What Is Average Position in Google Search Console? (And How to Improve It)

Average position is one of the most-checked metrics in Google Search Console — and one of the most misunderstood. If you've ever wondered why your "average position 3.5" page isn't getting the clicks you expected, or what counts as a "good" average position, this guide explains everything.

What Is Average Position in Google Search Console?

Average position in Google Search Console is the mean ranking of your page or website across all search queries and result pages where it appeared during the selected date range.

It's calculated simply:

Average Position = Sum of all ranking positions / Number of queries where you appeared

For example, if your page appeared:

Position 2 for "keyword A"
Position 5 for "keyword B"
Position 8 for "keyword C"

Your average position = (2 + 5 + 8) / 3 = 5.0

You can view average position in GSC under Performance → Search Results, then add the "Average position" column.

What Is a Good Average Position in Google Search Console?

| Average Position | What It Means | CTR Expectation | |---|---|---| | 1–3 | Top of page 1 | 15–30%+ | | 4–10 | Page 1 (lower half) | 2–10% | | 11–20 | Page 2 | 0.5–2% | | 21–50 | Pages 3–5 | Under 1% | | 50+ | Deep pages | Near 0% |

A good average position is generally considered to be under 10 (page 1). Position 1–3 is excellent. Positions 4–10 are competitive and still valuable. Below position 11 (page 2+), organic click-through rates drop dramatically — most studies show fewer than 1% of searchers click past page 1.

The GSC average for sites with any meaningful traffic is typically between 15–40 as new content mixes with established rankings.

Why Average Position Can Be Misleading

Average position is an average — and averages can deceive. Here's what to watch out for:

1. Queries with wildly different positions pull the average

If you rank #1 for 50 low-volume queries and #50 for one high-volume query, your average will look worse than reality for your core terms. Filter by specific pages or queries to get a true picture.

2. Position 1 doesn't always look the same

Google shows different SERP features (Featured Snippets, People Also Ask, local packs, image carousels) above regular organic results. Your "position 1" listing may actually appear below several SERP features and have lower click-through than expected.

3. Average position changes when you filter dates

If you compare two periods, make sure the query set is the same. Google often starts showing pages for new queries over time, which can pull average position down even as your actual rankings improve on existing queries.

4. Device breakdowns matter

Your average position on mobile vs. desktop can differ significantly, especially for local queries. Always check the Device filter in GSC Performance.

How to Find Average Position for Specific Pages or Queries

In Google Search Console:

Go to Performance → Search Results
Make sure Average position is checked in the "Graph" section
To analyze specific pages: click Pages tab, find your URL
To analyze specific queries: click Queries tab, find your keyword
To filter both: click a page first → then click Queries to see which queries drive that page's rankings

Pro tip: Use the Compare feature (date range dropdown) to track whether your average position is improving or declining over time.

What Is a Good Average Position Improvement Target?

The most valuable position improvements to chase are:

| Current Position | Target | Why It Matters | |---|---|---| | 11–20 (page 2) | Move to 1–10 | Page 2 → page 1 is the biggest CTR jump in SEO | | 4–10 (lower page 1) | Move to 1–3 | Top 3 captures 3–5x more clicks than positions 4–10 | | 1–3 | Maintain + optimize CTR | At the top, focus shifts to meta descriptions & title click-through |

These are called striking distance keywords — pages that are close to a ranking breakthrough. Moving a position-11 page to position 7 can increase traffic from that query by 5x or more.

5 Ways to Improve Average Position in Google Search Console

1. Identify and optimize striking distance pages

Filter GSC for queries ranking position 5–20 with high impressions. These pages are almost on page 1 — small content improvements can push them over. Add a new section addressing the query intent more directly, update the title tag, and strengthen internal links pointing to the page.

2. Update content freshness signals

Google favors recently updated content for many queries. Pages ranking position 8–15 that haven't been updated in 6+ months often benefit from:

Adding a "Last Updated: [month, year]" tag
Adding a new section on recent developments
Refreshing statistics and examples with current data

3. Improve page experience signals

Pages with poor Core Web Vitals (slow LCP, high CLS) rank lower than comparable pages with good performance. Check GSC's Page Experience → Core Web Vitals report for any flagged URLs, then fix them to potentially improve rankings.

