Case from the Tier1.shop collection · 2026 · ← all cases
For a promo code aggregator, we placed 78 links from PBN home pages. The links led to 50 pages of the acceptor site: the home page and 49 internal pages. During and after the campaign, the site grew significantly across several sources: Semrush estimates the final organic traffic to be 5.1 times higher than at the start, Ahrefs — 1.7 times, and Keys.so records a 3.3-fold increase in the number of queries in the top 3. In this case, we show the client's real result, but we do not present it as a laboratory experiment: the site simultaneously acquired other external links, Google updates were occurring, and the acceptor site pages may have changed.
Result in brief: 📈 Semrush RU: organic traffic grew from 2 859 to a peak of 35 520 visits/month, and in the last closed month it was 14 685 — 5.1 times higher than the start. 📈 Keywords in the top-3 became 109 versus 16, in the top-4–10 — 951 versus 244, total visibility in the top-50 grew from 1 906 to 9 204. Across 49 internal pages of the acceptor site to which our links led, Semrush recorded total traffic growth from 19 to a peak of 2 236 and 555 in the last closed month. At the same time, the result has mixed attribution: Keys.so discovered other external links leading to 44 of these 49 internal pages.
🔗 78 links from PBN home pages: 29 links for 1 year + 49 eternal
📅 28 July 2025 → 5 March 2026 — 8 months
⚙️ Links led to 50 pages of the acceptor site: the home page and 49 internal · 72 out of 78 anchor texts are unique (92.3%)
💰 254 000 ₽ for the entire period, on average ~31 800 ₽/month
Starting point
The site collects promo codes, coupons, and promotions of stores and online services. The continuous observation history in the Russian Semrush database begins in January 2025. The average for six months before the campaign was 2 798 visits/month, but individual months fluctuated from 548 to 5 368. This was a young and already volatile site, not a zero domain with a flat control line.
We do not consider seasonal adjustment: the continuous history is less than 24 months, therefore the calendar season cannot be reliably separated from the rapid growth of the project itself.
What was done
Over eight months, we placed 78 links from PBN home pages leading to 50 pages of the acceptor site. In July and August, the first batches of 17 and 13 links were released, and in September — another eight. After an October pause, placements continued steadily: eight links monthly from November to March.
Links were distributed widely. Of the 50 acceptor site pages, 38 received one link, seven — two, three — three, one internal page — four, and the home page — 13. An anchor text is the visible text of a link. Here, 72 out of 78 formulations were unique; only four groups repeated two-three times. We hide brands and exact URLs so that the client cannot be identified.
Traffic, batches, and accumulated mass
- 📈 The first jump came quickly, but was not sustained. After the first 30 links, traffic in August grew to 8 303 visits/month, then during the August 2025 Spam Update the estimate returned to 2 942.
- 📈 The steady rise began after four-five months. In December, Semrush showed 6 044 visits, in January — 9 794. This turning point coincided with the December 2025 Core Update, therefore we are talking about a coincidence in time, not a proven effect of a single factor.
- 📈 The peak came after the completion of placements. The last eight links were placed on March 5; Semrush estimated March at 24 065 visits, and May at 35 520.
- After the peak, the site rolled back, but not to the starting point. In July, Semrush estimated traffic at 14 685 visits/month — 5.1 times higher than the starting value. According to Semrush, the April, June, and July rollbacks coincided in time with confirmed Google updates; below we check these coincidences via Ahrefs.
How search traffic changed around Google updates
To avoid judging an update by a single random point, we compared Semrush and Ahrefs using different, pre-defined methods. For Semrush, we took the nearest monthly estimate before the start of the update and the first monthly estimate after its completion. For Ahrefs, we compared the median traffic for seven days before the start of the update with the median for seven days after completion. The "Our links" column indicates how many links had already been placed before the update and how many were added during its rollout. Dates are cross-referenced with the official Google Search Status Dashboard. This is a time correlation: the status shows only the direction of the estimates of the two services, not the impact or quality of the update itself.
