview SERPs (up from under 1% in March 2025) — meaning organic gets squeezed by AI summary + ads simultaneously.\n- **Ads physically push organic below the fold on commercial queries.** [Industry-consensus] On commercial-intent searches, a four-pack of top ads can push organic results entirely below the fold; older analysis found a large share of users on small screens see no organic result at all without scrolling when four ads show. See [[ads-push-organic-below-fold]].\n- **Implication for difficulty tiers:** query intent determines feature load. Informational queries are most exposed to AI Overviews; commercial queries to ads; local queries to map packs. A market where the money queries are commercial or informational-with-AIO is structurally harder regardless of how weak the named competitors are.\n\n### 5. Keyword/search difficulty as a concept\n\n- **What it is.** [Industry-consensus] Keyword Difficulty (KD), also \"SEO difficulty,\" is a 0-100 score estimating how hard it is to rank on page one. Most tools derive it primarily from the link profiles (referring domains, domain/page authority) of the pages currently in the top 10.\n- **How it is measured — and why it is unreliable.** [Verified] Each vendor uses a different proprietary formula, so the same keyword scores differently across tools (e.g., KD 23 in one, 58 in another, \"medium\" in a third). Ahrefs weights referring domains heavily; Semrush uses a multi-factor formula including authority scores and SERP features; [Moz](https://moz.com/) leans on Page/Domain Authority of the top 10. Independent practitioner tests show large divergences and frequent 0/100 \"edge\" values that misrepresent real difficulty. See [[keyword-difficulty-as-triage-only]].\n- **The core limits.** [Verified / Industry-consensus]\n  - **Low predictive validity:** at least one practitioner (Omniscient Digital) notes KD has \"low to no predictive validity\" and that he found only one study correlating KD with performance, with weak statistics.\n  - **It ignores your domain.** A \"low KD\" keyword is still unwinnable if the top results are Wikipedia, a government site, or a dominant brand — link counts don't capture brand/entity authority. (Semrush's \"Personal Keyword Difficulty\" attempts to fix this by factoring your domain.)\n  - **It ignores intent and commercial value.** A KD-5 keyword with no buyer intent is worth less than a KD-40 keyword with qualified buyers.\n- **Verdict:** KD is a legitimate **triage/prioritization** signal for the widget — useful to separate \"obviously brutal\" from \"possibly winnable\" — but it must not be presented as a precise difficulty measurement, and the actual SERP must be inspected.\n\n### 6. Trust and reviews as moats\n\n**(a) Reviews as a ranking moat (local).** [Industry-consensus] Covered in §3: review count, velocity, recency, and rating are among the strongest controllable local ranking factors and are rising in weight. A competitor with hundreds of recent reviews and a steady inflow presents a moat a new entrant cannot close quickly because reviews accrue only at the rate of real customer transactions.\n\n**(b) Reviews as a conversion moat (all markets).** This is where the evidence is strongest and largely non-vendor/academic:\n\n- **Revenue impact (causal).** [Verified] Michael Luca's [Harvard Business School](https://www.hbs.edu/faculty/Pages/default.aspx) study (working paper using a regression-discontinuity design on Washington State Department of Revenue data, exploiting Yelp's rounding thresholds) found a **one-star increase in Yelp rating leads to a 5-9% increase in revenue**, and that the effect is driven by independent (non-chain) restaurants. This is one of the few causal — not merely correlational — findings in this whole domain. See [[luca-hbs-yelp-one-star-revenue]].\n- **Conversion impact.** [Verified] The [Northwestern University Medill Spiegel Research Center](https://spiegel.medill.northwestern.edu/) study (June 2017, with PowerReviews; research director Professor Edward Malthouse) found that **\"the purchase likelihood for a product with five reviews is 270% greater than the purchase likelihood of a product with no reviews.\"** The lift is larger for higher-priced/higher-consideration products: displaying reviews raised conversion **190% for a lower-priced product but 380% for a higher-priced product** — \"when the price is higher, there is more risk involved in the consumer's decision.\" See [[spiegel-2017-five-reviews-270-percent-lift]].\n- **The \"too good to be true\" effect.** [Verified] The same Spiegel study found **purchase likelihood typically peaks at ratings in the ~4.0-4.7 range and declines as ratings approach a perfect 5.0** — \"products with an average star rating in the 4.7-5.0 range are less likely to be purchased than those in the 4.2-4.7 range… shoppers see ratings at the far end of the spectrum as 'too good to be true.'\" Consistent with the finding that many shoppers actively seek out negative reviews to gauge credibility. (Note: the primary PDF uses both \"4.0-4.7\" in its main text and \"4.2-4.5\" in its summary table.) See [[spiegel-too-good-to-be-true-effect]].