Weighted Share of Voice: The Real Scoreboard for Competitor Keyword Analysis

Rank tracking alone is a poor proxy for competitive strength. A competitor parked at position two across a handful of high-intent, high-volume terms can capture more qualified traffic than a site with dozens of position-five rankings. Comparing keyword rankings and share of voice means moving beyond a flat average position and building a weighted model that reflects how the SERP actually distributes attention. That model starts with keyword universe segmentation, then applies volume, intent, device, location, and SERP feature modifiers. Without those weights, you’re optimizing for a leaderboard that no longer exists.

The foundation is a clean keyword set. Pull competitor rankings from rank-tracking APIs, keyword gap platforms, and SERP scrapes. Enrich with Google autocomplete, People Also Ask, and competitor sitemap themes. Cluster terms by topic and intent rather than dumping them into one spreadsheet. Branded and non-branded queries should live in separate universes because branded share of voice is a retention metric, while non-branded share is acquisition. Informational, commercial, transactional, and navigational clusters each need their own CTR assumptions. A top-three ranking on a definition query is not equivalent to a top-three ranking on “best enterprise CRM for healthcare.“

Weighted share of voice is the percentage of estimated available clicks your domain earns across a keyword cluster. Calculate it by multiplying search volume by expected CTR for each ranking position, then applying a SERP feature multiplier. If an AI Overview, featured snippet, local pack, shopping carousel, or video block absorbs clicks, the organic CTR curve flattens. Your model should use observed CTR curves from Google Search Console where possible, supplemented by industry benchmarks for competitor-only terms. The result is a number that can be compared fairly: your weighted visibility divided by total weighted visibility in that cluster. It exposes where a competitor’s rank profile is actually dangerous, not merely prominent.

Comparing keyword rankings then becomes diagnostic. Track average position, but also top-three coverage, top-ten coverage, ranking volatility, and feature ownership. A competitor may have a worse average position yet dominate the commercial cluster because its top rankings sit on money terms. Segment by geography and device, especially for local and mobile-heavy niches. Pull data daily or weekly, then smooth with rolling twenty-eight-day windows to avoid overreacting to SERP fluctuations. Annotate algorithm updates, content launches, digital PR pushes, and internal linking changes so you can attribute share-of-voice movement to causes rather than coincidence.

SERP features deserve their own line items. If your keyword rankings live below an AI-generated answer, your effective share of voice may be half of what position data implies. The same applies to featured snippets, image packs, and review widgets. Track whether your brand is cited in AI Overviews, included in People Also Ask, or represented in local packs. For transactional queries, shopping and ads compress organic real estate further. A competitor with strong feed optimization and product schema can own the visual layer even when your organic URL ranks higher. Share of voice should reflect those layers or it will mislead budget decisions.

The most useful output is not a vanity dashboard but a prioritization engine. Map share-of-voice gaps to revenue potential using conversion rates by cluster. A five-point gain in a bottom-funnel comparison cluster may outperform a twenty-point gain in a broad informational hub. Identify striking-distance keywords where you rank four through fifteen and the competitor owns the top three. Those are often the fastest wins through content refreshes, internal links, schema, and on-page intent alignment. Meanwhile, defend clusters where your share of voice is slipping despite stable rankings; that pattern often signals SERP feature encroachment or competitor entity strength.

Avoid the common trap of treating every keyword equally or trusting a single tool’s opaque visibility score. Build a transparent model, document your CTR assumptions, and refresh them quarterly. Combine rank tracking with your own Search Console data, server-side conversion data, and manual SERP sampling for high-value terms. The goal is not perfect attribution. It is a repeatable competitive signal that tells you where attention is shifting, which rival is gaining, and which keyword clusters deserve the next sprint. When rankings and weighted share of voice disagree, trust the model that reflects clicks, intent, and SERP reality.