| Google Update | Semrush, before → after | Ahrefs, 7-day median before → after | Our links | Repeated anchor text by the end of the update | What is visible |
|---|---|---|---|---|---|
| August 2025 Spam 26.08–22.09.2025 |
8 303 → 2 064 −75% |
12 915 → 5 597 −57% |
30 before the update +8 during |
No | Both estimates decreased. A new batch of links was not accompanied by immediate growth. |
| December 2025 Core 11–29.12.2025 |
2 003 → 9 794 +389% |
5 118 → 8 552 +67% |
46 before the update +8 during |
Yes — 1 repeat | Both metrics increased. The update, the accumulated link mass, and the new batch cannot be separated. |
| March 2026 Spam + Core 24.03–08.04.2026 |
24 065 → 15 519 −36% |
39 065 → 27 446 −30% |
all 78 before the updates 0 during |
Yes — 6 repeats | Both estimates decreased. Links did not isolate the site from the drop; two nearly adjacent updates cannot be separated. |
| May 2026 Core 21.05–02.06.2026 |
35 520 → 22 556 −37% |
28 268 → 37 954 +34% |
all 78 before the update 0 during |
Yes — 6 repeats | Services provide conflicting assessments. A connection between traffic changes and the update cannot be established based on these data. |
| June 2026 Spam 24–26.06.2026 |
22 556 → 14 685 −35% |
30 917 → 35 319 +14% |
all 78 before the update 0 during |
Yes — 6 repeats | Services provide conflicting assessments. A connection between traffic changes and the update cannot be established based on these data. |
Why February 2026 Discover is not in the table. Google describes it as an update to the systems that show materials in Discover, initially only for English-speaking users in the US. The table above refers to regular search traffic in the RU database, so including growth during the Discover update would be methodologically incorrect. The calculation for this window is saved in the case's raw data, but is not used in the web search conclusions.
How repeated anchor texts were counted. The column is cumulative: for each update, all our links placed by its completion date are taken into account. A repeat is considered every subsequent occurrence of the same anchor after the first, regardless of case or extra spaces. There were no duplicates within each individual task; all six repeats occurred between different tasks.
How the table is linked to the graph. The December 2025 Core row refers to the same rise visible on the Semrush graph: 2 003 visits at the November point before the update → 6 044 at the December point during the update → 9 794 at the January point after its completion. The table shows the extreme points 2 003 → 9 794. The Ahrefs values 5 118 → 8 552 are not separate points on the graph, but seven-day medians before and after the update.
What is visible regarding the connection with placements. During the August spam update, the site fell according to both sources, although eight links were added during the rollout. During the December core update, the site grew according to both sources: 46 links were already in place by the start, and another eight were placed within the update window. This is a positive temporal correlation between link accumulation and growth, but it does not allow for the separation of the influence of links from the update and other work on the site. The cleanest negative coincidence is March–April 2026: all 78 links were already in place, there were no new placements, and both sources showed a decrease of approximately one third. Consequently, the preserved link mass does not guarantee protection against an algorithmic drop.
June 2025 Core Update ended 11 days before our first link, so it refers to the starting point, not the campaign result. August 2026 Spam Update is not included in the table: the last Ahrefs point is dated August 11, and the Semrush monthly point is August 15, meaning both series end before the start of the update on August 18.
Positions held up better than traffic
After the May peak, the traffic estimate decreased, and search visibility did not return to the baseline. Top-3: 16 → peak 162 → 109. Top-4–10: 244 → peak 1 413 → 951. The number of queries in the top-50 reached 9 571 in June and remained at 9 204 in July compared to 1 906 at the start.
This means that the result was not based on a single random high-frequency query: noticeably more keys of various levels appeared in the search results. However, domain positions alone do not allow for the separation of the contributions of links, content, technical changes, and algorithms.
Ahrefs: growth direction is confirmed, scale differs
Ahrefs shows the same general sequence: autumn slump, winter recovery, a strong first quarter of 2026, and a finish higher than the start. On the day the campaign began, the estimate was 15 864 visits/month, reaching 51 542 on March 2, 2026, and remaining at 26 654 on August 11 — 1.7 times higher than the start.
We do not directly compare absolute Semrush and Ahrefs values: the services have different databases and estimation models. The breadth of the result also diverges. Ahrefs shows an increase in top-10 queries from 979 to 2 680, but a slight decrease in the overall top-50 from 3 837 to 3 509. Therefore, the general conclusion from the two sources is formulated as follows: top positions and organic traffic are higher by the end of the period than at the start.
