Tiered link building

By Bill Hartzer · Hartzer Consulting · 25+ years in search, domains and expert witness work

What tiered link building is

Tiered link building is a scheme in which you build links to the pages that link to you, on the theory that strengthening those pages strengthens the link they send you.

The vocabulary is simple. Tier 1 is the set of pages linking directly to your site. Tier 2 links to tier 1, tier 3 to tier 2, and so on down. In the scheme, tier 1 is handled with some care - a guest post, a purchased placement - and then tiers 2 and 3 are blasted with whatever is cheapest, on the reasoning that those links do not point at your site and so cannot be held against it. Automated directory submissions, forum profiles, spun articles, comment runs, bookmark packages: the tactics documented elsewhere here as obsolete get a second life as tier fodder.

The theory is not stupid. It is a literal reading of how PageRank was described: a page's ability to pass value depends on the value flowing into it, so raise the value flowing into your linking pages and more arrives at yours. Somebody read the original description carefully and drew a diagram. The diagram is where it stopped being true.

It is sold as link pyramids, link wheels, tier 2 boosting and power-up packages. If a vendor's deliverable is described as links to your links, this is the page.

Tiered link building is not tiered link analysis

This section exists because the confusion is constant and consequential.

Tiered link building is a scheme. It is the manufacturing activity described above: acquiring links that point at your linking pages, in order to inflate them. It is what this page is about, and it does not work.

Tiered link analysis is a diagnostic discipline. It is reading upstream through the link graph - what links to the sites that link to you - to judge whether a link is real. It is one of the most valuable analytical skills in link building, it is taught in Module 3, and it has nothing to do with the scheme.

They share the word tier because both describe layers of the same graph. That is the entire relationship. One builds noise into the graph; the other reads the graph to detect noise. Confusing them leads people to dismiss the analysis as spam-adjacent, which costs them the best method available for telling manufactured authority from earned authority before they spend money on it.

The analysis answers questions the scheme cannot: is this prospect's authority borrowed from a network, how densely is this layer wired to itself, does this profile look assembled or accumulated. If you want that skill - and you should - go to the lessons on tiers two through five and on running a tiered link audit.

How it works, step by step

Described accurately, without supplying a recipe:

1. Acquire or identify tier 1 - the links pointing at the money site, usually purchased placements, guest posts on low-gate sites, or properties the operator controls.
2. Point volume at tier 1. Automated submissions, profile links, spun articles and comment runs, aimed at the tier 1 URLs. Because these links do not touch the money site, quality is treated as irrelevant and the operator optimizes purely for count.
3. Point more volume at tier 2. The same tooling one layer further out, where the volumes become absurd - tens of thousands of links is normal at this depth.
4. Push for indexation. Because a link on an unindexed page passes nothing, vendors bolt on an indexing service, itself usually another layer of automated links.
5. Cross-link the layers. Some operators wire the tier 2 and 3 properties to each other before pushing outward, to concentrate value first. This is the link wheel, and it produces the most visible signature of the lot.

What it costs in time and effort

Almost nothing per unit, which is the whole pitch. Tier 2 and 3 packages are priced by the thousand, because the links are produced by software against targets that accept anything. Cost per link approaches zero and vendors compete on volume, which tells you what the market believes it is selling.

The cost drivers are link count, whether an indexing service is included, and whether the operator supplies tier 1 too. Add a recurring charge for drip-feeding, which exists because a hundred thousand links appearing in one day looked wrong even to the people buying it.

What is genuinely expensive is the opportunity cost. A team that spends a quarter running tiered campaigns spends a quarter not building anything anyone would cite. I have spent billable hours separating tier fodder from signal in an inflated profile, and the client paid twice - once for the links, once for me to explain that they did nothing.

When it works and when it does not

The theory deserves to be taken seriously enough to show where it breaks.

Damping. Value passed along a link is reduced at every hop. Two hops of that reduction, through pages carrying hundreds of outbound links each, leaves an amount not meaningfully distinguishable from zero. The arithmetic that makes tier 1 matter is the arithmetic that makes tier 3 irrelevant.