\n- **Consumer reliance.** [Industry-consensus] BrightLocal's annual Local Consumer Review Survey (vendor, but the standard reference) finds **just 4% of consumers say they \"never\" read online reviews**, with a 2025 nuance: trust in reviews \"as much as personal recommendations\" has fallen from **79% in 2020 to 42% in 2025**, and consumers are growing slightly more lenient on review volume and slightly less swayed by raw star count, favoring detailed, recent, credible reviews. Google review usage was 83% in 2025 (down from 87% in 2023). See [[consumer-rating-thresholds-rising-brightlocal]] in sister brief [[research-brief-smb-widget-vertical-difficulty-june-2026]].\n\n**(c) Why reviews are a particularly durable moat.** Unlike content or links, which a well-funded newcomer can produce quickly, **review count and velocity are rate-limited by actual customer volume** — you cannot legitimately buy them, and incentivizing customers violates Google's guidelines ([[ftc-fake-review-rule-and-google-policies]]). A 500-review incumbent with a 4.5 rating and steady inflow is therefore one of the hardest competitive positions for an SMB to attack.\n\n---\n\n## Quarantined Vendor Sources (commercial incentive — used for concepts/direction only, not as authoritative numbers)\n\nEvery quantitative dataset cited above comes from a commercial vendor. None is neutral, academic, or governmental. Treat all figures as **indicative, not precise.**\n\n- **Ahrefs** — sells SEO tools. Source of the \"1.74% of new pages rank in a year,\" \"average #1 page is 5 years old / 72.9% are 3+ years,\" \"66.31% of pages have zero backlinks,\" and \"AI Overviews cut clicks 58%\" figures. Incentive: content marketing for its subscription tools; its KD and Domain Rating are proprietary metrics. Used here as directional only.\n- **Semrush** (owns [Search Engine Land](https://searchengineland.com/) and Datos) — sells SEO tools. Source of AI-Overview-plus-ads (25% of AIO SERPs) and KD methodology claims; publishes self-promoting \"our KD is most accurate\" studies. Quarantined for proprietary metrics.\n- **Moz** — sells SEO tools; originator of \"Domain Authority,\" a proprietary metric Google does not use. Quarantined.\n- **SE Ranking, Keywords Everywhere, SpyFu, Serpstat, Mangools, Ubersuggest** — SEO tool vendors; KD and domain-age stats are proprietary/self-serving.\n- **Whitespark, BrightLocal, Uberall, Sterling Sky, GatherUp** — local-SEO software/agencies. Their ranking-factor surveys and consumer surveys are the field's standard references but reflect practitioner opinion and have commercial incentives. Whitespark (Edmonton, Alberta) and BrightLocal are used here because they are the most-cited and partially corroborated by Google documentation and controlled tests — but flagged.\n- **Seer Interactive, NP Digital, various SEO agencies** — sell services; their CTR studies are directionally useful but self-interested.\n- **SparkToro** — sells audience-research software; the zero-click studies rely on third-party clickstream panels (Datos/Similarweb) and Fishkin is transparent about methodology limits. Treated as high-quality-but-commercial.\n\n**Non-vendor / higher-trust sources used:** Pew Research Center (browsing-data study), Michael Luca / Harvard Business School (peer-reviewed-grade Yelp study), Northwestern Spiegel Research Center (academic), US DOJ v. Google trial testimony, Google's own documentation and spokespeople (Mueller, Ohye, Nayak), and peer-reviewed review-impact literature (NIH/PMC, Chevalier & Mayzlin, Luca & Zervas), plus the 2024 Google API documentation leak ([[google-api-docs-leak-2024]]) which independently corroborates the trial testimony.\n\nConsolidated quarantine: [[caveats-market-difficulty-vendor-quarantine]].\n\n---\n\n## What a Non-Expert Owner Can Observe vs. What Needs Expert Tools\n\n**Plausibly self-observable (the widget can ask the owner to look):**\n\n- Whether the first page of results for their core query is dominated by big national brands vs. small local players. [Owner can eyeball]\n- Whether an AI Overview, a four-pack of ads, and/or a map pack appear for their money queries — and how far down the first real organic result sits. [Owner can eyeball]\n- Competitors' **review counts, star ratings, and recency of newest review** on Google — all publicly visible. [Owner can observe directly]\n- Whether top local competitors' Google Business Profiles are complete (photos, categories, hours, responses to reviews). [Owner can observe]\n- Their own and competitors' approximate **proximity** to the town/city center and to dense customer areas. [Owner can judge]\n- Whether people already search for competitors by name (brand demand) — observable via Google autocomplete and \"people also search for.