4. Build internal links to underperforming pages

Pages ranking 11–20 often lack internal link equity. Find your top-traffic pages and add contextual links from them to the underperforming page. Even 2–3 strong internal links from high-authority pages on your site can meaningfully boost rankings.

5. Improve title tags and meta descriptions for CTR

Higher click-through rate (CTR) is a ranking signal. If your page ranks position 8 but has an above-average CTR for that position, Google may boost it to position 5–6. Rewrite title tags to:

Front-load the primary keyword
Add a specific benefit or number ("5 ways to…", "Complete guide to…")
Match the searcher's exact intent

Average Position vs. Impressions vs. Clicks — What to Optimize First?

| If you have... | Optimize for... | |---|---| | High impressions, low position (20+) | Ranking improvement — content depth, links | | High impressions, good position (1–10), low CTR | CTR — title/meta description | | Good position, good CTR, low impressions | Keyword expansion — target more related queries | | Low impressions overall | Indexing — ensure pages are indexed; add more content |

How Google Actually Calculates Average Position (The Detail Nobody Reads)

The one-line definition — "sum of positions divided by number of queries where you appeared" — is directionally right and quietly wrong in the places that matter for decision-making. If you build a weekly review process on the one-liner, you will misdiagnose real problems and chase phantom ones. Four details change everything:

1. Position is per impression, not per query. When your URL appears in the SERP, GSC records the topmost position among your ranked results for that page-and-query combination for that specific impression. A user who reloads and sees you at position 4 the second time generates a second impression at position 4. Averages across impressions, not across users and not across unique queries.

2. The "topmost" rule masks a second listing. If your URL appears twice for a query (organic + sitelinks, or two of your pages both rank), GSC records only the highest position for that URL. A page that ranks 3 and 18 for the same query looks like a clean position-3 result — and you never see the 18. This is why average-position changes sometimes have no explanation in your content: a lower duplicate ranking dropped out and the aggregation shifted, not the visible rank.

3. SERP features count as positions. The infamous rule change from 2019 is still tripping teams in 2026. A featured snippet at position 1 is position 1. A People Also Ask carousel that expands? Still position 1 as a whole block. A four-item image pack that pushes your organic listing down the page counts as four positions above you. The number "3.5" on your dashboard may correspond to a listing that appears below the fold on mobile.

4. Impression counting is permissive. GSC counts an impression when your listing appears anywhere in the SERP, including when it is below the fold and the user never scrolls. This inflates impressions and, by extension, drags average position toward the middle. A page that "ranks 5" for a term where nobody scrolls past position 3 is not actually the fifth-most-seen listing — it is barely seen at all. This matters when you interpret impression changes: rising impressions with flat clicks often means you gained visibility in below-the-fold or SERP-feature-buried positions, not that demand shifted.

Take these together and the rule becomes: average position is a directional indicator of where you sit on the page, not a ranking scoreboard. Use it to spot movement, then drill into the query-level detail to understand cause.

The Six Buckets of Average Position Change (And What Actually Causes Each)

Every meaningful average-position change falls into one of six buckets. Different bucket, different fix. Confusing them wastes weeks.

Bucket 1 — True ranking shift

Your URL moved up or down for the same query set. This is what most SEOs think they are looking at, and it is usually 20-30% of movement, not 100%. Confirm by filtering to a single query, single page, single country, single device — if the position number moved with all four constant, you have a true ranking shift.

Bucket 2 — Query mix drift

You started ranking (or stopped ranking) for new low-position queries. If you launched a new topic cluster and now rank position 45 for 60 new long-tail terms, your average will look worse even though every existing query is unchanged. Diagnosis: filter to queries with impressions in both compare periods. If average is flat within that intersection, mix drift is the whole story.

Bucket 3 — SERP-feature crowding

Google added a featured snippet, an AI Overview, a shopping carousel, or a video block above your listing. Your organic rank did not change but your effective position number did. Diagnosis: run the query in an incognito window and count what sits above your listing.

Bucket 4 — Duplicate-listing collapse

A second URL of yours that was outranking the primary URL was deindexed, redirected, or consolidated. The primary URL's average position "improved" — not because anything about the primary URL changed, but because the topmost-rule aggregation shifted. Diagnosis: check for URL Inspection changes on adjacent pages in the same date range.

