A niche edit is a link added to an article that already exists. No new content is written, no guest post is pitched, someone simply edits a published page and puts your link in it.
That is the whole mechanic. The complicated part is that most of the guides explaining it are written by companies selling them, and they tend to leave out the same thing: in commercial practice, the site owner is being paid. This covers how they work, what they cost, how to judge a target, and what Google's guidelines actually say.
The short version
Niche edits are also called link insertions, contextual links or curated links. All four terms describe the same thing.
They are faster than guest posts because the page is already indexed and, if you picked well, already ranking and receiving traffic. There is no waiting for a new post to be crawled and assessed.
They are usually bought. There is a legitimate unpaid version, covered below, and it is a fraction of what gets sold under this name.
And the quality range is enormous. A link added to a genuinely relevant, well-trafficked article on a site with real readers is one of the better links available. A link added to a five year old post on a site that publishes nothing and exists to sell insertions is worthless at best.
Niche edits versus guest posts
| Niche edit | Guest post | |
|---|---|---|
| Content required | None | A full article |
| Time to live | Days | Two to eight weeks |
| Page already indexed | Yes | No |
| Page may already rank | Yes, and that is the point | No |
| Typical cost | $50 to $500 per link | $100 to $1,000+ per placement |
| Control over context | Low, you fit their article | High, you write it |
| Anchor text flexibility | Limited to what fits | High |
| How obvious it is | Low if placed well | Moderate |
The reason experienced link builders often prefer niche edits is the third and fourth rows. A link on a page that already gets traffic passes something a brand new post does not, and you find out within weeks rather than months whether it did anything.
The guidelines position, plainly
Google's link spam policy treats links exchanged for money, goods or services as link schemes when they are not disclosed with a rel attribute. That covers most commercial niche edits, because most of them involve a payment to the site owner.
That does not mean everyone doing this gets penalised. It means the risk is real, it scales with volume and with how obviously manufactured the placements look, and anybody telling you a paid link insertion is definitively safe is selling you something.
The version that sits inside the guidelines is straightforward: you find a page with a genuine gap, you contact the owner, you explain what you have that improves it, and they add it because it improves the article. No money changes hands. This works, it converts at a low single-digit rate, and it is the version almost nobody means when they say niche edits.
We are not going to tell you not to buy links. Plenty of competent businesses do, deliberately, having weighed it. The mistake worth avoiding is buying them without knowing that is what you bought, which is what happens with most cheap agency packages.
Step 1: find pages, not sites
The common error is evaluating a domain and then asking what pages it has. Do it the other way round.
Search for the phrases your prospects would already be covering: "best [thing] tools", "how to [task]", "[topic] guide", plus roundups and resource lists in your subject. What you want is a specific page that is missing something you have.
The three page types that work are educational articles where your resource genuinely adds detail, resource lists and roundups where you belong on the list, and comparison pages where you are a legitimate option that is not mentioned.
If you cannot describe in one sentence why your link makes that specific page better, it is the wrong page. That sentence is also your entire outreach email.
Step 2: evaluate the page, then the site
Check the page first.
- Does it get traffic? A page with zero organic traffic passes very little regardless of the domain's metrics. This is the single best filter and most people skip it.
- Does it rank for anything relevant? Look at its top queries. A page ranking for your topic is worth several ranking for nothing.
- How many outbound links does it already have? Twenty external links in a 900 word article means it is a link farm page.
- Is the content real? Read it. If it is thin, generic or was clearly written to host links, no metric saves it.
Then the site.
- Has it published anything in the last three months? An abandoned site is a bad neighbourhood.
- Is there a named author with a real presence? Editorial control is the thing you are actually buying.
- Does it have a "write for us" or "sponsored post" page? That is a marketplace, not a publication, and its links are priced accordingly by Google as well as by the owner.
- What is its traffic trend? Declining sharply usually means something happened, and you do not want to be attached to it.
On metrics: use them to exclude, not to select. Domain authority scores are easy to manipulate and a high number tells you a site has links, not that it has readers. Majestic's Topical Trust Flow is more useful than any single authority number here, because it tells you which subjects a site is actually trusted in.
Step 3: the outreach
The unpaid version is a specific email and it is short.
Name the exact page. Say what it is missing, in one sentence, referring to something actually in the article. Say what you have that fills the gap and why it is better than what they currently link to, if anything. Then stop.
No compliments about their "amazing blog". No three paragraphs about your company. No mail merge fields that make it obvious this went to 400 people. Reply rates on this run somewhere between two and eight percent when the personalisation is real, and near zero when it is not.
If you are buying instead, you will usually be dealing with a marketplace or a broker rather than the site owner, and the conversation is a transaction. In that case the diligence in step 2 matters more, not less, because nobody in that chain is filtering for quality on your behalf.