\" [Owner can roughly gauge]\n\n**Requires expert tools / data:**\n\n- Backlink and referring-domain counts and quality (requires Ahrefs/Semrush/Moz-type crawlers). [Expert]\n- Domain/page authority scores and keyword difficulty scores. [Expert — and themselves unreliable]\n- Actual branded-search volume (Search Console / keyword tools). [Expert]\n- Share of queries triggering AI Overviews across a keyword set, and CTR impact. [Expert]\n- Grid-based local rank tracking (how rankings vary by precise location). [Expert]\n- Whether a competitor's authority comes from clean vs. toxic links. [Expert]\n\nThis split is the architectural backbone of the widget's input layer — see sister brief [[research-brief-smb-widget-capture-layer-june-2026]] for the full design of the \"CAN vs CANNOT answer\" question split.\n\n---\n\n## Recommendations (for building the difficulty-tier widget)\n\n1. **Weight the tiers by market type first.** Branch the logic: if the business is local/service-based, lead with proximity + review prominence + GBP completeness. If it competes nationally/organically, lead with incumbent authority + content saturation + SERP-feature load. The factors are not interchangeable across these two worlds. Reinforces the two-axis tier model in sister brief [[research-brief-smb-widget-vertical-difficulty-june-2026]].\n2. **Make SERP-feature crowding a market-wide multiplier, not a per-competitor input.** Have the owner report what appears on their top three money queries (AI Overview? ads? map pack? where does organic start?). This is observable, high-signal, and rising in importance.\n3. **Use review gaps as the most actionable difficulty axis.** Ask the owner to compare their review count/recency to the top three competitors. This is fully self-observable, causally linked to revenue (Luca: 5-9% per star) and conversion (Spiegel: up to 270%), and rate-limited — making it both a difficulty signal and an action item.\n4. **Treat keyword-difficulty scores as one directional input, clearly labeled as an estimate.** If the widget ingests any vendor KD score, show it as a rough band and pair it with a \"look at who actually ranks\" prompt. Never output a precise difficulty number from KD alone.\n5. **Set expectations on timeframe explicitly.** Bake the 4-12-month (national) and 3-6-month (local) realistic-traction windows into the output so owners interpret a \"hard\" tier as \"long climb,\" not \"impossible.\"\n6. **Thresholds that should change the tier:**\n   - **Drop a tier easier** if: money queries show no AI Overview/ads, top results are weak/off-intent, and competitors have <30-50 reviews with stale recency.\n   - **Raise a tier harder** if: top results are entrenched national brands (national) or competitors have hundreds of recent reviews and complete GBPs (local), AND AI Overview + ads dominate the money queries.\n\n## Caveats\n\n- **Correlation vs. causation pervades this field.** Almost all ranking-factor evidence (links, age, brand) is correlational. The rare causal exceptions are the review-impact studies (Luca's regression-discontinuity Yelp study; the Spiegel conversion experiments) and the antitrust-confirmed existence of NavBoost.\n- **Survivorship/selection bias is rampant.** SEO case studies and \"we outranked a 2015 competitor in 8 months\" stories feature only winners; the far larger number of new sites that never ranked are invisible. The Ahrefs 1.74% figure is the closest thing to a base rate and it is sobering.\n- **Vendor incentive bias.** A large share of all published data in this domain comes from companies selling SEO tools or services, who benefit from portraying difficulty as both measurable (buy our tool) and surmountable (hire us). Numbers have been quarantined accordingly.\n- **The ground is shifting fast.** AI Overviews, AI Mode, and zero-click behavior are changing quarter to quarter; any difficulty model must be re-baselined regularly. Google's own claims that AI Overviews keep click volume \"relatively stable\" conflict with independent publisher data and Pew — an unresolved dispute to flag.\n- **Geographic scope.** Most data is US-centric; Whitespark (Edmonton, Alberta) and BrightLocal provide some Canadian-relevant local-search grounding, but vertical- and country-specific difficulty can differ materially from these aggregates.\n- **Tool numbers are not measurements.** Keyword Difficulty, Domain Authority, and Domain Rating are proprietary estimates, not Google signals; Google does not use them.\n\nConsolidated caveats: [[caveats-market-difficulty-vendor-quarantine]].