The Real Cost of an Editorial Link in 2026: We Opened Our Invoices

In 2018 we published a pricing study of sponsored posts across 229 websites, because at the time nobody would say out loud what placements cost. Eight years later the industry has repriced itself at least twice, the surveys still round up, the rate cards still round down, and actual invoices are still nowhere in public.

So this is the successor to that study, built the only way we trust anymore: from what we actually paid and what we actually charge. One number at a time, with the embarrassing parts left in.

The headline number: $455 per link, invoices included

Across one 12-month engagement we delivered 12 links for $5,463.99 total. Divide it out and you get $455 per link, and 33 cents we’re rounding away.

Three things about that number, so it can’t be misread:

• Every publisher had real, verified traffic. We check traffic before we pitch and we check the live link daily after it goes up. That’s the boring half of our process, and it’s why we can publish this number without an asterisk about where the links live.
• Two of the twelve dropped during the year. Both were replaced free under our 12-month guarantee. The cost of finding and placing those replacements is inside the $5,463.99, not billed on top and not quietly excluded to flatter the average.
• It’s one engagement. One client, one niche, one year. Twelve links is a ledger, not a dataset, and we won’t pretend otherwise. We’re publishing it anyway, because the alternative is that the only public numbers stay survey answers and vendor rate cards, and both have an incentive problem.

The guarantee, for the record, covers the link existing. It does not cover rankings. Nobody honest guarantees rankings.

Where $455 sits in the 2026 market

The best public pricing research we know of puts legitimate editorial links at $500 to $900 and up (Reporter Outreach, link-building pricing). The same firm’s State of Link Building survey of 500 practitioners puts the average acceptable price for a quality link at about $509 (Reporter Outreach, State of Link Building 2026).

So our $455 lands about 11% under the survey average and at the bottom edge of the legitimate band. Not because we have secret inventory. Because of four levers, which are the actual subject of this article.

One more market number before the levers, because it frames everything below the band: in the largest audit of guest-post marketplaces we’ve seen, 85.3% of marketplace inventory rated low quality (Xamsor, guest-posting marketplace study). When you see links priced at $60 to $150, that’s the shelf they mostly come from. It isn’t the same product at a better price. It’s a different product at its correct price.

What actually moves per-link cost

Four things set the price of an editorial link: the publisher’s quality, how you negotiate, whether you bundle, and who pays for the content. In our experience that’s also the order most buyers get wrong, so we’ll take them in order of leverage instead.

Negotiation craft: our actual rules, quoted

This part usually gets described in adjectives (“we negotiate hard”). Here is our buying policy as it exists in our production configuration, not as a philosophy:

1. Never accept the first quote. A first quote is an opening position, and publishers who sell placements know it. In our experience the counterparty expects the counter; the only party surprised by negotiation is the buyer who didn’t.
2. Open counters anchor at 65% of the ask. Quote us $600 and our counter opens near $390. The point of 65% is that it’s low enough to move the midpoint and not so low that it reads as an insult and ends the thread. Most deals settle between the anchor and the ask.
3. Maximum two negotiation rounds, then walk or escalate. A negotiation that survives two rounds without converging isn’t a negotiation, it’s a hobby. Walking is a fine outcome; there are more publishers than budgets.
4. Payment after publication, always. We never prepay unverified partners. Prepaying a stranger converts your worst case from “slow deal” to “total loss,” and at volume you will eventually meet the stranger who makes that math real.
5. Listicle paid inclusions auto-accept up to $150. Below that line, negotiating costs more in hours than it saves in dollars. Knowing when not to negotiate is part of negotiating.

None of this is clever. It’s just written down and enforced, which puts it ahead of most buying we’ve seen, including some of our own before we wrote it down.

Bundling: the single biggest lever we know

The best price we saw in this period came from a package. One publisher quoted £600 for three placements when we negotiated them together: about £200 per placement, far below what the same placements would have cost bought one at a time.

