TrainingPlan.dev: 884,799 Search Impressions in 239 Days

Jan. 9–3157

August4,050

The result

TrainingPlan.dev first recorded Google Search impressions on January 9, 2026. By September 4, the checked Search Console record had 884,799 impressions and 9,061 clicks. That period covers 239 days.

The latest full 28 days had 294,127 impressions and 3,998 clicks. Compared with the prior full 28 days, impressions rose 94.7% and clicks rose 113.3%.

Measure	Observed value	Evidence boundary
Study window	Jan. 9–Sep. 4, 2026	First Search Console impression through the last checked full day
Inclusive days	239	Calendar count
Total impressions	884,799	Checked Search Console files
Total clicks	9,061	Checked Search Console files
Latest 28-day impressions	294,127	Two full periods next to each other
Latest 28-day clicks	3,998	Two full periods next to each other
Impression growth	+94.7%	Versus the prior full 28 days
Click growth	+113.3%	Versus the prior full 28 days

These numbers show strong Google discovery. They do not show revenue, AI citation share, or the effect of one template, schema field, content change, or tech update.

Public proof: Download the month and fixed-period CSV or read the method and claim limits. Raw query and page files stay private.

The link claim

Bob did not ask for links, buy links, run a guest post plan, or run a digital PR campaign for TrainingPlan.dev during this period.

That does not mean the site had zero links. People may link to useful sites on their own. Automated sites may also make links. Link indexes miss some links. The safe claim is no planned link campaign.

What TrainingPlan.dev is

TrainingPlan.dev offers public endurance training guides, plans, tools, product data, research, and choice help. It started as a new site. It did not move from a known site with strong trust.

A live check on September 10, 2026, found:

• 589 article URLs.
• 77 tool URLs.
• 678 total URLs.
• 706 image entries.

The result did not come from one viral article. It came from many linked guides and tools built with one publishing system. Page count alone does not explain the growth. An August check found that 245 of 664 sitemap URLs had no impressions in the current 28 days. Those pages went on a clear list to improve or merge.

The main idea

The site was built around five jobs that help each other:

1. Finish a real task. Articles answer clear training questions. Tools do math or help readers choose.
2. Show the answer in plain HTML. Public pages use simple HTML, working links, and clear page data.
3. Connect related work. Topic pages, author pages, research, articles, and tools give readers a useful path.
4. Make trust easy to check. Authors, methods, sources, schema, and update rules support the content.
5. Use search demand to improve. Search Console shows new questions, weak pages, and groups that need better links.

This idea fits common search guidance. It is not a secret rank formula.

What changed and when

January to early August: build useful pages and tools

The early site added training guides, content templates, data pages, tools, site links, and page details. Search Console showed its first impressions on January 9.

By August 5, a site check found both growth and problems:

• 1,931 conversational query rows had generated 8,162 impressions and 66 clicks.
• 245 of 664 sitemap URLs had no impressions in the current 28 days.
• 113 posts still had July dates, which showed update problems.

These facts helped guide the work. They were not numbers to brag about. They showed where people had questions and which pages were not being found.

August: tech, trust, and content work

The work record included:

• Main URL, robots, H1, sitemap, author, method, identity, and schema checks.
• Better site links and topic groups.
• Image, metadata, and content-template work.
• AI visit tracking and bot access checks.
• Cloudflare cache and live site checks.
• Original research and more visible editorial methods.

The final release passed 48 of 48 bot access checks. Tests covered sample Google, Bing, OpenAI, Anthropic, and Perplexity paths.

August to September: measure fixed page groups

A set group of 343 changed pages grew from 592 to 770 clicks in one week (+30.1%). Impressions rose from 41,690 to 55,330 (+32.7%). Within that group, 318 shoe and article pages grew from 538 to 725 clicks (+34.8%).

The shoe group also relied less on one topic. Gel Contend pages drove 16 of the 187 added clicks. Growth across many pages is a better sign than growth from one page.

The timing does not prove cause. Some pages had only four full days after Google checked them again. Several changes overlapped. The site was already growing. Pages aged, demand may have changed, and Google changed on its own.

Why the architecture was bot-friendly

Public content worked without a login or browser app state. Standard links, main URLs, sitemaps, author pages, and method pages made the site easy to follow. The live stack used edge delivery and clear response tests.

