Most teams calculate share of voice SEO the lazy way: they count how many of their tracked keywords rank on page one, turn it into a percentage, and call that their visibility share. It’s a comforting number and it’s almost always wrong. A first-place ranking on a keyword nobody searches counts exactly the same as a first-place ranking on the term that drives your entire category — which means the metric can climb while your actual traffic falls. Done properly, share of voice in SEO is a weighted measure of how much of the available search attention in a defined market you actually capture, and getting the weighting and the market boundary right is the whole exercise.
Why “Percentage of Keywords Ranked” Is the Wrong Number
The unweighted version fails on two fronts. First, it ignores demand: a keyword with 40,000 monthly searches and one with 20 searches are treated as equal votes, so you can pad your score by tracking easy, irrelevant long-tail terms you happen to rank for. Second, it ignores the click distribution: position 1 and position 9 both sit “on page one,” yet position 1 earns something in the range of a quarter to a third of all clicks while position 9 earns low single digits. Counting both as a win pretends a booked ranking and a barely-visible one are the same asset. They are not.
The result is a vanity metric that moves in the wrong direction. Rank a dozen new zero-volume terms and your keyword-count SOV rises even as a competitor quietly overtakes you on the five queries that actually pay. If a measurement can improve while the business outcome it’s supposed to track gets worse, it isn’t measuring the right thing.
What Share of Voice in SEO Actually Measures
Share of voice in SEO is your portion of the total organic visibility available across a fixed set of keywords, weighted by how much attention each keyword and each ranking position genuinely commands. Borrowed from advertising — where share of voice meant your slice of total category ad spend — the SEO version swaps ad dollars for search demand. The question it answers is precise: of all the clicks realistically up for grabs in my market, what fraction do my rankings put within reach?
That reframing matters because it makes SOV a competitive metric, not a self-report. Your click count tells you how you did. Your share of voice SEO score tells you how you did relative to everyone else competing for the same demand — which is the only version of “are we winning” that survives contact with a growing market. Traffic can rise while your share falls, if the category is expanding faster than you are.
The Formula That Respects Volume and Click-Through
A defensible visibility share calculation has three ingredients per keyword: search volume (the demand), your ranking position, and an estimated click-through rate for that position (how much of the demand that position captures). Combine them:
- Per-keyword visibility = search volume × estimated CTR at your position.
- Your total visibility = the sum of that figure across every keyword in the set.
- Total available visibility = the same sum computed as if one player held the top-capturing position on every keyword (or, better, the summed visibility of all tracked competitors).
- Share of voice = your total visibility ÷ total available visibility, expressed as a percentage.
The CTR curve is the piece to handle honestly. Public studies converge on a steep decline — position 1 captures a large minority of clicks, position 2 roughly half of that, and the tail beyond position 5 thins out fast — but the exact figures vary by query type, device, SERP features, and study. Use a sensible curve as a weighting model and treat it as an estimate, not a physical constant. What you must not do is invent precise per-position percentages and present them as measured fact; the shape of the curve is what makes the math meaningful, not spurious decimals.
A Worked Micro-Example
Say your market is three keywords: “team scheduling software” (10,000 searches), “shift planner app” (4,000), and “free rota maker” (1,000). You rank 2, 6, and 1. A competitor ranks 1, 3, and 12. Using a declining CTR weighting, your position-2 on the big term is worth far more than the competitor’s position-1 on the small one, but their position-1 and position-3 on the two largest terms likely still edge out your mix. Compute each keyword’s volume × CTR, sum both brands, and you might land near a 45%/55% split — a genuinely close race the keyword-count method would have hidden entirely (you rank on all three; so do they).
Now watch the leverage: move from position 6 to position 3 on “shift planner app” and your share jumps more than it would from picking up three new zero-volume rankings. The formula tells you exactly where a rank improvement converts into visibility share, which is the entire point of measuring it.
Defining the Keyword Universe: The Measurement Is the Choice
Here is the lever almost every SOV report quietly abuses. Your share of voice is only as meaningful as the keyword set it’s computed over, and that set is a decision, not a given. Track your branded terms and a handful of niches you already own, and you’ll report 85% share and feel invincible. Track the entire category including head terms owned by incumbents, and the same site reports 3%. Neither number is a lie; they’re answers to different questions.
The discipline is to fix a keyword universe that reflects the market you’re actually competing in — the non-brand, commercially relevant terms your real customers search — and then hold that set constant over time. Change the set and you’ve reset the baseline; a share that “grew” from 20% to 30% means nothing if you swapped in easier keywords between reports. Define the universe once, defend it against convenient edits, and let the number move only because your rankings moved.
Where the Numbers Come From: Ground Truth vs Estimate
Every share-of-voice score blends two very different data sources, and confusing them is how teams talk themselves into false confidence. Your own performance has a ground-truth record: Google Search Console reports the actual impressions and clicks your site earned, straight from Google. Your competitors’ positions and the keyword volumes, by contrast, come from third-party tools like Ahrefs — modeled estimates built from clickstream and index sampling, refreshed on a lag, accurate as direction rather than to the decimal.
