We Rebuilt Our Entire Site in 10 Days. Here's the GSC Data, Including the Dip.
By Archie Cortés, Founder of AutoPilotPR. All data pulled September 4, 2026 via Google Search Console API. Screenshots available on request. This post will be updated with Day-30 data on September 25, 2026.
On August 25, 2026, we pushed the largest single update in AutoPilotPR's history. In roughly 10 days of work, we rewrote 48 blog posts, launched 12 new pages (pillar pages, geo pages, case studies, a free AI readiness tool), and permanently removed 3 posts that had zero traffic and weak targeting. We also redesigned the site architecture to reflect our repositioning to "Smart Websites" as a category.
No other marketing agency in Puerto Rico publishes their own GSC data during a rebuild. We are.
Here's exactly what happened, including the parts that look bad.
Table of Contents
- What We Changed and Why
- The Site-Wide Numbers
- Blog Performance: The Dip
- The New Pages: Too New to Judge
- What the Data Actually Means
- Day-30 Commitment: September 25
What We Changed and Why
Phase F launched August 25, 2026. The full scope:
- 3 posts killed with permanent 308 redirects to relevant pillar pages (zero traffic, wrong targeting)
- 29 posts updated with lighter edits: corrected internal links, refreshed CTAs, fixed dead URLs
- 16 posts updated with heavier rewrites: new structure, updated stats, stronger targeting
- 12 new pages launched:
/seo-agency-puerto-rico,/marketing-agency-puerto-rico,/ai-agents-puerto-rico,/web-design-puerto-rico,/custom-web-app,/marketing-agency-san-juan,/marketing-automation-puerto-rico,/how-it-works,/faq,/case-studies,/ai-readiness-score, and/results(portfolio)
The strategic goal: replace a site built for "AI marketing agency" positioning with one built for "Smart Websites," the category we're creating for the Puerto Rico market. More on that positioning here: What Is a Smart Website?
The Site-Wide Numbers
Measurement period:
- Pre-rewrite: August 1 to August 24, 2026 (24 days)
- Post-rewrite: August 25 to September 3, 2026 (10 days, starting Phase F launch day)
- Data source: Google Search Console Search Analytics API,
sc-domain:autopilotpr.com, pulled September 4, 2026
Site-wide impressions (all pages):
| Period | Days | Total Impressions | Daily Average | Clicks |
|---|---|---|---|---|
| Pre-rewrite (Aug 1-24) | 24 | 514 | 21.4 | 6 |
| Post-rewrite (Aug 25-Sep 3) | 10 | 257 | 25.7 | 4 |
Site-wide daily impressions are up 20% at Day 10. That's driven by the new pillar and geo pages beginning to index, plus the blog index page improving (from position 16.3 to position 4.2 for blog-related queries).
Key individual page movements at Day 10:
| Page | Pre Impressions | Post Impressions | Pre Position | Post Position |
|---|---|---|---|---|
| Homepage | 26 | 14 | 10.2 | 2.6 |
| /marketing-agency-puerto-rico | 45 | 21 | 30.9 | 16.1 |
| /seo-agency-puerto-rico | 4 | 4 | 20.5 | 10.0 |
| /ai-agents-puerto-rico | 11 | 14 | 14.8 | 11.9 |
| /marketing-agency-san-juan | 37 | 27 | 9.9 | 8.4 |
| Blog index (/blog) | 7 | 16 | 16.3 | 4.2 |
What jumps out: /marketing-agency-puerto-rico shows fewer impressions but moved from position 30.9 to 16.1. That is a meaningful position gain on a page that was barely on page 3. At position 16, it is just outside page 2. Fewer impressions at a better position is normal early-indexing behavior for a page Google is actively re-evaluating.
The homepage dropped in impressions but improved sharply from position 10.2 to 2.6. That number will likely stabilize as more data accumulates, but the directional signal is correct.
New pages (/ai-readiness-score, /case-studies, /contact) are showing their first impressions. /web-design-puerto-rico, /custom-web-app, /how-it-works, and /faq are not yet indexed with measurable impression data at Day 10. That is expected. New pages typically take 2 to 6 weeks to accumulate GSC data after launch.
Blog Performance: The Dip
This is the part we are publishing despite it looking bad.
