About SerpInsight

Why SerpInsight exists

This started as a spreadsheet. Years of working in-house meant signing off on link invoices every month, and then quietly opening each placement to see what had actually been bought. A lot of it was fine. A lot of it wasn’t — links dropped into articles that had never ranked for anything, on domains with a Domain Rating of 58 and a traffic graph flatlined at forty visits a month, in paragraphs that had clearly been written backwards from the anchor.

The reports never mentioned any of that. They listed DR, they listed the URL, and they moved on. So the spreadsheet grew a second tab: a checklist of things to verify before accepting a placement. Does the host page rank for a real query? Does anyone read it? Does the sentence still make sense if you delete the link? Does the site have any other articles on this topic, or is this the one lonely post about kitchen renovation on a general blog?

SerpInsight is that checklist turned into a service. Nothing more mysterious than that.

We’re new, and we’d rather say so

SerpInsight launched recently and we’re small. We don’t have a wall of client logos, a “500 placements delivered” counter, or five years of retention data, and we’re not going to invent any of it — you can find plenty of vendors who will. What we have instead is a method we’re willing to describe in complete detail, including the parts that make us look slower and more expensive than the competition.

That’s a deliberate trade. An established agency can point at results and stay vague about process. We can’t point at results yet, so the process has to carry the weight. Every standard on this page is written as a commitment you can hold us to and check yourself, because you can open any placement we report and verify it in about four minutes. Do that. We’d genuinely rather work with clients who check.

Work we turn down

Some of this costs us money, and we do it anyway.

We don’t place casino, sportsbook, pharmacy or payday lending links into general-interest sites. Not because those industries are beneath us, but because the placement is a lie — a home decor blog does not organically link to an online casino, and every party involved knows it. Those verticals have their own publications, and that’s where those links belong.

We also turn down clients whose page isn’t ready. If the URL you want to move is thin, slow, or doesn’t actually answer the query you’re targeting, links won’t fix it. They’ll just make an underperforming page slightly more visible to a crawler that already decided. In most of those cases the honest advice is to spend the money on the page first, and we’ll tell you that before we quote. Our SEO audit and on-page work exist partly because that conversation kept happening.

And we decline exact-match-only anchor demands. If you want every link pointing at your money page to read “best project management software,” we’re not the right vendor. A healthy profile runs somewhere around 40–50% branded, 25–35% partial or topic-based, 10–20% naked URL or generic, and only 5–10% exact-match. We’ll build to something like that mix or we’ll pass. The reasoning is laid out in our anchor text ratio guide.

What gets published on the blog

The blog has one editorial rule: we publish what we’ve actually tested or verified, and we say which is which. When something is industry-general knowledge rather than our own measurement, it’s hedged — “typically,” “in our experience,” “in most niches” — and when we don’t have data, we describe the method instead of inventing a statistic.

That includes writing up things that stopped working. Outreach angles that used to get replies and now don’t. Vetting signals that were useful in 2019 and are noise now. A tactic dying is more useful to you than another post explaining what a backlink is, and it’s the kind of thing agencies tend not to publish because it dates their playbook.

Our vetting standard

This is policy, not an aspiration. Every host page we propose has to clear all six before it reaches your shortlist. We’re not going to quote you an accept rate, because we haven’t placed enough links to have measured one honestly — but here’s the bar itself:

• At least 1,000 monthly organic visits to the host site — traffic, not Domain Rating alone
• The specific host page is indexed and ranks for a real keyword, not just the domain
• Shared search intent between the host page and your target page — adjacent topic, same reader
• The paragraph still reads correctly with the link removed — if it doesn’t, the link was the point
• At least 3 other articles on the host site in the same topical cluster
• A clean outbound profile — no wall of unrelated sponsored links stacked in the same post

Anything that fails one of these gets dropped before you see it. You then approve or veto what’s left, so the final call is yours. If you want to run the same checks on links you’ve bought elsewhere, our backlink quality evaluation guide walks through it step by step.

What we’d tell you not to buy

Don’t buy links if your site is under roughly DR 20 with a handful of thin pages. At that stage links have almost nothing to amplify, and you’ll spend $1,500 discovering that your content was the constraint.