What happened to the 49 internal pages of the acceptor site
The home page of the acceptor site did not enter the Semrush exact-URL export, so its metrics are not replaced by the traffic of the entire domain. For all 49 internal pages to which our links led, the same monthly series is preserved:
| Slice | Total traffic, visits/month | Queries in top-50 |
|---|---|---|
| Before the campaign, July 2025 | 19 | 65 |
| End of campaign, March 2026 | 864 | 395 |
| Peak, May 2026 | 2 236 | 427 |
| Last closed month, July 2026 | 555 | 528 |
Of the 49 internal pages of the acceptor site, 19 grew by the closed July, two decreased, and 28 remained unchanged; 15 pages rose from zero. Simultaneously, 29 internal pages ended July with a zero traffic estimate, and 13 never left zero throughout the entire window. This is an important practical result: wide link distribution did not make every page successful, but some of the internal pages to which our links led grew, and their overall visibility expanded from 65 to 528 queries in the top-50.
Separately, five current leading pages of the acceptor site, which provide 3 619, 1 973, 787, 544, and 269 visits/month, did not receive our links. This means the site's organic traffic grew beyond the list of pages to which our links led. Possible mechanisms include a general increase in domain authority, new or updated content, and internal linking; it is impossible to divide them by percentages without data from the owner.
Keys.so: traffic, positions, and parallel external links
We count referring domains (RD), not link rows. Keys.so sees 83 domains with active links: 21 match known donors for Tier1.shop, another 62 do not match. The origin of these 62 domains is unknown — we do not claim that the client purchased links on them.
In the export, 8 933 active rows fall on 62 third-party domains, but we do not use this number as the quantity of independent links or a measure of link weight. Two domains create 8 391 rows, and eight domains create 8 866 rows, which is 99.2% of the total volume. Such a ratio often occurs due to mass template or automatically created links, and less frequently due to sitewide link blocks on many pages. The DR of these eight domains according to the Keys.so export is in the range of 0–11. Meanwhile, only one active row was found for 47 out of 62 domains.
Superficial analysis of page types. Based on the URL, titles, and export structure, 41 domains lead from pages formatted as articles or news; four from forums, social, and user pages; three from profiles, app directories, or maps; eight relate to mass repeating template pages; the type for another six cannot be reliably determined. This is a classification of the placement form, not proof of editorial quality or SEO weight.
Additional domain check in Semrush. For all 62 third-party referring domains, we collected current domain metrics: for 49, Semrush does not see estimated organic traffic, for six the estimate is 1–99 visits/month, for five it is 100–999, and only for two it is 1 000 or more. Authority Score is distributed more widely: 19 domains are below 10, nine are 10–19, 23 are 20–39, and 11 are 40+. Particularly telling is the array of eight repeating sources that create 99.2% of the rows: seven out of eight have zero estimated traffic, the average Authority Score is 4, and the range is 2–12. This confirms that a large number of rows cannot be equated to a large number of strong links. Semrush does not determine the type of a specific source page, and zero estimated traffic or a low Authority Score do not in themselves prove zero SEO weight.
When they appeared in the database. Keys.so first detected six third-party domains before our campaign, 23 during it, and 33 after the last Tier1.shop link. The detection date by the index can lag significantly behind the actual publication date, so these numbers provide context but do not prove the exact sequence of placements.
Keys.so gives a more modest estimate of organic traffic: 3 072 in July 2025 and 3 544 in July 2026 (+15%). At the same time, top positions are growing noticeably: top-3 from 540 to 1 787, top-10 from 9 901 to 21 576, top-50 from 60 902 to 79 565. History also shows contextual advertising before and during the campaign; this is parallel marketing activity, but not proof of the advertising's influence on organic growth.
To check for a possible lag in Keys.so, on September 3, 2026, we opened one URL specified by the service on each of the 62 third-party domains. 54 pages responded with code 200, and on 49, a clickable link to the acceptor site remained; on five available pages, the link was not confirmed, and another eight URLs could not be checked due to response errors. This is a sample of one URL per domain, not a re-crawl of all rows in the export.
For page-by-page analysis, coverage is more important: additional external links were found on 44 out of 49 checked internal pages of the acceptor site. These 44 internal pages account for 540 out of 555 total visits. On five internal pages for which Keys.so found no other external links, total traffic grew from 0 to 15 visits/month; by the end of the period, non-zero traffic remained for one of the five pages.
How we interpret this: the client's result is real, including the growth of some internal pages of the acceptor site to which our links led. But the statement "exactly 78 Tier1.shop links independently created all the growth" is not supported by the data. The correct formulation is — the site grew during and after the Tier1.shop campaign against a backdrop of parallel external reinforcement and other work.
