Case 40. Task and errand service — links from PBN home pages (Tier1.shop)

Владислав Бобров (Mdevostator)

Founder of tier1.shop · PBN Specialist, in SEO since 2009

★ 5.0 on Kwork4 912 orders completed1 197 reviews receivedAbout the author →

Telegram channelCases and PBN breakdownsSubscribe →1 802 followersTier1.Shop

Case from the Tier1.shop collection · 2026 · ← all cases

Result in brief: 57 links over 5 months: a two-year downward trend reversed into growth. According to Semrush, organic traffic +23% (from ~120 to 147.5 thousand visits/month), keys in top-3 +69%. According to Ahrefs weekly data, traffic +30% (from 108 to 141 thousand, peak 152.7 thousand), keys in top-3 — growth by 2.5 times (from 3 753 to 9 260).

Starting point

A large service for tasks and assignments (services marketplace). By the time of request, organic traffic had been sliding smoothly for two years: according to Semrush — from ~160 thousand visits/month at the 2024 peak to ~110 thousand by the end of 2025; according to Ahrefs weekly estimates — from a peak of 171.5 thousand (April 2024) to ~105–108 thousand by the beginning of 2026. Both services showed the same picture of a prolonged decline.

What was done

From February 16, 2026 — regular placement of links from PBN homepages in monthly batches, distributed across 10 different pages of the site (not only the homepage). All links are anchor links, 42 unique anchors for 57 placements — without mass repetitions.

Dynamics by Semrush (monthly data)

Over the placement window, organic traffic grew from ~120 to 147.5 thousand visits per month (+23%). This is not seasonality: in the same months of 2025, the site added only ~8%, and in 2024 — lost ~9% (Semrush data, Tier1.shop calculation).

Keywords in top-3 became 4 564 (+69%) — a maximum since at least the beginning of 2024. In top-4–10 — 14 000 (+43%). Important detail: the total number of keywords for the site decreased during this period (the team was cleaning the long tail of semantics) — it was specifically the top of the search results that grew.

Dynamics by Ahrefs (weekly data)

The Ahrefs weekly grid shows the turning point more accurately:

At the end of March, the site dropped to ~95 thousand (the drop coincided with the Google algorithm spring update), and from mid-April, a sharp growth began — nine weeks in a row, up to a peak of 152,7 thousand at the end of May (+41% from the start of the window), stabilizing at ~140–147 thousand. Result by July 13 — 141,1 thousand, +30% compared to the week placements began.

Top keywords by Ahrefs confirm the Semrush picture and make it more vivid: top-3 — from 3 753 at the start of the window to 9 260 (a 2,5-fold increase, and from the April minimum of 2 952 — a 3,1-fold increase); top-4–10 — from 14 839 to 23 711 (+60%).

Breakdown by target pages: how many links, what growth, how much to expect

Links went not to the homepage, but to 10 internal sections of the service. Below is the fate of each (traffic of the target pages themselves by Semrush, visits/month; section names are given by topic — paths are hidden along with the domain):

Ahrefs daily data shows timing accurate to the day. Both February star pages first dipped along with the entire site during the March update, and then rebounded synchronously — April 10–16, after ~8 weeks from placement:

The "turnkey websites" section, which received the first batch of 7 links on April 2, broke through after ~2 weeks — and grew from ~4 to 11.2 thousand at its peak (×2.7):

The role of anchor text

The difference between "took off" and "did not" correlates with the quality of the anchor list. Both stars had 100% unique anchor text with commercial intent, entity, and geo: "buy term paper", "term papers to order in Moscow", "order apartment design project", "best masters for apartment design project". Weak pages had repetitions and vague phrasing: in the general tutors section, one anchor text appeared three times out of seven links, in the courier section — twice out of six ("delivery service" — without intent and geo). The sample is small, we are not making hard conclusions, but the picture matches our rule: anchor text must be unique and assembled from intent, entity, and key with dilution by synonyms (Tier1.shop observations).

Conclusions from the case: how to order links to get an effect

Everything below is a guideline from THIS case, not guarantees: the site is large, the niche is its own, your numbers will be different. But the patterns are clearly readable.