The source has nothing to pass. This is the fatal one. Value flows from trusted pages, and tier fodder is by construction the least trusted material on the web. Here is the number that settles it: on my own site, 69.3% of 22,260 referring domains have a Trust Flow of 0, on a profile where no link was ever bought. Junk accumulates on its own, free, at enormous scale. If volume of zero-trust links did anything, every site on the web would already have the benefit.

You do not control your tier 2 anyway. Real profiles inherit whatever their linking sites attract - there is a network in my own second tier that I did not place and cannot remove. Depth is context, not credit.

The shape gives it away. A manufactured profile is thin at tier 1 and dense underneath, because the money went into the layer feeding the layer you can see. An earned profile is the reverse. That inversion shows up in a footer statistic, as the next section demonstrates.

Where does it ever work? In fields with no competition, briefly, and indistinguishably from what the tier 1 links would have done alone - the measurement nobody runs.

Common mistakes

• Assuming tier 2 is consequence-free because it does not touch you. It cannot be disavowed, which also means it cannot help you.
• Buying indexation services to fix the problem. If a page must be forced into an index, its links were never going to be worth anything.
• Cross-linking the tiers. The link wheel is the most detectable configuration in the category, because a self-wired layer stands out immediately in a graph.
• Pointing volume at innocent third parties. If a real publisher gave you a link, blasting their page with junk is a poor way to repay it.

A worked example

Two complete tier walks, captured the same day with the same tool - one earned profile, one manufactured. The figure to watch is links per domain: how densely a layer is wired to itself.

Read the earned column first. It runs between 1.35 and 1.96 at every depth: it wobbles, it does not spike, and no layer is noticeably wired to itself. Irregular but flat is what accumulation looks like.

Now the manufactured column. Tier 1 is 30 domains sending exactly 30 links. Every referring domain sends precisely one, and not one sends two. Real linking is lumpy: somebody mentions you, mentions you again later, the archive page carries both, the sidebar widget carries the title. A flat 1.00 across a whole tier is what buying one placement at a time looks like.

Then tier 3: eight domains, 36 links, 4.5 links per domain - more than three times the density of anything in the earned profile, on the smallest population in the table. Eight sites are not sending 36 links outward at random; they are linking to each other. That is a link wheel, and it is the scheme showing up as a number in a footer.

The deep tiers of that profile also mix a Dutch airport taxi service, NFL merchandise, a pharmacy spam site, stand mixers and dolls: unrelated subjects, one wiring pattern.

How to measure it

There is no honest way to measure whether tier 2 spending produced a return, because the claimed effect is not separable from the tier 1 links it was pointed at. What you can measure is whether a profile has been through this, and the routine takes about ten minutes per domain:

1. Load the domain in a tool that tiers the whole graph natively, rather than walking one chain by hand - a hand-walked chain describes a thread and calls it a fabric.
2. Record the all-tier footer: domains, links, links per domain.
3. Repeat for tiers 1, 2, 3 and 4.
4. Put the readings beside a comparable competitor; the numbers only mean something relative to something.
5. At each tier, scan for runs of near-identical Trust Flow and Citation Flow pairs, and for naming and extension clusters.

Stop at tier 4. Beyond that the population is enormous and shared with half the web. Treat the counts as shape rather than accountancy.

Two readings to remember: a tier at exactly 1.00 means placements bought one at a time; a small deep tier well above 2 means properties wired to each other.

The verdict

Tiered link building takes a correct observation - that a link's value depends on the linking page's own inbound links - and acts on it with the only material cheap enough to deploy at volume, which is material that has no value to pass. The theory eats itself. The execution, meanwhile, leaves the most distinctive shape in the whole link graph: a first tier at exactly one link per domain, a deep tier wired to itself, topical incoherence at depth.

But do learn to read tiers. The analysis is genuinely valuable, it is the fastest way to tell manufactured authority from earned, and it is a different thing entirely from the scheme on this page. They share a word and nothing else.