TrafficTriage

Example report, delivered as an email

Deep Traffic Analysis: ugursaritepe.com

This is a real analysis of our own site, shown in full so you can judge the depth before you pay. Yours covers your site and, if you send it, your Search Console data.

The verdict

Almost all of this site's Google traffic ever came from a single article: the guide on installing minor Python versions with pyenv (pyenv is a tool developers use to switch between Python versions). In your Search Console export that one page was shown 146 times, more than half of the site's 262 total. Google put it on page one and two for a while, saw almost nobody click, and quietly pushed it down to page five and beyond. Nothing new was published to take its place, so monthly impressions fell from 133 in May to a trickle now. The plumbing is fine; every mechanical check passes. The problem is that the whole site rests on one aging page.

FIX THIS WEEK

Rewrite and expand the pyenv article so it fully answers the searches people already make (installing an exact version like 3.12.7, troubleshooting, pyenv versus pipenv). Put the working commands in the first screen, update the date, then leave it alone for three weeks. It is your only page with proven demand and it is slipping away.

What we checked

  • The 8 site-level checks that every free triage runs: fake-URL handling (soft 404s), sitemap, robots rules, canonical tags (the tag that tells Google which address is the real one), one site version, page basics, whether Google gets real text when it fetches a page or just an empty shell, and index signals. All 8 came back healthy on July 28, 2026.
  • Three key pages read by hand: the homepage, the Projects page, and the pyenv article. We looked at the title, description, headline, structured data (the hidden labels that tell Google what kind of page this is), and how much real content each one has.
  • Your Search Console export for April 11 to July 10, 2026 (91 days): 262 impressions, 2 clicks, and which pages and searches produced them.

Page-by-page findings

Homepage (ugursaritepe.com)

  • The basics are in place: a clear title ("Uğur Sarıtepe - Developer & Data Engineer"), a real description, a canonical tag pointing to itself, and Person plus WebSite structured data. This is the only page earning clicks (2 of the site's 2), so nothing here is broken.
  • What matters: the homepage has 238 words and links to only one blog post, and it is not the pyenv article. Your best page, the one Google actually tested, gets no link from your most-visited page. Add a clear link from the homepage to the pyenv guide so both the reader and Google can find it.

Projects page (/projects)

  • The title is duplicated: "Projects | Uğur Sarıtepe | Uğur Sarıtepe". Your name appears twice. It looks careless in a search result and wastes room that could describe the page. Fix it to read once, for example "Projects | Uğur Sarıtepe".
  • The page is thin: about 144 words for five projects, just a name and a one-line label each. A sentence or two per project on what it is and who it helps would give it something to rank on.
  • This page has no structured data at all, while the homepage and the article both do. Adding a simple description of the list helps Google understand it.
  • One of its detail pages, /projects/personal-blog, is the quiet opportunity in your export: shown 12 times at position 3.3, which is near the top of page one, with zero clicks. The sample is tiny, but a page that high earning nothing usually means the title and description give the searcher no reason to click. Rewrite them to say what the page actually offers.

Pyenv article (/blog/install-minor-python-version-with-pyenv)

  • The structured data is good (BlogPosting, Person, and an image object) and the technical basics all pass. This is not a broken page; it is an under-fed one.
  • The title and description are generic. The title is "Installing and Managing Minor Python Versions with Pyenv", but the searches in your export are specific: "pyenv install 3.12.12", "pyenv install python version". Neither the title nor the description promises the exact outcome (install a precise version, fast), so people skip it.
  • It is shallow for the topic: about 640 words covering find, install, check, and set a version. It never covers installing a specific minor version step by step, troubleshooting a failed install, or the common pyenv versus pipenv confusion, all of which people search. Depth is what would pull it back up the rankings.
  • It has not been touched since it was published (both the published and modified dates read December 27, 2025). Google demoted it over the spring while it sat untouched. Expanding it and giving it an honest new date is the single highest-value edit on the whole site.

What the search data says

These come from the site's own Google Search Console export for April 11 to July 10, 2026.

  • The trend is fading, not shrinking. Impressions by month: April 81, May 133, June 32, and just 16 in the first ten days of July. In the last 30 days the site was shown 36 times, and 12 of those days it was shown zero times. This is a site going quiet, and the cause is one page losing its ranking, not a competitor.
  • Only 7 pages have ever appeared in Google. On a site this small the ceiling is the number of useful pages, not the tuning of the ones you have. More good pages is the lever.
  • The pyenv page is the only proven demand, and it is being demoted. It earned 146 impressions at an average position of 11.5, but the individual searches now sit at positions 45 to 94, deep on page five and beyond. Google tested it, measured no engagement, and moved it down.
  • Half of the visible searches are not your audience. About 15 of the 26 named searches are for "Sarıtepe Kampüsü", a university campus bus schedule in Istanbul that shares the surname. Those people want a bus timetable, will never click a developer blog, and should be ignored when you read your numbers. Your real footprint is even smaller than 262 impressions looks.
  • The audience is global developers. 95 percent of impressions are desktop (248 of 262) and the United States leads with 171. That fits technical Python content and tells you who to write for.

Your prioritized fix list

  1. This week: expand and refresh the pyenv article. Add the exact-version walkthrough, troubleshooting, and the pyenv versus pipenv section, put the commands up top, rewrite the title and description around the precise outcome, and update the date. It is your only page with proven demand and it is slipping.
  2. Publish one narrow developer-problem post every week. The pyenv page is the template: one specific problem, a worked fix, real terminal output. On a 7-page site, adding pages Google can rank is the only thing that moves the numbers.
  3. Link the homepage to the pyenv article. Your most-visited page currently does not point to your best-performing one, so both readers and Google have a harder time reaching it.
  4. Fix the Projects pages: rewrite the title and description of /projects/personal-blog, and clean up the /projects index. The personal-blog page ranks near the top (position 3.3) but earns zero clicks, so its snippet needs a reason to click; the index page needs the duplicated name removed from its title and a sentence per project.
  5. Ignore the "Sarıtepe Kampüsü" searches. They are a name collision with a campus bus schedule, not an audience. Do not write anything to chase them and do not let them flatter your topline numbers.

Want this for your site?

Your key pages plus your search data analyzed by hand, with a prioritized fix list for your traffic problem. 10 EUR, delivered to your email within 2 days.

Get the deep analysis