How often and how many to order:

• In this case, links were placed in regular batches approximately once a month — from 7 to 20 pieces — over five months. Each subsequent batch layered on top of the effect of the previous one: links are permanent, they do not disappear, and the site's link profile grew step by step. This cumulative mechanic is visible directly on the graphs: a breakthrough after the first batches, acceleration after the next ones.
• Honest disclaimer: there was no "place all 57 links at once" scenario in the case, so we cannot compare the regular mode with a one-time mode — we only assert what we saw: the regular mode worked.
• A one-month pause (there were no batches in March) did not break the trend. What matters is not perfect rhythm, but that the work continues.
• The first stable shifts for the domain occurred after ~23 placed links and two months of work. Therefore, it makes sense to plan link building for at least 3 months — on a shorter horizon, the effect simply will not have time to manifest and be measured.

Which pages to place on:

• The best result was given by live commercial sections — pages that were already in the search engine index and were collecting at least some organic traffic. It was these pages that grew 3.6 and 20 times, and each required only 2–4 links.
• From this follows the distribution strategy: place 2–4 links on a page — the next ones are more profitably invested in another section, rather than continuing to strengthen the same one. In this way, for the same money, several entry points to the site grow.
• What didn't work: 11 links went to two sections that had no organic traffic at all — and gave no measurable result. If the search engine does not rank the page (problems with content, indexing, cannibalization), the links simply have nothing to strengthen. First, we bring the page into order — then we strengthen it with links.

With which anchor text:

• Anchor text is the text in which the link is placed on the donor. The following worked unique anchor text, in which the entire essence of the query is collected: what the person wants to do (buy, order), what exactly (term paper, apartment design project) and where ("in Moscow"): "buy term paper", "order apartment design project", "term papers to order in Moscow". With such anchor text, both the search engine and the person understand what the page is about.
• What didn't work: anchor text repetitions — in the weakest section, the same anchor text was used three times out of seven links (patterning, which search engines dislike), and vague formulations without intent and geo ("delivery service") — it is not readable from these which query the page is being promoted for.
• We build the diversity of the anchor list with synonyms and variations of the key — not with empty anchor text like "here" or a bare URL.

How long to wait for the result:

• Internal pages reacted after 2–8 weeks: the section that received a batch on April 2 turned around after ~2 weeks; pages with February links — after ~8 weeks (between them occurred the March Google update, which initially pulled the entire site down).
• The domain as a whole turned around after ~2 months from the start of work.
• Hence the practical advice: do not draw conclusions in the first 4–6 weeks. In this case, after the start of placements, the site managed to dip in an update — and only then emerged from it upwards. If the work had been stopped in March with the verdict "it doesn't work", no one would have seen the result.

Verifiability

The graphs above are built from service exports. Here are screenshots of the interfaces themselves as of 14.07.2026 — the client's domain is hidden per our privacy policy:

Numbers in the interfaces may differ slightly from the graphs: services continuously recalculate estimates, Ahrefs calculates on a weekly grid, Semrush — on a monthly one, and each has its own index. The direction and scale of changes, however, coincide.

How we measured

Two independent sources: Semrush (monthly dynamics, worldwide) and Ahrefs (weekly dynamics, all locations). Placement window — from the date of the first link (16.02.26) to the date of the last (10.07.26). Seasonality was checked for the same calendar months of 2024 and 2025. All figures are estimates from analytics services, not data from the client's counters.

Honestly: the client was simultaneously performing their own work on the site — links from home pages are part of the overall result here, not the sole factor. The March dip and April reversal also coincided with Google algorithm updates: link reinforcement helped the site emerge from the update upward, but breaking down the contribution by percentages would be speculation.

All cases  ·  ← Case 41 (newer)  ·  Case 39 (older) →

Position and traffic dynamics are individual and not guaranteed; we guarantee indexing and the preservation of links. * Costs — at the time of order, current prices in tariffs.

Want the same — place links from home pages: donor selection, tier-2, guarantees included in the price. Questions — @tier1shopsupport_bot.