Author: Vladislav Bobrov, founder of Tier1.shop. In SEO and webmastering since 2009.
External links remain one of the strongest ranking factors: according to a Backlinko study of 11.8 million search results, a page in the first position of Google has on average 3.8 times more backlinks, than positions 2–10. The question is not whether links are needed, but how to obtain them. Four working channels: exchanges, crowd marketing, outreach, and PBN. Each has its own economics, its own risks, and its own percentage of "dead weight". Below is an honest comparison based on research data from 2024–2026 and our own Tier1.shop experience: thousands of placements and a network of more than 7 000 donors. Here we compare only link-based methods; a full map of all channels — including media, guest posts, and mentions, with 2026 prices — is in the analysis of external website promotion.
Four methods: what they are
Link exchanges
Sape (since 2007), Miralinks, GoGetLinks, and others — intermediaries between site owners and link buyers. Pros: accessibility and automation. Cons are fundamental: the link almost always goes to an internal page of someone else's site, donor quality is mixed, and the public catalog of sites is visible to both competitors and search engines. It was mass rental exchange links with commercial anchor text that became the target of the Yandex "Minusinsk" filter in 2015 — at that time, hundreds of sites promoted primarily through exchanges saw a drop.
Crowd marketing
Links from comments, forums, review sites, and Q&A-platforms. Pro — diversity and a "natural" profile appearance. Cons: minuscule weight (deep pages without link mass), frequent noindex/nofollow for profiles and comments, moderation cleans out posts, and the indexing of such pages is a lottery: as a reminder, according to IndexCheckr, 61.94% of web pages do not enter the Google index at all, and deep forum threads are first in that queue.
Outreach
Manual negotiations with site owners regarding a guest article or mention. Donor quality can be high, but the economics are harsh. Backlinko analyzed 12 million outreach emails: only 8.5% receive a response, and link builder surveys show an average success rate of ~7.9% — approximately one link for every 12–13 emails sent. Prices: according to BuzzStream (26 000+ sites), a guest post directly costs on average $364, via an agency intermediary — $1 459; according to an Editorial.link survey (518 SEO specialists), the acceptable price for a quality link is considered to be ~$509, and since 2022, prices have risen by 20–35%. Plus, weeks of correspondence for each link. A full breakdown of the method — response funnel, RF-2026 prices, and when outreach is justified — is in the separate article on outreach. What you get for your money — the economics and weak points of the format itself — is analyzed in the article on guest posts.
PBN (Private Blog Network)
A network of sites created on aged drop-domains specifically for link placement. The key difference from all the methods above: the link is placed on the home page — the strongest page of the site, where all internal weight and the domain's accumulated external link mass converge. The second difference is control: over the donor's topic, the anchor text, the link's surroundings, and the placement duration. What such a network consists of and how much it costs to build and maintain is in a separate breakdown PBN networks from the inside.
Why the placement location matters more than the method
Let's formulate the thesis that we have been testing for years on our own placement analytics:
A link from the home page of a donor that is weak in metrics usually works better than a niche or guest post link from an internal page of a "trusted" site with high DR/TF. The reason: the home page has incoming external mass and huge internal weight, while an internal page of even an authoritative site has almost none. One link from the home page often gives more than dozens of guest posts.
Exchanges, crowd marketing, and classic outreach in 95% of cases place you specifically on internal pages. Then a cascade of problems occurs, which we analyzed in the article on the indexing of purchased links: deep pages of weak sites are not crawled (Botify: Google crawls only ~40% of significant URLs monthly for unoptimized sites), are not indexed (88% of indexing refusals are due to quality, according to Indexing Insight), and drop out of the index (21% of indexed content is de-indexed). A link that is not in the index transmits nothing — regardless of the method used to obtain it.
A detailed breakdown of the mechanics of home page weight is in a separate article: why links from home pages weigh more than internal ones.
