What link building actually costs, and why

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

Why price per link is the wrong unit

Almost every conversation about the cost of link building starts in the same place: what do you charge per link? It is a reasonable question and it produces a useless answer, because a link is not a manufactured object with a bill of materials. It is the outcome of a decision made by a person you do not employ, on a site you do not control, for reasons that have nothing to do with your budget.

What you are actually buying is a process with a known failure rate. Somebody builds a list of candidate sites. Somebody judges which of those candidates are worth contacting. Somebody creates or improves the thing being pitched. Somebody writes to a named human being and follows up. Most of those contacts go nowhere. The links that do arrive have to carry the cost of every attempt that did not.

Once you see it that way the pricing stops being mysterious. Two agencies quoting wildly different numbers are usually not selling the same work at different margins. They are running processes with different labor content and different failure rates, and in many cases they are producing genuinely different products that happen to share the word link.

This page describes cost structurally. I am deliberately not putting figures on it: figures go stale, they vary by market, and a number quoted out of context is how people end up buying the wrong thing.

The six things that set the price

Every serious link building engagement I have looked at prices out along the same six axes. If a proposal is expensive, it is expensive because of one or more of these. If it is cheap, one or more of these has been removed.

1. Prospecting time

Somebody has to build the list. Pulling ten thousand rows out of a link index takes minutes; deciding which two hundred of them are real, relevant, alive and worth a pitch takes days. Prospecting is where quality is decided, and it is the first thing cut when a provider needs to hit a price point. A cheap engagement usually has a list somebody else built, reused across clients.

2. Asset creation

Some tactics require something to exist before you can pitch it: a dataset, a tool, a study, a genuinely useful guide, an original photograph, a piece of expert commentary. That work is design, development, research and writing, and it is priced like design, development, research and writing. Other tactics require nothing but an email. The gap between those two starting points is enormous and it shows up in the quote.

3. Outreach labor

Finding a named contact, writing something that is not obviously a template, sending it, tracking it, following up once, and handling the reply. This is per-prospect work and it does not compress. Automation reduces the cost per send and simultaneously reduces the response rate, which means the cost per link often stays flat or gets worse.

4. Seniority of the person writing

A pitch written by somebody who understands the subject reads differently from one written by somebody following a script. In technical, medical, financial and legal subjects the difference is not stylistic — a junior writer produces something an editor can identify as uninformed in one paragraph. Seniority is the single largest lever on cost inside outreach, and it is invisible on an invoice.

5. Response rates

Cost per link is cost per attempt divided by the conversion rate. A tactic that converts at a few percent needs many times the contacts of one that converts well, and every one of those contacts costs the same to produce. This is why response rate, not price, is the number worth interrogating.

6. The tactic itself

Some tactics have a ceiling on quality and a floor on cost. Others have neither. Choosing digital PR over resource page outreach changes the cost structure of the entire engagement, not the unit price of the output.

Why a digital PR link and a directory submission are not the same product

This is the confusion that costs buyers the most money, in both directions. People pay digital PR rates for work that is really list submission, and people expect directory prices for work that is really journalism.

A directory submission is a form. Somebody finds a directory that accepts entries in your category, checks it is not obvious junk, fills in the fields, and waits. The work is clerical, the outcome is close to certain, and the resulting link is one that anybody in your industry can obtain by doing the same thing. Its value is largely that you are listed where a customer might look. As a ranking signal it is weak, because it is available to everyone and it is not evidence of anything.

An earned editorial link from a news organization is the end of a completely different chain. Somebody had to find an angle worth publishing. Somebody had to gather or produce information that did not already exist. Somebody had to write it up in a form a journalist can use inside their deadline. Somebody had to know which journalists cover that beat, reach them, and survive the fact that most of them will not reply. And after all of that, an editor makes a free decision, and the link may be nofollow, or the coverage may not link at all.

Those two products differ in acquisition cost, in certainty, in durability and in what they mean. My own profile makes the durability point better than any argument: the median referring domain stops linking after about 1,080 days, and survival scales with the quality of the linking site. Domains at Trust Flow 0 — Majestic's measure of how trustworthy the sites linking to a page are — lasted a median of 859 days. Domains at Trust Flow 61 and above lasted a median of 3,353 days, roughly four times as long. You are not just buying a link. You are buying how long it lasts.

So when a menu offers you both at prices that differ by a small multiple, something is wrong with the description of one of them. Either the directory is being oversold, or the editorial link is not what it says it is.

What a retainer actually buys

Most serious link building is sold as a monthly retainer, and buyers frequently misunderstand what they are paying for. A retainer buys capacity — a certain number of hours of a certain quality of person, applied to your site every month. It does not buy a quantity of links, and any retainer that claims to is quietly making the quantity the priority, which changes what gets built.

Within a retainer, the money typically funds:

• Continuous prospecting. Lists decay. Editors change jobs, sites go dark, opportunities get taken by competitors. A standing program rebuilds its list constantly.
• Asset production. One substantial linkable asset per quarter is a common and realistic cadence for a small team.
• Outreach cycles. The actual sending, following up and negotiating.
• Reclamation and maintenance. Recovering links that were removed and mentions that were never linked. This is the highest-yield work in most programs and the most commonly skipped.
• Measurement and reporting. Not a screenshot of a rankings tool. A record of what was attempted, what landed and what it cost.

The honest way to read a retainer is as a rate for a team, multiplied by how much of that team you get. If a provider will not tell you who is doing the work and how senior they are, you cannot evaluate the price at all.

The costs nobody itemizes

Several real costs sit inside every engagement and almost never appear on a proposal.

Tooling and data. A link index subscription, a crawler, an email finder, a verification service, a rank tracker, a project system. Individually modest, collectively a standing overhead that a provider spreads across clients and a DIY team pays alone.

Dead ends. Campaigns that produce nothing. Every experienced practitioner has run them. A provider who has never had one is either new or not telling you about them. The cost of failed campaigns is priced into successful ones, which is one reason very cheap providers cannot afford to experiment and therefore only run the tactics with the most predictable — and least valuable — output.

Your own time. Approvals, subject matter interviews, brand and legal review, getting a quote signed off. In larger organizations this is frequently the biggest single cost in the program, and you pay it, not the agency.

Reputation risk. Cheap link building is cheap because somebody is buying placements or spinning content, and both create a cleanup cost later. That cost is deferred, and it lands on the client rather than the vendor.

How to compare two quotes honestly

Put the price to one side and compare the mechanisms. Five questions do most of the work.

1. What tactic is this, precisely? Not "outreach" — which outreach. Broken links, resource pages, digital PR, reclamation, guest contributions. Different answers mean different products.
2. Who writes the pitches, and what else do they know? Ask for the person, not the process.
3. What response rate do you plan around? A provider who cannot answer this has not modeled their own business.
4. What happens in month one if nothing lands? The answer tells you whether the price is buying effort or output — and whether they will start buying placements to make a number.
5. What is the expected durability of what you build? Anyone who has never thought about link loss is measuring only acquisition, and half of all links are eventually lost.

If two quotes survive that comparison and one is materially cheaper, the cheaper one is a better deal. If only one survives it, price was never the deciding factor.