85 Days of Doing Our Own SEO: What Changed Since the 60-Day Case Study

In September we published a 60-day case study on our own website and said we would check the numbers again in thirty days. This is that check. Every number was pulled again on October 7, 2026, from the same two places: Google’s search data (Search Console, United States searches only) and a rank tracker.

The new site went live on July 14, 2026. October 7 is day 85. Google’s search data runs about three days behind, so the newest full day we have is October 4. The case study’s “most recent month” was the 30 days from July 29 to August 27. Our “now” window is the 30 days from September 5 to October 4.

One thing up front. This is a visibility update. It proves what Google shows, not what the phone did. One of the main numbers went down.

The Same Numbers, Then and Now

Searches the site ranks for. The case study said 257. Counted the same way today, that same August window reads 259, and the September 5 to October 4 window reads 395.

That count includes searches Google’s data still lists for our site even if we were not shown for them once during the window. That is a loose count, and the case study did not say so. The stricter count is searches where Google showed our site at least once in the 30 days. By that count: 187 then, 310 now. We will print both from here on.

Searches on page one. The case study said 127. The same count today reads 128 for the August window and 166 for the latest one. The stricter count, shown at least once and averaging a top-ten spot: 56 then, 81 now.

Not all 81 are worth having. Thirty-eight are searches for a service we sell or for our own name. The other 43 are broad one-word searches, or searches that have nothing to do with what we do. Of the 38 that matter, 8 average first place. That is a standing, not a gain. We did not sort the August list the same way.

Times shown in Google results. The case study said 3,266. That window reads 3,263 today. The latest window, September 5 to October 4, reads 2,967. That is 296 fewer. The 30 days in between, August 28 to September 26, read 2,989. More on it below.

Clicks. Flat. For United States searches, 10 in the August window and 11 in the latest one.

Tracked terms in the top three and on page one. The case study counted 77 tracked terms: ten in the top three, eighteen on page one, two in first place. Those same 77 terms today: fifteen in the top three, twenty-three on page one, twelve in first place. Fifty-two of the 77 still do not rank at all, against 53 at the time of the case study.

We have since added terms, so the board is now 109. Across all 109: eighteen in the top three, thirty-one on page one, fourteen in first place, and 73 not ranking. The tracker’s newest data is dated October 2.

What Went Up

Almost all of it is our core service-plus-town searches, each on a local SEO page for one service or one trade.

• Ten of those searches moved into first place on the tracker since the case study. Four of them were off page one at the time: the contractor, heating and cooling, nonprofit and sports organization searches, which sat between 11th and 19th, are all first now.
• “Digital advertising Manitowoc WI,” 3rd in the case study, is now first.
• “Website design Manitowoc WI” was the slip we reported in the case study, down to 11th on the tracker. It reads 7th today. Better, but not back to the 2nd place it started from.
• One search on reputation follow-up moved from an average of 19th to an average of 4th in Google’s own data.

One was not a straight line. The contractor search dropped off page one for a week in the middle of September. We added a short question-and-answer section to the page built for it and pointed two newer articles at it. It reads first today.

What Went Down

Google showed us less often. 3,263 times in the August window, 2,967 in the latest one. The likely read, not a finding: four broad searches about AI consulting, none with a town in them, account for most of it. Google showed us for those four 183 times in the August window, always on page three or four. In the latest window it did not show us for them at all. Showings that deep rarely bring a visitor, but the drop is real. What we are doing: we publish an AI article most weeks, and those keep pointing at the pages built for these searches.

“AI consultant Manitowoc WI” slipped in both sources. The tracker had it 3rd at the time of the case study and 5th today. Google’s own data had it averaging 5th on 11 showings in August and 11th on 8 showings now. The likely read, not a finding: the tracker shows the same page of ours answering that search as before, so this looks like other sites moving up, not something breaking on ours. What happens next: our standing rule when both sources agree is to refresh the page, link to it from newer articles, and post about that service on our Google listing. That is this month’s job for this search.