Bucket 5 — Personalization and geo-drift

Google's SERPs vary by location, device, and browsing signals. If a large share of your impressions come from one region and Google changed its regional balance for a topic, your average drifts even though no query-level change occurred. Diagnosis: check Country and Device tabs — the movement should live disproportionately in one segment.

Bucket 6 — Reporting artifact

Data anomalies, delayed backfill, sampling changes on the 1,000-query cap, or GSC infrastructure incidents. Rare but real. Diagnosis: does the movement resolve on its own within 5-7 days with no action? Then it was probably an artifact.

The single most important habit is asking "which bucket?" before proposing a fix. Content improvements do nothing for Bucket 3. Backlinks do nothing for Bucket 4. Republishing does nothing for Bucket 6. The cannibalization audit is a Bucket 1/Bucket 4 diagnostic tool specifically.

The 15-Minute Weekly Average-Position Review

The dashboard invites you to stare at a line chart. Staring produces nothing. Here is an operator cadence that ships decisions in 15 minutes.

Minute 0–3 — Set up the compare. Open Performance → Search Results. Set date range: Last 28 days vs Previous 28 days. Enable all four metrics (clicks, impressions, CTR, average position). Set country = your primary country. This eliminates the biggest sources of geo-noise.

Minute 3–6 — Scan Pages tab, sorted by "Position Difference" descending. The top of this list is pages whose position number got worse. The bottom is pages that improved. Look for movements of ≥3 positions on pages with ≥200 impressions in the recent window. Under those thresholds, noise dominates signal.

Minute 6–10 — For each meaningful mover, ask which bucket. Open the page → Queries tab → sort by "Position Difference." If the queries all moved together, you likely have Bucket 1 (real ranking shift). If a handful moved and the rest are stable, look for Bucket 2 (mix drift) or Bucket 3 (SERP-feature crowding — run the top query in incognito).

Minute 10–13 — Log the diagnosis, not just the movement. A weekly review with entries like "cannibalization guide dropped from 8.2 to 12.4 (Bucket 3 — AI Overview added, no ranking loss)" is a system that compounds. A log with just "cannibalization guide dropped" produces panicked republishing that makes rankings worse.

Minute 13–15 — Queue at most 3 concrete actions. If you queue 15, you will do 0. The three should be the highest-impact single actions per diagnosis: a new section for a Bucket 1 loser, an internal link burst for a Bucket 3 crowded page, a canonical cleanup for a Bucket 4 collapse.

Ship this every Monday and you will outperform 90% of teams still doing 90-minute audits that produce a document nobody reads.

Position 8-to-3 Playbook (The Highest-ROI Movement in SEO)

Moving a page from position 8 to position 3 is worth more clicks than launching most new pages. The math is unglamorous: position 8 gets ~2-3% CTR on non-featured queries; position 3 gets ~10-13%. That is a 4-5x traffic multiplier on a page you already have. Here is the sequence that consistently ships it.

Step 1 — Confirm the query is actually competitive at position 3. Some queries have such heavy SERP-feature stacking that position 3 organic sits below an AI Overview, a video, and a knowledge panel. In those cases, position 3 is worth about what position 8 is worth today — the ceiling has moved. Run the query. If the first organic result is more than one scroll below the fold, redirect effort to a query where the ceiling still matters.

Step 2 — Read the top-3 pages critically for structural moves you are missing. Not "keyword density." Look for: schema types they use, question-answer sections you skipped, comparison tables, embedded tools, downloadable templates, structured how-tos with numbered steps. Position 8 → 3 rarely comes from more prose; it comes from matching the format Google has decided the query wants.

Step 3 — Add the missing format element. If the top 3 all have a comparison table, add one. If they all answer "how much does X cost," add a cost section with a real number. If they all have a summary at the top, add one. The compounding effect: format-match closes the "why doesn't Google rank us at 3" gap that content depth alone can't close.

Step 4 — Rewrite the meta description around the winning promise. Position 8 pages usually have generic meta descriptions written for the article, not the query. Write one that promises the specific outcome the searcher wants (a checklist, a template, a calculator, a decision) and delivers on it above the fold.