Comparison of methods: table
| Criterion | Exchanges | Crowd | Outreach | PBN from home pages (Tier1.shop) |
|---|---|---|---|---|
| Placement location | Internal pages | Deep branches/comments | Internal (articles) | Main pages |
| Passed weight | Low/medium | Minimum | Medium | Maximum |
| Link indexing | Not guaranteed | Lottery, often noindex | Usually ok, not guaranteed | Guaranteed, 1–14 days |
| Acquisition speed | Days | Weeks | Weeks–months (response to 8.5% of emails) | Days (up to 10 business days for us) |
| Control of anchor and environment | Partial | Low | Partial | Full |
| Donor relevance | As luck permits | Various | High | High (selection based on acceptor) |
| Durability | Rental: removal upon non-payment; owner may remove | Moderation removes | Owner may remove | Controlled (year/forever) |
| Visibility to competitors | Public catalogs | Public platforms | Public sites | Private network, not on exchanges |
| Price per link | ≈500–2 000 ₽ forever (Trust 30+) + risk of spam | ~50–100 ₽/pc in packages | $364–1 459 | 2 000 ₽/year · 4 000 ₽/forever |
Regarding crowd links at “50 ₽ per link,” it is important to fully understand the economics: on Kwork, a package of 15–20 crowd links actually costs 750–1 000 ₽, and a run of 500 links costs 2 000 ₽. But the price of a working link is calculated differently: if only a third of the package is indexed and survives, and the weight of each is close to zero, you didn't pay 50 ₽ per link, but an unknown amount for a lack of effect. Cheap does not mean profitable; in link building, this rule works without fail.
What changed in 2026: links are read by more than just search engines
A fourth consumer of your links has appeared — neural networks. According to an Ahrefs study of 75 000 brands, visibility in AI responses correlates most strongly with brand mentions in page text (0.664), rather than the raw number of backlinks (0.218). Semrush, across 1 000 domains, adds: nofollow links for AI visibility work almost on par with dofollow — the language model “reads” the text, not the attributes.
This changes the placement requirements: the ideal signal unit is now an indexed thematic page where the link is placed inside meaningful text and mentions the brand. A forum signature or a line in an exchange footer does not meet these requirements; a link in a thematic paragraph on the homepage meets them completely. We analyzed how to get into neural network responses in a separate article with data.
When which method makes sense
PBN from homepages — the foundation of the profile: young sites (fast mass accumulation), competitive niches, pages with positions in the top-100 that need a push. The most noticeable response comes specifically from young sites for low-frequency queries (source: practice and client reviews of Tier1.shop — “on a young site for low-frequency queries, there was a good boost”).
Crowd — as a way to dilute the profile for naturalness, not as a source of weight. Our clients do exactly that: “I dilute it a bit with crowd-forums for naturalness” — quote from a real dialogue. It is reasonable to spend the remaining budget on crowd, not its core.
Outreach — when single image placements in media and industry publications are needed (a PR task as much as SEO), and there is a budget of $400–1 500 per link and weeks for correspondence. If the goal is specifically media, see the separate breakdown of links from media and Digital PR: how much publications cost and what they actually give to SEO.
Exchanges — honestly: we don't see a niche where renting internal links from a public catalog would be the best choice. If the budget is minimal, the same money spent on 1–2 links from homepages will give a more measurable effect than a scatter of rentals.
The working combination we recommend: links from homepages as the foundation + crowd as dilution + targeted outreach for image. And the distribution rule: statistically, 10 links to 10 different pages are more effective than 10 to one.
Risks and how to mitigate them
Let's be honest: every method, including PBN, has risks. The difference is whether they are manageable.
- Anchor over-optimization is risk #1 for any method. A duplicate incoming anchor when placing from homepages gives a chance of page penalization higher than 50% (source: Tier1.shop placement analytics). The rules are simple: a unique anchor for each link, repeating a previously used anchor — no more than once a month, the same target page — no more than three times a month. How to compile an anchor list — we analyzed separately.
- Network quality. Key questions for a PBN provider: are the sites interlinked (interlinking, especially in a "ring", is the main path to a network ban; we do not interlink donors), how many outgoing links are on the homepage (we have a strict limit of 5 + 1 sitewide link), who are the neighbors (we place theme to theme and do not place "grey" niches next to white ones), is the network in exchange catalogs (ours is not).
- Empty metrics. Do not choose donors by DR/TF/Trust — this is synthetic data that can be manipulated and is weakly linked to the effect. Working indicators: unique referring domains (our threshold is RD ≥ 20), thematic relevance of the donor to the anchor and acceptor, indexing of the homepage, domain mentions in Google. I will give an expert recommendation for those who still purchase article links not from homepages: the RD ≥ 20 rule works there too, but apply it to page, not to the domain — at least 20 dofollow links should lead to the specific page containing your link. Without this, the internal page simply does not gain the weight that it could pass to you.
