SEO Case Study · Batch 2 · CTR at Scale

Scaling the system: 25 pages, a control group, and a rebrand shipped at once

Batch two took the workflow from 10 pages to 25, added a site-wide control group so the results can't hide behind seasonality, and folded in a company-wide brand-language cleanup. Search demand fell across the whole site this period — and that's exactly why the result matters.

25 pagesedited in one day (June 22), 2.5× the size of batch 1
+11.2%relative CTR lift on edited pages (0.88% → 0.98%)
+7.6%site-wide CTR lift over the same window — the control group
28-daymatched windows: May 25–Jun 21 vs Jun 23–Jul 20
The challenge

Prove it works at scale — during a traffic downturn.

Batch 1 proved the workflow on 10 pages. Batch 2 had to answer a harder question: does it hold up at 2.5× the volume, and can the results survive scrutiny when overall search demand is falling?

It also carried a second mission. Adoption Network was mid-rebrand — retiring its legacy "Law Center" name — and legacy brand language was scattered across titles, metas, and body copy on dozens of ranking pages. Every batch-2 edit did double duty: sharpen the page for CTR, and scrub the legacy branding while I was in there. One page alone had 140+ legacy brand instances in its testimonials.

How I measured it

Results

Demand fell everywhere. The edited pages kept more of it.

This is the honest headline: summer seasonality hit the whole site. Site-wide impressions fell 18.8% and clicks fell 12.7% between the two windows. No on-page edit stops a seasonal tide — so the fair question isn't "did traffic go up," it's "did the edited pages outperform the site they sit in?" They did.

CTR change, edited pages vs site-wide control

Blended CTR · May 25–Jun 21 (before) vs Jun 23–Jul 20 (after) · bars scaled to largest value
Edited 25 pages0.88% → 0.98%
Site-wide (control)0.80% → 0.86%
BeforeAfter

The edited pages' CTR rose +11.2% relative (0.88% → 0.98%) against the site's +7.6%. In plain terms: while there were fewer searches to win, the rewritten pages captured a larger share of the clicks that existed. That's the piece on-page edits actually control — and the control group is what makes the claim defensible instead of decorative.

How I read these results

Being straight about the rest: clicks on the edited set fell 13.2%, in line with the site's 12.7% — that's demand, not damage. CTR improved on 11 of 25 pages. And key events on this batch were flat (16 → 15): these are informational pages, so the conversion story still belongs to batch 1's lead pages, not this set. I'd rather show that than pretend every metric moved.

Spotlight

One page beat the downturn outright

The Inspirational Adoption Quotes page — a birth-mother audience page — got a title and meta rewrite plus the brand cleanup, and posted the batch's clearest win against a falling market: clicks up 70% while its impressions fell 12%.

3.05% from 1.59%CTR — nearly doubled
280 ↑70%GSC clicks (from 165)
6.6 from 7.4Avg. position, up
+83GA4 sessions vs baseline

Runner-up wins on the same pattern: History of Adoption (CTR +0.55 pts), Celebrating Birth Mother's Day (+0.64 pts, position +1.4), Philippines Adoption (+0.26 pts, +28 sessions), and Adoption Name Change, where fixing a duplicate meta description lifted both CTR and position.

The second mission

A rebrand shipped through the same pipeline

Alongside the CTR work, batch 2 executed the brand transition on every page it touched: legacy "Law Center" naming and restricted terms came out of titles, metas, H1s, and body copy — including a bulk find-and-replace of 140+ legacy signatures on the reviews page. On brand pages the traffic numbers were flat, and that's fine: the goal was risk removal, and it was tracked, shipped, and verified with the same spreadsheet rigor as the SEO work. Treating brand governance as a measurable pipeline — not a someday cleanup — is the operational win.

Full transparency

All 25 pages, wins and misses

Every page in the batch on matched 28-day windows — Search Console clicks, impressions, CTR, and position, before → after. Five pages earned a re-edit flag; that's the system working, not failing.