Similarweb: three-month snapshot of total traffic
Similarweb is available only for May–July 2026, so it has no "before links" point. Total visits were 245 682 → 123 333 → 153 599. For the available period, the service attributes 60.3% of visits to organic search, 22.3% to direct visits, and 14.9% to referral links. These data help understand the current scale and confirm the drop after May, but do not participate in the assessment of growth over the entire campaign.
Conclusions from the case
How often and how many. After the starting batches of 17 and 13 links, the campaign moved to a rhythm of eight links per month. The first spike appeared after one month, steady growth after four-five months, and the peak two months after the last placement. There was no control comparison of one large batch versus regular placements on this site, so one scheme cannot be considered better than the other.
Which pages. Links led to 50 pages of the acceptor site: the home page and 49 internal pages. Meanwhile, 38 pages received only one link each. The entire domain and 19 out of 49 internal pages grew, but traffic was distributed unevenly: 29 internal pages ended the window with a zero Semrush estimate. Wide distribution creates more growth points, however, it does not guarantee a result for every page by itself.
With which anchor texts. 92.3% of anchor text was unique, there were no clustered repetitions. Such a profile accompanied the site's growth, but we found no clean comparison that would allow us to attribute a specific result to a specific type of anchor text. We still do not recommend repeating one incoming anchor text for one page: it is safer to use different natural formulations and non-anchor options.
How long to wait. The first jump appeared in about three weeks, but did not hold. We waited about five months for a steady level above the starting point according to Semrush and Ahrefs — it formed in December 2025. The peak came approximately nine and a half months after the first link and two months after the last placement. This is an observation of one site, not a promise and not a guarantee of the timeframe.
All conclusions are benchmarks from this case. For a different site, niche, and set of parallel works, the result will differ.
Verifiability
The domain, URL, and page names are hidden per the privacy policy. Graphs are built based on saved exports from Semrush, Ahrefs, Keys.so, and Similarweb; the original series and page-by-page data are stored in the case archive. Below is an anonymized screenshot of the Semrush interface and two Ahrefs screenshots. The client's domain and address bar are hidden.
Semrush Domain Overview: organic traffic and positions over two years.
Ahrefs Site Explorer: organic traffic over five years, weekly scale.
Ahrefs Organic positions: dynamics of queries by position ranges.
How we measured
The source for placements is the Tier1.shop record: 78 completed links for one client, which led to 50 pages of the acceptor site — the homepage and 49 internal pages; the date, type, and anchor are saved for each link. Eight incomplete placements are excluded. Domain traffic and positions are from Semrush RU and Ahrefs RU. Separate Semrush series are saved for all 49 internal pages of the acceptor site; homepage metrics were not replaced by domain data. Keys.so was used as an additional source for organics, advertising, and incoming links: we count third-party links primarily by referring domains, and we use the rows only to check concentration. Page types are determined superficially by URL, Title, and export structure. Similarweb was used as a three-month snapshot of overall traffic and channels. The technical sample consisted of 14 pages of the acceptor site: all 14 respond with code 200, have self-canonical, and are allowed for indexing. August 2026 is not used as an open month for final traffic and position metrics.
Honestly
- This is not a laboratory experiment. We show the actual result of the client during and after the campaign, but we do not promise that all growth was created solely by our links.
- The site had parallel external reinforcement. Keys.so found 62 additional referring domains that did not match the known Tier1.shop list; the origin of these links is unknown.
- Some inflection points coincided with Google updates. Semrush and Ahrefs simultaneously show a drop during August 2025 Spam, growth during December 2025 Core, and a drop after March 2026 Spam/Core. For May Core and June Spam, the service directions diverge, so we do not claim a connection with these updates.
- Sources diverge by scale. In the last closed month, Semrush shows traffic 5.1 times higher than the start, Ahrefs — 1.7 times, Keys.so — by 15%. The common signal for them is the strengthening of top positions and a final point higher than the starting one.
- Not every internal page of the acceptor site grew. According to Semrush, 19 out of 49 internal pages grew, 29 finished July with a zero traffic estimate. A zero in a third-party service is not equal to the absence of real transitions.
- Only the site owner sees real clicks. Semrush, Ahrefs, Keys.so, and Similarweb use their own databases and models; accurate data is found in Google Search Console, Yandex.Webmaster, and the client's internal analytics.
* Costs are calculated based on prices at the time of orders. Positions and repetition of the result are not guaranteed.