- Pace. Not 100 links in a day, but a steady cadence: 1–3 links to the promoted page every 3–4 weeks. A smooth increase is natural for both Yandex (especially relevant after "Minusinsk") and Google.
- Links do not cure on-page. If the site does not respond to link building, honest advice is to redistribute the budget in favor of internal optimization rather than buying more.
Frequently Asked Questions
Is it safe to use PBN?
With the correct network construction — yes. Sanctions are given to networks with interlinked donors, grey neighbors, and anchor spam. Our donors are not interlinked, the outgoing limit is 5 + 1 sitewide link, neighbors are only thematic, and regarding anchors, we warn clients about the risks of duplicates ourselves. Manual sanctions in the RU segment are rare, and we will replace your link in any case.
How does a PBN link differ from an exchange link in essence?
By placement and control. An exchange almost always sells an internal page of someone else's site without indexing guarantees and with a public catalog of donors. PBN placement from the homepage means maximum page weight, thematic selection, control over the anchor and environment, guaranteed indexing, and network privacy.
How much does a working link cost in each method?
Do not look at the price list, but at the cost of a link that is indexed and passes weight. Outreach: $364–1 459 per placement + weeks of correspondence (response rate of 8.5% of emails). Crowd: pennies per piece, but after the indexing and weight filter, the effect tends toward zero. Exchanges: hundreds-thousands of rubles for an internal page without guarantees. Links from homepages: 2 000 ₽/year or 4 000 ₽ lifetime — with an indexing guarantee, meaning the stated price is the actual price of a working link.
How many links are needed for a result?
Minimum program: 1 link to the target page every 3–4 weeks. Maximum program: 1–3 links per page with the same frequency. Young sites usually need 10–20 quality links for the first noticeable shifts; in competitive niches, the count goes into the dozens over several months.
Can methods be combined?
They should be. The foundation is links from homepages, dilution is crowd and non-anchor links, and targeted outreach is for image-based platforms. This makes the profile look natural and covers all types of signals, including mentions for neural networks.
Why don't you show donors before the order?
For reasons of network security and client anonymity: a public catalog is a vulnerability of the exchange model, which we consciously do not reproduce. Instead, every placement is delivered with a report: the URL of the donor's home page + a screenshot of the link location — you verify everything post-factum, including indexing.
Briefly about the main points
The link acquisition method is secondary; what is primary is where it stands and whether the search engine sees it. Exchanges, crowd, and outreach in the vast majority of cases provide internal pages with non-guaranteed indexing — hence the "drained budgets." A link from the home page of a thematic donor wins in terms of weight, indexing, control, and — considering the percentage of dead links in alternatives — in real price per working link. The optimal strategy for 2026: a foundation of links from homepages + crowd for diversity + targeted outreach, with unique anchor text and a steady pace.
Prices for all formats for 2026 and five typical budget-drain scenarios are collected in a separate guide on buying links.
Try it simply: place your first link from the main page — with an indexing guarantee and a report. Questions — to the support bot @tier1shopsupport_bot, we will answer.
Data sources: Backlinko (analysis of 11.8 million Google results; study of 12 million outreach emails), surveys of link builders Aira/OutreachDesk, Ahrefs (cost of paid links and guest posts; link rot study on 2+ million domains; AI visibility study on 75 000 brands; content study on ~14 billion pages), BuzzStream (guest post prices, 26 000+ sites), Editorial.link (survey of 518 SEO specialists, 2026), uSERP State of Backlinks (survey of 800 specialists), IndexCheckr (indexing study on 16 million pages), Indexing Insight (1.7 million pages, 2025), Botify (crawling budget of enterprise sites), Pew Research Center "When Online Content Disappears" (2024), Semrush (backlinks and AI search, 1 000 domains), Princeton GEO paper (KDD 2024), Seer Interactive (CTR and SearchGPT citations), official Google statements (John Mueller) and Google Search Console help, history of the Yandex "Minusinsk" filter (2015, analyses by SEO.ru and SiteClinic), as well as Tier1.shop's own analytics: thousands of placements, a network of 7 000+ donors, a corpus of client dialogues 2018–2026.
