worldwide88
Newbie
- Oct 15, 2019
- 48
- 31
Sharing this because I have not seen the pattern discussed much: Google swapping which of your own URLs it ranks for a money term, and the swap itself costing you the ranking.
Setup: Swedish roof-cleaning site, ~180 pages. Money term is our "roof cleaning price" query. Two URLs on the site can answer it: the main price page (rebuilt in stages, last big rebuild went live Nov 2) and an old 2025 blog article with a long price table that we had stopped caring about.
What GSC shows for that exact query (impression-weighted avg position per month, 16 months):
Main price page: Oct 2025 pos 3.6, Dec 10.7 (winter), Mar 2026 5.1, Apr 2.8, Jul 3.2, Aug 2.5, Sep 2.4, Oct 3.8. Then week 44: 7.3. Week 45: 24.5. Daily since Nov 2 it sits at 41-48 in our rank tracker (mobile), GSC avg 24 because desktop is kinder.
Old 2025 article: position 96 in Nov 2025, 60 most of 2026, basically dead. Week 43: 11.9. Week 44: 16.8. Week 45: 18.2 with 3 clicks. It is now the URL Google shows on page 2 for the term while the real page is on page 3-5.
So the pattern is: Google re-evaluated the site (we had a site-wide drop Oct 30 after publishing too many thin town pages at once, separate thread), and in the same window it stopped trusting the rebuilt page for the term and started surfacing the stale article instead. The rebuilt page was first recrawled Nov 8, so for a week Google ranked a version it had not even seen.
Screenshot is GSC, 6 months, this query only. Orange is position (inverted axis), blue is clicks.
Questions:
1. When Google picks a stale URL over your updated one for the same term, do you consolidate (301 the old article into the main page) or keep both and let it sort itself out? I lean 301 because the old article has a year of history and no reason to exist, but a 301 right after a site re-evaluation feels like adding noise.
2. Anyone measured how long a rebuilt page takes to regain a position it held for a year, when the content is objectively better but the URL was unstable for two weeks?
3. Is there any evidence that "price" pages with calculators/interactive parts get treated differently than static price tables? The old article is a plain table, the new page has a calculator and a lot more structure.
Numbers and dates from GSC, happy to share more.
Setup: Swedish roof-cleaning site, ~180 pages. Money term is our "roof cleaning price" query. Two URLs on the site can answer it: the main price page (rebuilt in stages, last big rebuild went live Nov 2) and an old 2025 blog article with a long price table that we had stopped caring about.
What GSC shows for that exact query (impression-weighted avg position per month, 16 months):
Main price page: Oct 2025 pos 3.6, Dec 10.7 (winter), Mar 2026 5.1, Apr 2.8, Jul 3.2, Aug 2.5, Sep 2.4, Oct 3.8. Then week 44: 7.3. Week 45: 24.5. Daily since Nov 2 it sits at 41-48 in our rank tracker (mobile), GSC avg 24 because desktop is kinder.
Old 2025 article: position 96 in Nov 2025, 60 most of 2026, basically dead. Week 43: 11.9. Week 44: 16.8. Week 45: 18.2 with 3 clicks. It is now the URL Google shows on page 2 for the term while the real page is on page 3-5.
So the pattern is: Google re-evaluated the site (we had a site-wide drop Oct 30 after publishing too many thin town pages at once, separate thread), and in the same window it stopped trusting the rebuilt page for the term and started surfacing the stale article instead. The rebuilt page was first recrawled Nov 8, so for a week Google ranked a version it had not even seen.
Screenshot is GSC, 6 months, this query only. Orange is position (inverted axis), blue is clicks.
Questions:
1. When Google picks a stale URL over your updated one for the same term, do you consolidate (301 the old article into the main page) or keep both and let it sort itself out? I lean 301 because the old article has a year of history and no reason to exist, but a 301 right after a site re-evaluation feels like adding noise.
2. Anyone measured how long a rebuilt page takes to regain a position it held for a year, when the content is objectively better but the URL was unstable for two weeks?
3. Is there any evidence that "price" pages with calculators/interactive parts get treated differently than static price tables? The old article is a plain table, the new page has a calculator and a lot more structure.
Numbers and dates from GSC, happy to share more.