Step 5 — Build 3 internal links from your highest-authority pages using the target query as anchor text. Not the URL slug. Not "click here." The actual query. This is a 15-minute step that most teams skip and that reliably closes half a position on its own.

Step 6 — Request re-indexing via URL Inspection. Google will re-crawl in days rather than weeks. See the URL Inspection tool guide for the exact resubmit flow.

Step 7 — Wait 21 days. Check. Position movement on established pages lags content changes by two to three weeks. Checking on day 4 tells you nothing. Checking on day 21 with the same query filter tells you everything.

The teams that ship this consistently double the traffic of pages they already own faster than they can launch new ones. It is, dollar for dollar, the single best use of a mid-week SEO block.

The Position Distribution Curve (What Your Real Average Looks Like)

The number on the dashboard is a mean. The mean lies when the distribution is skewed, and search-position distributions are always skewed. Two sites can both have "average position 14" and be in wildly different competitive shape. One might have a bell curve centered on 14 (uniformly mediocre — a slow slog). The other might have a bimodal distribution — 40% of queries at position 1-3 and 40% at position 40-50, averaging to 14 (a strong core plus an experimental long tail).

The bimodal site is dramatically healthier and should be operated dramatically differently: harvest the winners for internal linking, prune or consolidate the long tail, and stop optimizing for the mean.

How to see your own distribution in 5 minutes: Export the Queries tab as CSV, filter to queries with ≥50 impressions, then bucket positions into 1-3 / 4-10 / 11-20 / 21-40 / 41+ and count. If more than 25% of your qualifying queries sit in the 1-3 bucket, you have a strong-core site and should keep pushing content that resembles your winners. If more than 40% sit in 21+, you have a long-tail-heavy site and your best move is a pruning + consolidation pass, not another content burst.

Averages hide this shape. Distribution reveals it. The teams that operate on distribution outperform the teams that operate on averages by roughly the ratio of a 90-day compound: small differences in what you look at become large differences in what you ship.

The Five Most Common Average-Position Misdiagnoses

Misdiagnosis 1 — "Our average position got worse, we must have been penalized." Almost never. Manual actions and algorithmic penalties usually manifest as sudden click collapse, not gradual position drift. Check Manual Actions in GSC. If empty, look at bucket 2 or 3 first.

Misdiagnosis 2 — "This page dropped from 5 to 12, we need to republish it." Not before checking whether the query mix changed. If GSC started attributing 30 new long-tail queries to the page, the "drop" is a mix-drift artifact and republishing does nothing. Filter to the top-5 queries by impressions and check position on each individually.

Misdiagnosis 3 — "Our average is 14 across the site, we're mid." A site-level average is nearly useless. Break it down by page-type (homepage, category, product, blog) and by intent (informational vs transactional). The averages will be dramatically different and one of them is almost always the real problem.

Misdiagnosis 4 — "Position improved but clicks didn't, GSC is broken." GSC is not broken. You are seeing SERP-feature crowding — your organic rank improved but a featured snippet, AI Overview, or ad block absorbed the click. Run the query and count what sits above your listing.

Misdiagnosis 5 — "Our position numbers look great, we're winning." Winning is measured in clicks, not positions. A "position 2.3" page with 40 impressions and 1 click is not winning. Sort by clicks first, always. Position is a lever, not a scoreboard.

When Average Position Doesn't Matter (Skip It)

Not every page should be judged on average position. Skip the metric entirely for:

Branded queries. You are the answer. Position 1 is default. Movement here is noise.
Navigational queries where you are the destination. Same reason.
Pages targeting SERP features exclusively (definitional queries where you own the featured snippet). Track "in snippet or not" as a binary; position doesn't add information.
New pages under 21 days old. Google is still learning the page. Positions are unstable by design. Wait for the 21-day mark before drawing conclusions.
Pages with fewer than 100 impressions in the review period. Below this threshold, the average is one or two data points and drifts wildly. Ignore the number.

The best SEO teams deliberately narrow what they measure on which pages. The teams that don't drown in noise and mistake it for insight.

Automating Average Position Monitoring

Checking average position manually in GSC works for a site with a handful of pages — but for sites with hundreds of pages and thousands of queries, manual monitoring misses most opportunities.

