Five of nine Charleston results changed in seven days
I log the same Charleston SEO queries every week. In seven days, five of the nine page one results turned over, including mine. Here is what to do about it.
Last Monday my site was sitting at number five on page one for "charleston seo company." This morning it is not on page one at all. I did not touch that page in between. No edits, no new links, no redirect, nothing. The page is byte for byte what it was when it placed.
That is the part of local search nobody sells you on, so I want to show you the actual numbers from my own logbook rather than tell you about it.
What I actually measure, and what I do not
Every week I run the same set of Charleston commercial queries through the same tool and write down the top results in a file in the repo. Not a rank tracker with a dashboard. A JSON file I append to, because I wanted a record I could argue with later.
Two honest limits before the numbers, because they change how much weight you should put on this. First, this is a logged search result set, not my personalized Google on my phone in my house in Charleston. Yours will differ, and so will your neighbor's. Second, I capture nine or ten results per query, which is roughly page one and not exactly page one. So read what follows as a consistent instrument pointed at the same target every week, not as the literal ten blue links any one person saw.
With that said, the instrument did not change. The results did.
The same query, seven days apart
| Position | September 29 | October 5 |
|---|---|---|
| 1 | thriveagency.com | thriveagency.com |
| 2 | designrush.com (directory) | newmedia.com |
| 3 | newmedia.com | myvirtualtalent.com |
| 4 | palmettosoft.com | palmettosoft.com |
| 5 | charlestowne.marketing | dmthriveagency.com |
| 6 | buzzcube.com | holycityweb.com |
| 7 | charleston-seo.com | baldwinseo.com |
| 8 | rankthecoast.com | themanifest.com |
| 9 | themanifest.com | varundigitalmedia.com |
Four results held: Thrive, NEWMEDIA, PalmettoSoft, The Manifest. Five were replaced: DesignRush, me, Buzzcube, charleston-seo.com and Rank The Coast. Five brand new entrants walked in: My Virtual Talent, DM Thrive, Holy City Web, Baldwin Digital and Varun Digital Media.
Fifty six percent turnover in a week on a head commercial term, and not one of us did anything on the Tuesday to deserve it.
Look at which four held. Thrive is a national agency out of Texas. NEWMEDIA is national with a Rivers Avenue office. The Manifest is a directory owned by Clutch. PalmettoSoft is the one genuinely local, genuinely established shop in that group. The stable part of page one is big and old. The churn is happening in the bottom five, which is where every small Charleston shop including mine lives.
(I want to be careful here. One week of data is one week of data. If I write a post in three months saying the bottom five churns and the top four holds, that will be because I kept logging it, not because I decided it today.)
Then I added one word
Same day, same tool, five minutes later. I searched "best charleston seo company" instead of "charleston seo company."
Without "best," two of nine results were a directory or a listicle, and both sat in the bottom half. With "best," four of nine were directories, and Clutch took the number one spot outright, with DesignRush at three, Onfolio at five and GoodFirms at six.
One adjective, and Google decided the query wanted a list of agencies rather than an agency. That is not a ranking change. That is Google changing its mind about what kind of page answers the question.
This is the thing I wish more Charleston owners understood before they pay anybody for "ranking for Charleston roofer." There is no single page one. "Charleston roofer," "best Charleston roofer," "Charleston roofer near me" and "Charleston roofer reviews" can return four different kinds of page, and your service page is only eligible to win some of them. Chasing a keyword without checking what page type currently wins it is how good money gets spent on a page that was never allowed to compete.
The detail that stung most
When my site did place at number five on the 29th, the page that placed was /services/, my services hub. Not /services/charleston-seo/, which is the page I built specifically for that term.
So I counted. Across the 99 pages in my own build, the money page has 150 internal links pointing at it from 95 pages. The hub has 123 from the same 95. The money page runs 2,703 words against the hub's 1,594. The money page contains the exact phrase "seo company" eight times. The hub contains it zero times.
The page that lost on every signal I control is the page Google picked.
I have written before about a blog post from Christmas Eve outranking my own service page, and this is the same lesson arriving from a different direction. You do not get to nominate the page. You get to build candidates. Google selects, and its selection logic is not a scoring sheet you can reverse engineer from your own word counts.
At this point a lot of people are thinking: so rankings are meaningless
No. They are a noisy measurement of something real, which is a very different thing from meaningless.
The mistake is treating a single week's position as a signal worth reacting to. I have watched owners do real damage this way. Rank drops on Tuesday, they rewrite the H1 on Wednesday, swap the title tag on Thursday, add four hundred words on Friday. Now they have changed four variables on a page that was fine, the position recovers or does not recover for reasons unrelated to any of it, and they have learned nothing except a superstition.
What I actually act on:
- A trend across four or more checks, not one. If a term has been outside the top ten for a month and a competitor has been inside it for a month, that is a finding. One Tuesday is weather.
- Page type mismatch. If the query returns directories and listicles and I am sending a service page at it, no amount of copy fixes that. Different asset, or skip the term.
- Which of my pages Google picked. If it keeps choosing a page I did not build for the term, my internal linking and my page differentiation are the problem, not my word count.
- Leads, which do not churn fifty six percent in a week. Calls and form fills are the number that pays the mortgage, and they are far more stable than any position.
- Whether the page is honest about where it is. I ran that check on myself a few weeks ago and found sixteen of my twenty six service pages never named South Carolina. That was worth fixing. A one week rank wobble was not.
Do this on Sunday, not tomorrow
Here is the actual homework, and it takes about fifteen minutes.
Open a plain text file. Write down today's date. Search your three most valuable service terms, exactly as a customer would type them, and write down the top five results for each plus where you appear. Then search each one again with "best" in front of it and write down whether the results are businesses or lists. Save the file.
Do it again next Sunday. And the one after. By the fourth entry you will know which of your terms are stable, which churn, which want a list instead of a page, and whether your own position is actually moving or just breathing. That file is worth more than any rank tracker subscription, because you built it and you trust it.
This is how we run Charleston digital marketing for our service business clients: four weeks of evidence before anyone touches a page, and leads as the number we report. If you would rather not keep the logbook yourself, that is fine, ask whoever you hire to show you theirs. An agency that cannot produce a dated record of your positions is guessing, and so are you.