Step 4: the link itself
Anchor text should read as though a writer chose it. Exact-match commercial anchors at volume are the clearest possible pattern, and they are what gets link profiles reviewed.
Prefer the anchor that fits the sentence. If the article says "several tools handle this automatically", the natural anchor is a few words of that clause, not your product name plus your primary keyword.
Placement matters more than people think. A link in the body, in a paragraph that is genuinely about your subject, is worth several in an author bio or a resource list at the foot of the page.
Point it at the page that deserves it. Sending every link to your homepage wastes them. The page you want to rank is the page that should receive the link.
Step 5: verify, then keep verifying
Check the link is live, that it is not nofollow or sponsored unless you agreed to that, and that the page is still indexed. Do this a week after placement and again at 90 days.
Links bought through brokers disappear. Sites get sold, redesigned, or the owner quietly strips old paid links when a new SEO arrives. A monitoring tool that alerts you on lost links pays for itself the first time it catches one.
Keep a record of what you placed, where, at what cost, and what the target page did afterwards. Almost nobody does this, and it is the only way to find out whether any of it worked.
What a niche edit should cost
Prices vary by market and by how much traffic the page has, but the shape is consistent.
- $50 to $100. Low-traffic pages on sites that exist largely to sell links. Occasionally fine, usually not.
- $100 to $300. The bulk of the market. Real sites, modest traffic, mixed quality. Diligence decides whether it is worth it.
- $300 to $800. Pages with genuine traffic on sites with real editorial standards. This is where the value usually is, if you have the budget.
- Above $800. Established publications. Rare for insertions, more common for guest posts.
- $0. The outreach version, where the cost is your time and your reply rate.
If a vendor offers you 50 links for $1,000, you are buying the first band, and 50 of them will not outperform five of the fourth.
How to spot a bad vendor
They will not tell you the sites in advance. Any vendor unwilling to name the placement before you pay is protecting a list they know you would reject.
They lead with domain authority and say nothing about traffic. DA is the metric that can be bought; traffic is the one that cannot.
They guarantee rankings. Nobody can, and the guarantee is a signal about what else they are willing to say.
Their sample sites publish on every topic. A site with posts on crypto, dentistry, casinos and dog food is not a niche site, whatever the invoice says.
The alternatives, honestly
Digital PR. Slower, more expensive up front, and the links are earned rather than bought. If you can produce something genuinely newsworthy, this is the version with no downside. Our guide to digital PR for link building covers the process.
Brand mention reclamation. Finding places that already mention you without linking, and asking. Free, high conversion, and finite. Covered in our brand mention link building guide.
Exchanges. Two businesses agree to link to each other's relevant content. Both sides consent, no money moves, and the links sit on real operating sites rather than on link inventory. This is the model we run, and the honest trade-off is that you do not choose the specific site: matching is by relevance and quality threshold. See the backlink exchange, or the wider comparison in best backlink exchange platforms.
Being worth linking to. Not a tactic and the reason most link building is difficult. A page with original data, a genuinely useful free tool or a real opinion attracts insertions without anyone asking.
When niche edits are the right call
When you need a specific page to move and it is close, within striking distance rather than on page five. A handful of relevant contextual links can be enough at that margin.
When you have a genuinely good asset and outreach is the constraint rather than the asset. Then the unpaid version works and you should be doing it at volume.
When your competitors are all doing it in a market where that is the norm, and you have weighed the risk deliberately rather than by default.
They are the wrong call when your site has thin content, when you are buying at the bottom of the price range, or when you are hoping links will substitute for having anything worth ranking. Links amplify a good page. They do not create one.
For the tooling side of all of this, see our breakdowns of link building automation tools and how to automate SEO content creation and link building.
Frequently asked questions
Are niche edits safe? The unpaid version, where a site owner adds your link because it improves their article, is inside the guidelines. The paid version is not, and the risk scales with volume and with how obviously manufactured the pattern looks. Most of what is sold as niche edits is the second kind.
Do niche edits still work in 2026? Yes, when the target page has real traffic and real relevance. What has stopped working is volume on low-quality inventory, which is most of what the cheap end of the market sells.
How long until a niche edit affects rankings? Typically two to eight weeks, faster than a guest post because the page is already indexed. If nothing has moved by twelve weeks, the link was probably on a page that passes very little.
How many do I need? There is no number. A page within striking distance may move on three good links. A page ranked 60th will not move on thirty. Position, competition and the page's own quality decide it, not a quota.
Should I disclose paid links? Google's guidelines say yes, via a rel="sponsored" or rel="nofollow" attribute. Doing so removes most of the SEO value, which is why almost nobody does it, and that gap between the stated rule and common practice is the whole risk conversation in this category.
What is the difference between a niche edit and a link insertion? Nothing. The same product is sold under both names, plus contextual links and curated links. Different vendors, same mechanic.