\n","rationale_body":"Compiled June 2026. Determines the FACTOR WEIGHTS the widget uses to compute its difficulty tier. Tells the algorithm what to ask about and how much each input matters.","metadata":null,"links":{"outgoing":[{"slug":"incumbent-authority-the-deepest-moat","title":"Factor 1 — Incumbent authority accumulation (the deepest moat)","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"ahrefs-backlink-distribution-tail","title":"Ahrefs — 66.31% of web pages have ZERO backlinks; ~94% get no Google traffic (vendor)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"ahrefs-page-age-top-10","title":"Ahrefs — average #1 ranking page is 5 years old; 72.9% of top 10 is 3+ years","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"navboost-doj-trial-confirmation","title":"NavBoost — Google's 13-month click-based re-ranking, confirmed under oath in DOJ v. Google 2023","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"proximity-and-reviews-dominate-local","title":"Factor 2 — Proximity + review prominence dominate LOCAL markets","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"serp-features-difficulty-multiplier","title":"Factor 3 — SERP feature crowding (AI Overviews, ads, packs) — a market-wide difficulty multiplier","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"sparktoro-datos-zero-click-2024","title":"SparkToro/Datos 2024 — 58.5% US / 59.7% EU Google searches end with zero click","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"pew-2025-ai-overview-click-reduction","title":"Pew Research 2025 — AI Overviews roughly halve traditional clicks (non-vendor)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"ads-push-organic-below-fold","title":"On commercial queries, four ads can push organic entirely below the fold","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"content-saturation-per-query-not-niche","title":"Factor 4 — Content saturation is per-query, NOT per-niche","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"reviews-as-conversion-moat","title":"Factor 5 — Trust/review moats (conversion, not just ranking)","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"luca-hbs-yelp-one-star-revenue","title":"Luca HBS WP 12-016 — one-star Yelp = 5-9% revenue (regression-discontinuity, the causal anchor)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"spiegel-2017-five-reviews-270-percent-lift","title":"Spiegel 2017 (Northwestern) — 5 reviews = 270% purchase likelihood lift; 380% for higher-priced","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"spiegel-too-good-to-be-true-effect","title":"Spiegel — purchase likelihood PEAKS at 4.0-4.7 stars and DECLINES toward 5.0 (too-good-to-be-true)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"keyword-difficulty-as-triage-only","title":"Factor 6 — Keyword difficulty as triage only, not a measurement","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"caveats-market-difficulty-vendor-quarantine","title":"Caveats — vendor quarantine and the correlational-vs-causal divide","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"research-brief-smb-widget-vertical-difficulty-june-2026","title":"Research brief: SMB widget vertical difficulty — two-axis tiering by industry (June 2026)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"ahrefs-40-percent-reach-top-10-within-month","title":"Ahrefs — of pages that DO reach top 10, ~40.8% did so within 1 month (early momentum matters)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"adilo-2000-keyword-low-da-outranks","title":"Adilo — 66% of cases lower-DA site outranks higher-DA (incumbency not destiny, on narrow queries)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"sterling-sky-review-velocity-case-study","title":"Sterling Sky / Joy Hawkins — rankings rise and fall in direct correlation with review velocity","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"hawkins-business-name-keyword-test","title":"Joy Hawkins controlled test — adding service term to GBP name moved listing from unranked to #4","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"google-api-docs-leak-2024","title":"Google Search API documentation leak — March 2024 corroborates NavBoost-style click signals","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"spp-soci-126-percent-local-pack-traffic-uplift","title":"SOCi — local-pack visibility drives ~126% more traffic and ~93% more actions vs below","kind":"reference","scope":"business","link_type":"relates-to"}],"incoming":[{"slug":"research-cluster-smb-difficulty-widget-june-2026","title":"Research cluster: SMB digital-difficulty self-assessment widget (six briefs, June 2026)","kind":"reference","scope":"business","link_type":"depends-on"},{"slug":"research-brief-smb-widget-capture-layer-june-2026","title":"Research brief: SMB widget capture layer — what owners can vs cannot self-report (June 2026)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"research-brief-smb-widget-spend-benchmarks-june-2026","title":"Research brief: SMB widget spend benchmarks — feasibility of a \"digital-minus-ads\" % of revenue (June 2026)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"research-brief-smb-widget-difficulty-to-work-june-2026","title":"Research brief: SMB widget difficulty-to-work mapping — three tiers of work for three sizes of gap (June 2026)","kind":"reference","scope":"business","link_type":"relates-to"},{"slug":"research-brief-smb-widget-presenting-tiers-june-2026","title":"Research brief: SMB widget presentation layer — tiered results without overclaiming (June 2026)","kind":"reference","scope":"business","link_type":"relates-to"}]},"created_at":"2026-06-23T19:16:04.161Z","updated_at":"2026-06-23T20:13:54.645Z"}