Search Console Tools automatically identifies:

Pages where average position improved or declined week-over-week
Striking distance queries (pos 5–20) by impressions
CTR anomalies (high position, low click-through)
Queries where you jumped from page 2 to page 1

Instead of spending 30+ minutes in GSC every week building pivot tables, the analysis runs automatically and surfaces the highest-impact actions.

Frequently Asked Questions

What does average position mean in Google Search Console?

Average position in Google Search Console is the mean ranking position of your website or a specific page across all search queries and dates in the selected time range. A position of 1 means your page appears first in Google search results; position 10 means it's last on page 1. Lower numbers are better.

What is a good average position in Google Search Console?

A good average position in Google Search Console is generally under 10, meaning your content appears on the first page of Google results. Positions 1–3 are excellent (15–30%+ click-through rate). Positions 4–10 are competitive. Position 11+ (page 2 or beyond) typically sees less than 1% CTR and is considered an area for improvement.

Why is my average position going up (getting worse) in GSC?

If your average position number is increasing (e.g., from 8 to 14), your rankings are declining. Common causes: new competitor content outranking you, content freshness issues (page hasn't been updated), loss of backlinks, Core Web Vitals degradation, or Google algorithm updates. Use GSC's date comparison feature to identify which queries dropped.

Why is my average position going down (improving) in GSC?

If your average position number is decreasing (e.g., from 25 to 12), your rankings are improving — the lower the number, the higher the position. This typically follows content updates, new backlinks, improved Core Web Vitals, or increased click-through rate signals.

Does average position affect how much traffic I get?

Yes, directly. Pages ranking position 1–3 receive roughly 15–30% of all clicks for a query. Pages at position 4–10 receive 2–10%. Page 2 (positions 11–20) receives under 2% total. Moving from position 12 to position 7 can increase traffic from that query by 3–5x.

How do I improve average position for a specific keyword in GSC?

To improve average position for a specific keyword: (1) identify the page ranking for it in GSC, (2) update the content to more directly match the search intent, (3) improve the title tag to front-load the keyword, (4) add internal links from other high-authority pages on your site, (5) check for page experience issues (Core Web Vitals) in GSC that may be suppressing rankings.

Why did my average position get worse but my clicks went up?

Both can happen at the same time, and it usually means Google started ranking your page for more queries (query-mix drift, "Bucket 2"). The new queries rank lower than your established ones, dragging the average up (worse) even as total impressions and clicks grow. Filter to the queries that had impressions in both compare periods — if position is flat within that intersection, mix drift explains the whole gap and it's actually a good sign.

Does an AI Overview count in average position?

Yes. AI Overviews are treated as SERP features and occupy positions above organic results. If Google adds an AI Overview to a query where you rank organically at position 3, your "position 3" is now sitting below the AI Overview block — the number is unchanged but the effective position on the page is lower. Track this via the AI Overviews report to see which queries added the feature.

What's the difference between average position and rank tracking tools?

Average position in GSC is calculated across every impression Google recorded for your page, aggregated over the date range you selected — it reflects actual user-seen SERPs including personalization, device, and location variation. Third-party rank trackers pull a single scraped SERP per query per day from a fixed location, so they show a cleaner but less representative number. Use rank trackers for competitive tracking and GSC average position for what real users actually saw.

How many impressions do I need before average position is reliable?

At least 100 impressions in the compare window, and ideally 500+, before drawing conclusions about a single page or query. Below 100, the average is one or two data points and swings wildly with normal SERP variation. This is the single most common source of false-positive "ranking drops" in weekly reviews — a page that "fell from 4 to 22" on 12 impressions has not actually fallen.

Does GSC average position include image and video search?

Only if you have those search-type filters enabled. By default, the Performance report shows Web search only. Change the "Search type" filter at the top of the report to include Image, Video, or News to see those numbers separately. Combined average positions across search types are not shown; each type is reported independently to prevent muddying the metric.

How long does it take for average position to reflect a content update?

Typically 14-21 days for established pages, longer for new pages. Google needs to re-crawl the page, re-evaluate it against ranking signals, and accumulate enough new impressions at the new position to shift the average. If you check on day 3 and see no movement, that is not a failed update — that is Google not having finished re-crawling. Use URL Inspection to request re-indexing and shorten the tail.

What Is Average Position in Google Search Console? (And How to Improve It)