So a competitor SOV figure is always part measurement, part model. That’s fine as long as you read it correctly: trust Google for your own slice, trust the third-party estimate for competitive direction, and never treat a 42.7% share as if the decimal is real. This data-trust hierarchy is exactly how SEO Rocket frames its numbers — GSC and GA4 as ground truth for your site, Ahrefs volume and position data as estimated competitive signal — so you weight each source by what it can actually know rather than reading them all as equally hard fact.
Share of Voice vs Traffic: Why They Diverge
Teams expect SOV and organic traffic to move together, then get confused when they don’t. They measure different things at different points. Traffic is a downstream outcome that depends on total market size, seasonality, SERP features stealing clicks, and conversion of impression to visit. Share of voice is a positional metric — your competitive standing on a fixed keyword set — that deliberately strips out market growth so you can see whether you’re gaining or losing ground independent of the tide.
That independence is the value. If traffic is up 10% but your share of voice is flat, the category grew and you merely kept pace — you didn’t win anything, the market handed it to you. If traffic is flat but share is up, you’re taking ground in a soft market and will pull ahead the moment demand returns. SOV is the leading indicator; traffic is the lagging confirmation.
Measuring It in Practice
You can assemble a share of voice SEO score by hand: export a fixed keyword set with volumes, pull your and your competitors’ current positions, apply a CTR curve, and compute the ratio in a spreadsheet. It works, but it’s fragile — someone has to refresh positions on a schedule, keep the keyword set locked, and resist the urge to swap terms. The practical path is a tool that tracks positions continuously and computes the weighted share for you.
SEO Rocket handles this as one workflow: AI keyword research on real Ahrefs data to build the universe, rank tracking that captures positions as trends rather than one-off spot checks, and competitor gap analysis to see which terms are moving your share. Because rankings jitter two or three places day to day for reasons that have nothing to do with your site, read share of voice as a trend line across weeks, not a reading you refresh and react to every morning.
Share of Voice in AI Answers
The SERP is no longer the only place visibility is won. AI Overviews, ChatGPT, Perplexity and other answer engines increasingly intercept the query before a user ever sees ten blue links, and being cited in those answers is becoming its own share-of-voice battle. The concept transfers directly: of the AI responses generated for your category’s questions, what fraction mention or cite your brand? It’s earlier and messier to measure than classic rankings, but the strategic question is identical — what slice of the available attention do you own?
This is why measuring only page-one positions is starting to undercount your real exposure, in both directions. SEO Rocket’s AI-visibility tracking exists to close that gap, surfacing where your brand shows up in AI-generated answers so your share-of-voice picture reflects how people actually search now, not only the classic results page.
Turning Share of Voice Into Decisions
A number you don’t act on is just a dashboard ornament. A share of voice SEO metric earns its place when it drives three specific moves. First, priority: the keywords where you hold a small share of high demand are your biggest upside — a rank gain there converts to more visibility than anywhere else. Second, defense: watch for terms where your share is slipping while a competitor’s climbs, and treat that as an early warning before the traffic loss shows up. Third, reporting: a single trended share figure tells a client “you’re taking ground in your market” far more honestly than a screenshot of ten rankings, which is exactly the kind of outcome-focused view a live client dashboard is built to show instead of a static emailed PDF.
The playbook proven across 1,000,000+ ranking pages never optimized for how many keywords ranked — it optimized for share of the demand that mattered, on a fixed set, tracked as a trend. That’s the difference between a metric that flatters you and one that tells you the truth.
Frequently Asked Questions
How is share of voice in SEO calculated?
Weight each keyword by its search volume and by an estimated click-through rate for your ranking position, sum that across a fixed keyword set to get your visibility, then divide by the total visibility available (or the summed visibility of all tracked competitors). The result is your percentage share. Avoid the unweighted “percent of keywords ranked” shortcut — it ignores demand and click distribution and can rise while your traffic falls.
What is a good share of voice percentage?
There’s no universal target, because the number is entirely dependent on how broadly you defined the keyword universe. A 60% share of a narrow niche and a 15% share of a broad competitive category can represent the same real business. What matters is the trend on a fixed keyword set: rising share against constant terms means you’re winning; the absolute figure only makes sense relative to your defined market and your named competitors.
Can I measure share of voice from Google Search Console alone?
Only partially. GSC gives you ground-truth impressions and clicks for your own site, which anchors your half of the calculation, but it can’t see competitors’ positions — so any true share-of-voice figure has to bring in third-party position and volume estimates for the rest of the field. Treat GSC as authoritative for your slice and the modeled competitor data as directional, and remember GSC itself carries a roughly two-day lag and hides rare queries.