PageClicksImpressionsCTRAvg. pos.Read
US Adoption Statistics540 → 44847,359 → 36,9681.14% → 1.21%6.96 → 7.2— Hold
Ethiopia Adoption61 → 357,778 → 6,1110.78% → 0.57%6.86 → 7.32— Hold
How Long Does It Take to Adopt35 → 377,096 → 6,8680.49% → 0.54%7.91 → 8.22— Hold
History of Adoption62 → 7013,150 → 6,8920.47% → 1.02%5.71 → 8.06✓ Win
Home Study Questions Checklist340 → 20814,399 → 7,6242.36% → 2.73%7 → 9.32✓ Win
Adoption Definitions & Terms16 → 169,382 → 5,5720.17% → 0.29%7.8 → 8.58— Hold
Inspirational Adoption Quotes · birth-mother165 → 28010,391 → 9,1831.59% → 3.05%7.41 → 6.6✓ Win
Domestic vs International Adoption155 → 16011,761 → 12,3121.32% → 1.3%9.11 → 9.38— Hold
Birth Father Rights · birth-mother6 → 35,695 → 4,0370.11% → 0.07%8.81 → 9.7↻ Re-edit
Do I Have to Be Married to Adopt30 → 244,812 → 5,1830.62% → 0.46%5.83 → 6.64— Hold
The Adoption Symbol26 → 163,448 → 3,1140.75% → 0.51%6.33 → 5.57— Hold
Adoption Name Change7 → 143,169 → 3,2680.22% → 0.43%9.07 → 8.91✓ Win
Requirements: Can I Adopt?34 → 145,650 → 4,7140.6% → 0.3%10.63 → 14.25↻ Re-edit
South Korea Adoption39 → 256,874 → 4,7830.57% → 0.52%8.36 → 9.65— Hold
Open, Semi-Open & Closed Adoption27 → 254,564 → 3,6110.59% → 0.69%9.93 → 10.3— Hold
What It Means to Be an Adoptee2 → 13,812 → 3,2870.05% → 0.03%7.13 → 7.1↻ Re-edit
Long Island Adoption11 → 52,733 → 2,7430.4% → 0.18%4.91 → 4.66— Hold
Why Adoption Network · brand32 → 166,881 → 3,7280.47% → 0.43%4.49 → 4.52⚑ Brand
Philippines Adoption27 → 403,574 → 3,9370.76% → 1.02%9.63 → 10.06✓ Win
Pregnancy Lifestyle Changes · birth-mother0 → 038 → 930% → 0%6.84 → 4.31— Hold
Celebrating Birth Mother's Day1 → 4552 → 4900.18% → 0.82%8.22 → 6.82✓ Win
Same-Sex Adoption49 → 178,790 → 7,3300.56% → 0.23%14.23 → 17.03↻ Re-edit
How Do I Tell Him I'm Pregnant · birth-mother, priority13 → 62,968 → 2,7020.44% → 0.22%10.46 → 12.2↻ Re-edit
Adoption & Obituaries16 → 113,484 → 1,3050.46% → 0.84%4.35 → 3.95✓ Win
Adoption Network Reviews · brand47 → 378,767 → 8,1460.54% → 0.45%6.51 → 6.59⚑ Brand

Clicks, impressions, CTR, and position from Search Console compare export (May 25–Jun 21 vs Jun 23–Jul 20). "Brand" = page edited primarily for brand-language cleanup; flat traffic was the expected outcome. Sessions and key events tracked separately in GA4 on the same windows.

What I learned

Skills this batch built

Next steps

Where this goes

  1. Re-edit How Do I Tell Him I'm Pregnant first — it's a birth-mother lead page and the batch's priority miss
  2. Rework Same-Sex Adoption with content depth, not just metadata; it's stuck on page two
  3. Second pass on Requirements: Can I Adopt and the Adoptee page, where the first rewrite didn't land
  4. Re-pull all 25 after the next 28-day window to separate settling from real decline
  5. Batch 3: roll the workflow plus control-group measurement onto the next tier of high-impression pages

The takeaway

Batch 1 proved I can lift pages. Batch 2 proves I can run a program: scale it, control for what I don't control, ship a rebrand through it, and report the results the way a skeptical stakeholder would want them reported.

Google Search ConsoleGA4WordPressAIOSEOGoogle SheetsControl-group analysisCTR optimizationBrand governanceReporting