There is no verified, published weight hierarchy among these factors, and any specific percentage breakdown you encounter claiming otherwise is not something Google has confirmed. Google names exactly three local ranking factors in its own documentation: relevance, distance, and prominence (support.google.com/business/answer/7091). Category selection primarily feeds relevance, proximity is distance, and review signals are one component of prominence. But Google has explicitly declined to publish how these three factors are weighted against each other, and that weighting plausibly varies by query type, industry category, and market density, which is precisely why no single verified hierarchy can exist as a general truth.
Why the premise doesn’t hold
The three-factor framework is real and it’s the only officially named structure for local ranking. What doesn’t exist is a disclosed formula translating those three factors into a precise, universal weighting. Anyone presenting a specific hierarchy, such as “reviews count for 30%, proximity for 45%, category for 25%,” is presenting an estimate or a guess dressed up as verified fact. These numbers circulate in SEO industry content, sometimes derived from correlation studies run by third-party tools, but correlation-based studies measure what ranks alongside what, not causal weight, and none of them are Google-confirmed methodology.
The reason Google likely avoids publishing an exact hierarchy is the same reason it avoids publishing exact organic ranking weights: the system is dynamic, it’s evaluated per query, and a fixed public formula would be immediately gamed and would also misrepresent how the system actually behaves across the enormous variety of business categories and query intents it has to handle. A hair salon query in a dense urban core likely leans on different relative signal strength than a specialty equipment rental query in a rural market, simply because the competitive density and available signal richness differ.
This is also worth examining through the lens of what third-party local-pack correlation studies actually measure, since these are the studies most SEO tools and blogs cite as if they were closer to verified fact than they are. A correlation study of this kind typically samples some number of local pack results across a set of queries and markets, then measures which observable attributes (review count, proximity to centroid, category match, and so on) correlate with higher-ranking positions in that sample. This methodology has real, structural limits. Correlation across a sample doesn’t establish causal weighting, a factor can correlate strongly with top rankings simply because businesses that already rank well for other reasons also tend to accumulate more reviews over time, not because the reviews caused the ranking. The sample itself, the specific queries chosen, the markets included, the time window studied, shapes the result, which is exactly why different correlation studies run by different tools at different times routinely produce different, sometimes contradictory, rankings of factor importance. None of this makes such studies worthless as directional, exploratory signal, but it does mean they cannot responsibly be reported as a “verified weight,” since the methodology isn’t built to establish causation in the first place, and it isn’t Google’s own data.
Google’s public statements, in various Search Central office-hours sessions and documentation, have repeatedly reinforced that ranking factor importance varies by query rather than following one fixed global weighting, a pattern also consistently described for core organic ranking. There’s no reason to expect local ranking works differently in this respect, given it’s evaluated by systems that are themselves query-dependent and context-sensitive. Applying a single, universal hierarchy to every local search, regardless of category, market density, or query specificity, contradicts this repeatedly stated behavior, even though Google hasn’t produced a query-by-query breakdown of exactly how the weighting shifts.
A concrete contrast makes the context-dependence tangible. In a dense urban market with dozens of similarly-categorized, similarly-reviewed competitors clustered within a short radius of each other, distance differences between competitors may be small in absolute terms, which plausibly leaves more room for prominence and relevance differences to become the deciding factors between near-identical options. In a low-competition rural market where only one or two businesses of a given category exist within a reasonable radius, distance may functionally decide the outcome almost by default, simply because there’s no dense field of alternatives close enough to compete on relevance or prominence differences at all. Neither scenario proves a fixed hierarchy, they illustrate the opposite: the same three named factors produce different practical outcomes depending on the competitive and geographic context they’re operating in, which is precisely why no single fixed hierarchy could describe both situations accurately.
What you can say with confidence: category accuracy affects relevance, proximity to the calculated search point affects distance, and review count/rating/recency along with other authority indicators (citations, links, general online prominence) affect prominence. None of the three factors is disclosed as dominant across the board, and no factor is described as capable of fully overriding the other two in every case, since Google’s documentation explicitly frames all three as interacting rather than existing as a strict priority stack.
It’s also worth flagging a related but distinct methodological problem with how a lot of “ranking factor” content gets produced and shared. Some of it isn’t even a genuine correlation study, it’s an aggregation of practitioner opinion surveys, where a number of SEO professionals are asked to rank factors by perceived importance and the resulting average gets presented with the same visual authority as an empirical measurement. Opinion surveys of experienced practitioners have real value as a gauge of prevailing industry belief and shared practitioner experience, but that’s a fundamentally different kind of evidence than an actual measurement of ranking behavior, and the two get blended together in how this content typically gets presented, often without a clear disclosure of which type of study produced a given number. A practitioner reading a cited “ranking factor” percentage should ask not just what was sampled, but whether the number reflects measured search results at all, or reflects what other SEO professionals believe influences search results, since conflating the two is common and the difference matters enormously for how much weight the number deserves.
What to do about it
- Stop optimizing as if a precise formula exists to reverse-engineer. Instead, treat all three factors as levers you should improve in parallel, since neglecting any one of them (an inaccurate category, a location genuinely far from your target searches, or a thin review profile) creates a real weakness regardless of how the undisclosed weighting actually works.
- Be skeptical of any local SEO tool, course, or consultant presenting an exact percentage-based hierarchy as verified. Ask for the primary source. If it traces back to a third-party correlation study rather than Google’s own statements, treat it as a directional hypothesis, not a formula.
- Focus effort where you have real control: category precision, complete and accurate profile data, genuine review acquisition over time, and consistent citations. These improve all three named factors simultaneously rather than betting on one theorized weight over another.
- When diagnosing why a competitor outranks you, resist forcing the explanation into “it must be because they have more reviews” or “it must be proximity.” Check all three factors honestly, since the real answer is usually a combination, not a single dominant variable.
- When a correlation study is cited to justify a strategy, ask what queries and markets it sampled and over what time window, since a study’s conclusions are bounded by its sample and don’t necessarily generalize to your specific category, market density, or query type.
- Calibrate expectations by market context rather than importing a one-size strategy. A dense urban category with many close, similarly-matched competitors likely rewards differentiation on relevance and prominence more than in a sparse rural market where proximity alone may already be doing most of the work of deciding who’s even in contention.
The honest, defensible position for a practitioner is that three named factors exist and interact, no verified numeric hierarchy among them has ever been published, and building a strategy on an invented formula is optimizing for a number that doesn’t exist rather than for the real, interacting system Google has actually described.