Bot access did not mean trusting all traffic. The site kept public pages apart from private or costly tools. It tested sample known bots on the live site.

The site had llms.txt and other discovery files. This case claims no rank or citation effect from them. They are cheap page lists, not the heart of the plan.

Why useful tools mattered

TrainingPlan.dev did more than publish articles. Its free tools help readers do math, compare choices, and plan. A tool can earn repeat use and links because it finishes a job that an article only explains.

The lesson is not “publish 77 tools.” Each tool needs one clear task, a plain guide, checked inputs, useful output, a stable URL, and upkeep. Broken or copied tools only add clutter.

How the totals were checked

The fixed total combines:

1. An all-time Search Console export through August 5.
2. Daily rows for August 6 through September 4 that do not overlap.
3. Two full 28-day periods next to each other for the growth check.

The math does not add date ranges twice. Query and page files stay private because they show sensitive site details. A future public proof set should add safe daily totals, formulas, file hashes, and images of the set Search Console periods. This would let others check the math without seeing private query data.

The current public aggregate package also exposes the monthly time series:

Period	Clicks	Impressions
Jan. 9–31	57	21,966
February	184	32,297
March	226	52,717
April	467	84,979
May	682	101,745
June	1,111	104,566
July	1,543	137,448
August	4,050	295,351
Sep. 1–4, partial	741	53,730

The monthly values add up to 884,799 impressions and 9,061 clicks. These periods show what happened. They are not separate tests.

Jan21,966

Feb32,297

Mar52,717

Apr84,979

May101,745

Jun104,566

Jul137,448

Aug295,351

Sep*53,730

What likely contributed

The proof suggests these possible causes:

• A large and growing set of distinct training questions and tools.
• Static, accessible delivery and crawlable site structure.
• Search data used to improve groups of pages.
• Clear authors, methods, research, and sources.
• Internal connections between articles, data, and interactive resources.
• A new site gaining age and more search history.

“Possible” matters here. This was not a controlled test. The changes were not random or made one at a time.

What the case does not prove

It does not prove that:

• Publishing hundreds of pages causes growth.
• A specific schema type raises rankings.
• llms.txt earns AI citations.
• Cloudflare caching raises search position.
• No links existed or links did not contribute.
• The August work alone caused the September growth.
• Search impressions produced a corresponding level of revenue.
• TrainingPlan.dev has leading visibility in AI answer products.
• Another site, market, or team will get the same result.

The result is strong proof that Bob can build the tech and content base that SEO and AEO need. AI answer results need their own fixed prompt and citation tests.

The reusable playbook

1. Create one complete useful loop

Choose one useful reader problem. Publish the guide, choice tool, proof, author, method, site links, and tracking together. Prove the system before you make it larger.

2. Measure before celebrating

Record the first result date, starting page and query data, sitemap size, lead actions, bot access, and release history. Pick the date ranges before you read the result.

3. Add distinct jobs, not keywords

Each new URL should serve a distinct reader, question, proof need, or tool result. Merge pages that give the same answer.

4. Audit production like a crawler

Test status, redirects, page text, main URL, robots rules, schema, links, sitemap, cache, and firewall from the live site.

5. Treat zero-impression URLs as research questions

A page with no impressions may be new, blocked, repeated, weak, poorly linked, or outside the test dates. Check the cause before you delete it. Merge it if it does not need its own URL.

6. Report page groups and limits

Fixed page groups make changes easier to read. Full periods next to each other reduce date errors. They do not remove seasons, search updates, site age, or work that overlaps.

Next measurements

We should compare the case with the September 4 starting point after 90 days and one year. The next release should include:

• Safe daily search totals and source file hashes.
• A dated URL and deployment timeline.
• Link checks from named data tools, without claiming the list is complete.
• Groups of indexed and sitemap pages.
• A fixed-prompt AI mention and citation baseline.
• Good lead measures when privacy and volume allow them.
• Poor results and pages that were removed or merged.

That follow-up will show if growth lasted and if the site gained AI answer visibility beyond search.

The practical conclusion

TrainingPlan.dev shows that a new site can earn strong Google discovery without a planned link campaign. It invested in useful content, free tools, bot access, site links, public trust, and steady updates based on data.

This is not a shortcut. It is a clear record of careful work over time.