“Online marketing Manitowoc WI” fell from 3rd to 12th on the tracker. Off page one. Google’s own data disagrees: an average of 2nd, on 9 showings, in the latest window. The likely read, not a finding: Google switched which of our pages answers that search, and the tracker’s one fixed check caught the weaker one. What we are doing: checking it again next Monday, and fixing the page if Google’s own data follows the tracker down.

“Marketing agency Manitowoc Wisconsin” went from 2nd to 4th on the tracker. Google’s own data has it averaging first on 5 showings. The likely read, not a finding: the same page-switching. What we are doing: watching it.

A search on a short version of our own name slipped. The tracker reads 11th today, just off page one. Google’s own data moved the same way: an average of 9th on 37 showings in August, 12th on 13 showings now. The likely read, not a finding: other companies share those three letters. What we are doing: nothing special. The fuller version of our name reads first on the tracker.

Three things did not move. Clicks, as above. The reputation search that improved in Google’s own data still reads 12th on the tracker, the same as in August. One broad AI consulting search has read 30th since August. And 73 of the 109 tracked terms do not rank, nearly all of them searches with no town in them.

What We Are Doing Next

We are not adding more pages aimed at the town searches we already hold. That is how the website design slip happened. The work is on existing pages: the AI consulting page first. The cleanup where two of our own pages on website design compete for one search is still open.

For searches with no town in them, the case study said three to six months. Day 85 says the same.

We will pull all of this again in about thirty days and add it to the case studies, up or down.

Common Questions

Why did you show up for more searches but get shown less often?

Because the two numbers measure different things. The count of searches went from 187 to 310, but most of the new ones are small and specific, shown once or twice each. Meanwhile four broad searches that had shown us 183 times in the August window stopped showing us at all. A long list of small searches did not make up for losing a few big ones.

Is a drop in times shown a bad sign?

Not by itself, but we are not going to wave it off. The showings we lost were on page three and four for broad searches, where almost nobody looks. The searches a nearby owner would actually type held or improved. Still, 3,263 to 2,967 is a decline, and the fix is better pages for those broad searches, not a better explanation.

Why does the rank tracker say one thing and Google’s data say another?

The tracker checks one search, from one fixed spot, once in a while. Google’s data averages every real person who searched over 30 days. They often disagree by a few places. We only call a slip real when both move the same way, which this month happened on two searches. When only the tracker moves, we watch for a week before changing a page.

Did the clicks ever follow the rankings?

Not yet. United States clicks were 10 in the August window and 11 in the latest one. Our town is small, and the searches we rank first for are typed a handful of times a month. First place on a search typed a handful of times a month is worth having, but it does not fill a calendar, and we would rather say that than hide it.

Why did you change how you count the searches you rank for?

The first count included searches that Google’s data still listed for our site even when we were not shown for them during the window. That inflates the total. Counting only searches where we were shown at least once is stricter and easier to defend. We printed both counts, for both windows, so nothing from the original case study is swapped out quietly.

What happens when one of your own rankings drops?

We check whether both sources agree. If they do, we fix the page built for that search: tighten the section that answers it, link to it from newer articles, and post about that service on our Google listing. We do not build a second page for the same search. Two of our own pages competing is how the website design slip happened in the first place.

How long before the searches without a town name start ranking?

Longer than 85 days, on our own site. Seventy-three of our 109 tracked terms still do not rank, and nearly all of them are searches with no town in them. The case study said three to six months for those. Nothing in this month’s numbers lets us shorten that, and we would be guessing if we promised you faster.

Would you show me these same numbers for my business?

Yes. Search terms, page-one terms, times shown, clicks, and the ones that went the wrong way, pulled from your own Google data. If people find you but cannot reach you, that is a separate problem, and it is the one our AI receptionist exists for. The audit covers both.

If you want to see where your own business stands, book your free audit or call (920) 383-3387. You get the numbers that went up and the ones that did not.