Blog-specific impressions (/blog/*):
| Period | Days | Total Impressions | Daily Average | Clicks | Daily Average |
|---|---|---|---|---|---|
| Pre-rewrite (Aug 1-24) | 24 | 343 | 14.3 | 4 | 0.17 |
| Post-rewrite (Aug 25-Sep 3) | 10 | 122 | 12.2 | 2 | 0.20 |
Blog daily impressions are down 15% at Day 10.
Here is what causes this: when Google recrawls and re-evaluates a rewritten page, it temporarily reduces how often it shows that page in results while it re-assesses quality and relevance. This is not a penalty. It is the standard recrawl cycle. The algorithm is essentially saying "we noticed this page changed significantly, let us re-rank it." Pages that were rewritten with stronger targeting typically recover and improve; pages that were only marginally updated often recover faster.
The 3 killed posts are no longer generating impressions directly. Their redirects are passing link equity to the pillar pages, but that transfer is not instant in GSC data.
Industry data on post-rewrite recovery: most SEO practitioners observe a 3 to 8 week dip after significant content rewrites before rankings stabilize or improve. This is consistent with Google's documented behavior around major page updates.
We are not adjusting the narrative based on what the data shows. The dip is real. It is also expected. We will update this post on September 25 with the Day-30 data. If the blog impressions have recovered, we will say so. If they have not, we will diagnose why in public.
The bet the receipts brand makes: we publish what happened, not what we wish happened.
The New Pages: Too New to Judge
Pages with no meaningful data yet at Day 10:
| Page | Purpose | Day-10 Status |
|---|---|---|
| /web-design-puerto-rico | Primary pillar: web design | Not yet indexed with data |
| /custom-web-app | Pillar: web app builds | Not yet indexed with data |
| /how-it-works | Process / trust page | Not yet indexed with data |
| /faq | FAQ for AI queries | Not yet indexed with data |
| /ai-readiness-score | Free tool, lead magnet | 1 impression, position 1.0 |
| /case-studies | Portfolio / social proof | 1 impression, position 1.0 |
We launched 12 new pages. At Day 10, most have not accumulated enough GSC data to report. That is not a failure state. Google typically needs 2 to 6 weeks after a site crawl to surface new pages in Search Analytics with reliable data.
The /ai-readiness-score and /case-studies pages each showing position 1.0 on their first impression is a good early signal: Google matched them to a relevant query, ranked them first, and will likely expand their visibility as it sees more user signals. But 1 impression is 1 impression. We are not calling this a win yet.
Day-30 data will tell the real story for these pages.
What the Data Actually Means
Here is our honest read of the Day-10 snapshot:
What is working:
- Position improvements on core pillar pages (
/marketing-agency-puerto-ricofrom 30.9 to 16.1;/seo-agency-puerto-ricofrom 20.5 to 10.0) - Blog index ranking dramatically better (16.3 to 4.2)
- Site-wide daily impressions up 20%
- New pages beginning to surface in index
What is in the dip:
- Blog impression daily average down 15% (14.3 to 12.2)
- Homepage impressions down, though position improved (10.2 to 2.6)
- Several pillar pages show fewer impressions at better positions (early recrawl pattern)
What we cannot measure yet:
- 8 new pillar/geo/tool pages have no data yet
- The full link equity transfer from the 3 killed-post redirects
- Whether the blog dip is temporary (expected) or structural (requires diagnosis)
The Day-30 snapshot will answer most of these. If position gains are holding and blog impressions have recovered toward baseline, the rewrite worked. If positions slip back or blog impressions continue falling, we will say so and adjust.
Day-30 Commitment
September 25, 2026: We will update this post with the Day-30 GSC data using the same methodology. Same pages, same metrics, same API pull.
What we will specifically report:
- Did blog daily impressions recover to pre-rewrite baseline?
- Did the pillar page position improvements hold?
- What data is now available for the new pages?
- Did the 3 redirect link equity transfers show up in pillar page rankings?
If the rebuild worked, Day-30 will show it. If it did not, Day-30 will show that too, and we will diagnose in public.
Data note: All figures pulled September 4, 2026 via Google Search Console Search Analytics API. Comparison periods: Pre-rewrite = August 1 to 24, 2026 (24 days). Post-rewrite = August 25 to September 3, 2026 (10 days, beginning Phase F launch). Site property: sc-domain:autopilotpr.com. Impression counts are non-normalized across periods; daily averages are used for comparison. Position figures represent average ranking position for pages with at least 1 impression in the period.
Questions about the methodology or data? Contact us here.