Preliminary results
Snapshot from 16 August 2026, 10:35 Moscow time. 100 URLs per service.
Indexing leaderLink Indexing Bot · 72/100
| Metric | IndexBotik | Link Indexing Bot | SpeedyIndex |
|---|---|---|---|
| Indexed on August 13 | 62/100 | 71/100 | 78/100 |
| Indexed now | 67/100 | 72/100Best | 68/100 |
| Gained / lost | +6 / −1 (+5)Best | +6 / −5 (+1) | +11 / −21 (−10) |
| Dropped from index | 1Best | 5 | 25 |
| At least 3 days after Googlebot* | 67/84 | 71/82 | 55/82 |
| Cost per indexed URL | $0.0182Best | $0.0278 | $0.10 |
| Estimated spend | $1.22Best | $2.00 | ≈ $6.80 |
| Googlebot pages* | 84/100 | 82/100 | 82/100 |
| Total visits* | 169 | 246 | 188 |
| Repeat visits* | 85 | 164 | 106 |
| Visits per visited page* | 2.01 | 3.00 | 2.29 |
| Median first observed visit* | 5 h 37 min | 6 h 46 min | 6 h 17 min |
| Median indexing | 24 h | 24 h | 6 h 1 minBest |
| Score | 72 | 76Best | 74 |
I ran an independent page indexing services test to compare SpeedyIndex, IndexBotik, and Link Indexing Bot under the same conditions. This is a preliminary result snapshot: monitoring is still running, and this page will be updated as more data is collected.
Methodology correction, August 13. Server logs confirm real Googlebot requests, but this experiment had no untreated control group. The visits therefore cannot be attributed to a specific indexer, and Googlebot metrics no longer declare a winner.
What was compared
The public comparison includes three services: SpeedyIndex, IndexBotik, and Link Indexing Bot.
The test used 20 donor sites that I own. Serparium was not used for publication. On every donor, there were 15 public test pages: 5 pages assigned to each of the three services.
In total, this public comparison contains 300 URLs: 100 URLs per service. Each URL was sent only to its assigned service.
Methodology
- All articles were produced with one fixed model configuration, one prompt, and the same humanization pass so content conditions were consistent.
- Googlebot logs were collected every 10 minutes.
- Historically, a visit counted after reverse and forward DNS verification. The August 13 audit found that all 121 stored IPs also belong to Google’s published common crawler ranges.
- The revised collector counts only a 2xx response from the test page, also checks the official Google range, and stores the response code, range version, and log-line hash. A 301 response no longer counts as a separate visit.
- Google index presence was checked through XMLStock by exact URL.
- Each URL was submitted only to the service it had been assigned to.
Metric definitions
- Median first observed visit* is the historical estimate from submission to the first Googlebot log record among visited pages. It confirms a server observation, not indexer influence.
- Median indexing is the median time from submission to the first scheduled XMLStock check where the exact URL was found, among indexed pages. It is not the exact second when Google indexed the page.
- Repeat visits means total verified Googlebot visits minus the number of pages visited at least once.
- Visits per visited page means total visits divided by pages visited at least once.
- Cost per indexed URL means the estimated test spend for a service divided by the number of exact URLs found in the index.
- Dropped from index means URLs previously found in the index but absent from the latest conclusive check.
- Score is a 0–100 composite: 80 points for indexed share and up to 20 points for indexing speed. Googlebot no longer affects the score.
* Googlebot metrics from this first run are observational. The old collector could count a 301 response and the following 200 response as two visits, so total and repeat visits are historical estimates. The first-visit median changes by only about one second, but complete raw lines for the whole period are no longer retained.
How cost was calculated. Prices were checked on August 13, 2026. The calculation uses base prices without volume discounts: SpeedyIndex at $0.10 per confirmed indexed URL, IndexBotik at $0.0122 per submitted URL, and Link Indexing Bot at $0.02 per submitted URL. SpeedyIndex has an estimated spend of ≈ $6.80 for 68 found URLs. IndexBotik cost $1.22 for 100 submitted URLs, or $0.0182 for each of 67 found URLs. Link Indexing Bot cost $2.00 for 100 URLs, or $0.0278 for each of 72 found URLs.
Services and pricing sources: SpeedyIndex, IndexBotik, and Link Indexing Bot; the IndexBotik cost comes from the test receipt.
How to read the data
In the August 16 snapshot, Link Indexing Bot moved into first place with 72 indexed URLs out of 100. SpeedyIndex has 68 and IndexBotik has 67. Compared with August 13, SpeedyIndex is down 10 URLs net, IndexBotik is up 5, and Link Indexing Bot is up 1.
Server logs observed Googlebot requests in all three cohorts. For example, a successful 200 response for an IndexBotik cohort page was recorded on August 10 at 05:13:47 Moscow time; one for a SpeedyIndex cohort page was recorded at 05:54:05. These records prove a visit occurred but do not show how Google discovered the URL.
Causality cannot be determined in the first run: pages were public and there was no untreated control group. The next run will include pages not submitted to any indexer and excluded from sitemaps, feeds, and internal links.
Current conclusion
Link Indexing Bot leads indexed URLs and total score; IndexBotik leads retention, cost, and estimated spend; SpeedyIndex leads median indexing time. Green highlighting is limited to indexing and cost metrics. Googlebot metrics remain visible for transparency without declaring a winner.
The experiment is not final. I will continue monitoring logs and index checks, and this page will be updated when the next stable data snapshot is available.