Search Console exports: three audits the GSC UI hides from you

Most SEOs look at Google Search Console the same way: open the Performance tab, glance at the totals, maybe sort the queries table by clicks, leave. The dashboard is fine for vibes. It's terrible for diagnostics.

The actual data — even with sampling and filtering — is dense enough to power three high-leverage audits that the UI cannot show you. All you need is one Performance export per month and 20 minutes.

The 1000-row problem (and how to actually solve it)

GSC's UI caps any view at 1000 rows. Most sites have 5-50K queries with at least one impression in a 28-day window, which means you're seeing somewhere between 2% and 20% of your dataset.

Three ways out:

1. Search Analytics for Sheets (free Chrome extension by Mike Richardson). Pulls up to 25K rows per query into Google Sheets. Easiest path for most people.
2. Looker Studio native GSC connector. Same data, prettier dashboards, harder to manipulate.
3. GSC API via Python or Apps Script if you want full control. ~50 lines of code.

For everything below, I'm assuming you have an export with at least 5K rows. If you're stuck with the UI, you can still do the audits — just on a much narrower slice.

Audit 1: Striking distance opportunities

The premise: queries currently ranking 11-20 are statistically the cheapest to move. They've already passed Google's quality bar (you're on page 2), and a content refresh, internal link push, or title tweak can often bring them onto page 1.

What to do:

1. Open your Queries export. Filter Position ≥ 11 and Position ≤ 20.
2. Sort by Impressions descending.
3. Filter out branded queries (variations of your brand name) — they don't count.
4. Take the top 20 queries.

For each one, click through to the actual SERP, check who ranks 1-10, and ask: what would have to be true about my page for it to deserve to overtake position 10? Usually the answer is one of:

• Add a section that directly answers a sub-query the top results cover.
• Add internal links from related, higher-authority pages on your site.
• Update the publish date and refresh the post.
• Tighten the title and meta to match query intent more directly.

This is the highest-ROI activity in SEO and almost nobody does it systematically because the GSC UI hides which queries are at position 11-20 unless you specifically filter for it.

Audit 2: Keyword cannibalization

The premise: when multiple URLs on your site rank for the same query, Google often serves the wrong one — diluting your CTR. Cannibalization is the most common cause of "we have the content but ranking is mediocre."

What to do:

1. Open your Pages-by-Query export (or join Pages and Queries datasets).
2. Group by Query. Count the number of distinct Page entries per query in the same period.
3. Flag any query where 2+ pages have at least 50 impressions. These are your cannibalization candidates.

For each candidate:

• Decide which URL should rank (usually: the one with the most search-intent match, or the most authority, or the one that already has more clicks).
• For the others: either redirect them to the canonical URL, change their title/H1 to target a different query, or merge them.

I've seen sites recover 30-40% of organic traffic just by resolving 5-10 cannibalization conflicts. The GSC dashboard cannot tell you about this — it shows the single best-performing page per query and hides the conflict.

Audit 3: CTR outliers

The premise: at any given position, there's a normal range of CTR. Pages well below the normal range have a snippet problem — bad title, bad meta, or bad URL. Pages well above the normal range have a snippet doing extra work, and you should figure out why so you can replicate the pattern.

Reference CTR ranges (informational queries, 2024-2026 averages):

These vary by SERP type — branded, transactional, and high-AI-Overview queries skew lower. Treat them as a baseline, not a contract.

What to do:

1. Filter your export for Impressions ≥ 200 (statistical noise floor).
2. Compute expected CTR per row using a simple position lookup against the table above.
3. Compute delta = actual_ctr - expected_ctr.
4. Sort by delta. The bottom 10 are underperformers. The top 10 are outperformers.

For underperformers: rewrite the title and meta. Use the Snippet Optimizer and AI Snippet Writer to draft variants.

For outperformers: examine why. Is there a number in the title? An emoji? A power word? A particularly punchy meta description? Whatever pattern is working — apply it to your underperformers.

Cross-source enrichment

GSC alone tells you what's happening. To understand why, join it with two other sources:

• Lighthouse / PageSpeed — pages with poor Core Web Vitals can have suppressed CTR even at good positions. Slow pages get clicked through less.
• Screaming Frog Internal_All — pages with missing or duplicate titles, weak H1s, or noindex tags will have weirder GSC data than they should.

The Report Builder does this join for you — drop in a GSC export, a Lighthouse JSON, and a Screaming Frog Internal_All, and it surfaces things like "high-impression pages with broken titles" automatically.

What not to trust in GSC data

• Sampling. GSC samples queries above a certain threshold. Long-tail rare queries are underrepresented; common queries are accurate.
• Filtered queries. Queries that include personal info or that GSC considers privacy-sensitive get hidden. The "Anonymous queries" line in your export captures these in aggregate.
• CTR for branded queries. Always inflated. Filter them out for any analysis.
• Position averaging. A "position 5" can mean ranked #5 every day, or #1 half the days and #9 the other half. Look at the distribution, not just the average.
• Date alignment. GSC data lags 2-3 days. Don't compare today's report to yesterday's traffic.

A monthly audit cadence

Once you've got the export pipeline set up, this becomes a 20-minute monthly ritual:

1. Pull last 28 days of Queries and Pages.
2. Run the three audits above.
3. Pick 2-3 actions you'll take this month (one striking distance, one cannibalization fix, one CTR fix).
4. Document and ship.

That's 24-36 high-leverage actions per year, each backed by your own data. It's the most effective SEO process I know and it costs almost nothing.

Tools

• Report Builder — drop in your GSC export, Lighthouse, Screaming Frog and others; assemble a sectioned audit. Print to PDF for the client.
• Snippet Optimizer — for fixing CTR outliers via title/meta rewrites.
• AI Snippet Writer — for generating fresh CTR-tuned variants when your own copy is stale.

The GSC UI is for skimming. The export is for working. The gap between the two is where the audits live.