The mechanism is ordinary economics. For the publisher, three placements as one deal means one negotiation, one invoice, one counterparty they’ve already vetted. They price the relationship cheaper than the transaction, because it is cheaper. For the buyer it means the fixed cost of finding and qualifying that publisher amortizes across three links instead of one.

The practical rule: before you ask a publisher for a single-placement price, check whether they fit two or three of your targets, and if they do, ask for the package price first. In our numbers nothing else, not templates, not timing, not charm, moves per-link cost as much.

Publisher quality: the floor you shouldn’t fight

Real publications have real costs. Editors who say no to things are the entire reason the link is worth buying, and their salaries are in the price. You can negotiate a legitimate publisher toward fair; you cannot negotiate them to marketplace prices, and if you succeed, update your opinion of the publisher rather than your opinion of your negotiating.

This lever works in one direction: it sets the floor. The other three levers decide how close to the floor you land.

Content production: the cost that hides off the invoice

Somebody writes the article the link lives in, and that somebody gets paid whether or not the line item is visible. A “free” guest post with $300 of writing and editing inside it is a $300 link before any placement fee. When you compare quotes, normalize for who produces the content, or you’re comparing prices of two different things.

This is also the quiet reason agency quotes vary so much for what looks like the same placement: some quotes include a writer, an editor, and revisions, and some include a prayer.

Why cheap links cost more: the survival math

Here’s the part of the 2018 study we’d write differently today, because we now have drop data of our own.

Even at the quality end, on vetted publishers with verified traffic, 2 of our 12 links dropped within the year. That’s a 17% in-year drop rate from an operation that checks every link daily and picks publishers carefully. We replaced both at our own cost because the guarantee is ours to honor, but we want the uncomfortable number in print: link attrition happens at the good end too.

Now scale that intuition down-market. Industry-wide, 66.5% of links built between 2013 and 2024 are now dead (Ahrefs, link-rot research). And 85.3% of marketplace inventory, per the Xamsor study above, is low quality to begin with: sites built to sell placements, which are exactly the sites that get cleaned out, deindexed, or abandoned.

So do the arithmetic buyers skip. The real price of a link is the sticker price divided by the odds it still exists when you need it:

• A $509 link with a replacement guarantee behind it costs $509 per surviving link, because the drop risk is the seller’s problem. That’s what the guarantee is: transferred risk, priced in.
• A $150 marketplace link with even a coin-flip chance of surviving the year (generous, given the rot data) costs $300 per surviving link. Before you count that the surviving link lives somewhere no editor, reader, or AI model has ever been.
• And nobody at the $150 end replaces drops. The replacement is a second purchase, at full price, from the same shelf.

Cheap links aren’t a budget version of expensive links. They’re a slower way to spend the same money for less. That’s the finding we’d put on the cover if research had covers.

What we’d tell a buyer in 2026

If you skim one section, skim this one:

1. Budget $500 to $900 for legitimate editorial. Anything far below that band is a different product, whatever the listing says.
2. Never take the first quote. Counter around 65% of the ask. Two rounds, then walk.
3. Ask for package pricing whenever one publisher fits more than one target. Biggest lever on this list.
4. Pay on publication. Prepaying unverified partners is how link budgets disappear without a trace.
5. Don’t negotiate under ~$150. Your hours are a cost too. Pay the small inclusions and move on.
6. Price the guarantee. A link that drops unreplaced doubles its own price. Ask any vendor what happens at month eight when a placement vanishes, and listen for a specific answer.
7. Normalize for content production. A quote that includes the writing and a quote that doesn’t are not the same number.

Our own $455 is what those seven rules produce when they’re enforced by config instead of willpower, on publishers good enough that we’ll show the traffic data. What we charge for that work is public on our pricing page, and the delivery mechanics are documented on the editorial links service page; the numbers there reconcile with the numbers here, which we’d argue should be the minimum bar for anyone publishing a pricing study.

We put our invoices in public for the same reason we put reply rates in public: this industry runs on adjectives, and adjectives are how bad links get sold. If you’d rather see the proof habit than read about it, the free baseline we run takes us about a day and shows where your brand actually stands before anyone spends $455 on anything.