Don’t buy packages sold purely on DR. “50 DR 50+ backlinks for $499” is arithmetically impossible to deliver honestly — that’s under $10 a link against four to six hours of real work per placement. Somebody owns those sites.

Don’t buy links to a page that hasn’t been touched in three years. Refresh it first, then build. And don’t buy sitewide footer or sidebar links at any price; they’re the easiest paid pattern in the world to spot, and they’ve been getting discounted for well over a decade.

Paid links are against Google’s guidelines. That applies to us as much as anyone else, and we can’t make the risk zero — vetting, anchor discipline and sensible pacing reduce it, they don’t remove it. If a vendor tells you their links are guideline-compliant and completely safe, that’s the moment to close the tab.

Work with us

If the standard above matches how you’d want your own links vetted, we’re probably a fit. Start with the link insertions service to see the full process, check the pricing, or send us the page you’re trying to move — we reply within 24 hours, and if links aren’t the right spend for you right now, we’ll say that instead of quoting.

What makes us reject a host site

The vetting standard describes what a host page has to pass. This is the other half of it — what we actually find when we look, and which findings end the conversation. Most rejections happen in the first two minutes, before we’ve contacted anyone, which is why the prospect list you eventually see is shorter than you’d expect.

Two of these have a grey zone worth naming honestly. A sponsored-posts page doesn’t disqualify an otherwise excellent site, and plenty of legitimate publications run one. And a single strong article in your cluster can occasionally be enough if the article itself ranks well and the site is genuinely authoritative on an adjacent subject. Where we land in a grey zone, we tell you it’s a grey zone in the shortlist notes rather than presenting it as a clean pass. You can veto it, and some clients do.

What we’re building toward

It would be easy to fill this page with numbers. Durability percentages, accept rates, average ranking movement — every one of them is a plausible-sounding figure that nobody checks, and a new agency could invent all three this afternoon. We’re not going to, and it’s worth being specific about what that means rather than just claiming virtue.

Here is what we don’t have yet, stated plainly: we have not placed enough link insertions (also called niche edits) to report a durability rate, so we don’t have one. We have not run enough prospect lists to report an accept rate, so we don’t have one. We have no client results to publish, because we have not yet had clients whose campaigns have run long enough to produce results worth publishing. Any site telling you otherwise about its own first months is describing an ambition in the past tense.

What we’ve committed to instead:

• We monitor every placement for 12 months from day one, so the durability data accumulates whether or not it flatters us. When there’s enough of it to mean something, we’ll publish it — including the removals
• We’ll publish an accept rate once it’s measured across enough prospect lists to be a rate rather than an anecdote, and we’ll say how many lists it came from
• Client results get published only with written permission, and only with the parts that didn’t work included. A case study that reports a win and omits the two months of nothing beforehand is advertising
• Until then, the case studies section stays honest about being early, and the blog keeps labelling which claims are measured and which are industry-general

The obvious objection is that a vendor with nothing to show has an easy time promising future transparency. That’s fair. The counter is that everything we do have — the vetting standard, the relevance test, the anchor ratios, the pricing, the rejection reasons above — is already published in enough detail to hold us to, and you can verify any placement we deliver in about four minutes without taking our word for anything.

What happens when we get it wrong

We will get things wrong. A host site that looked healthy will get sold and stripped. An editor will agree to a placement, take payment, and go quiet. A link will go live with an anchor that isn’t the one we agreed, because someone edited it on the way through.

What we do about it is fixed in advance, because policies written after a problem always favour whoever wrote them. If a link is removed, nofollowed without agreement, redirected, or the host page falls out of the index inside twelve months, we replace it free on a site of equal or better tier. You don’t have to spot it first — we re-check every placement on a schedule, and in most cases we’d rather email you about a removal you hadn’t noticed than wait to see whether you do. If we can’t place something you’ve paid for, you get a refund on that line or a substitute at the same tier, and it’s your choice, not ours.

The part that isn’t a policy is the telling. If we place a link on a site that turns out to be worse than we thought, the useful thing is an email saying so, with what we missed and what we’re doing about it. That’s a worse conversation in the short run than hoping you never look. It’s the only version that survives a client who checks, and we’d rather build the business around clients who check. The formal terms sit on the refund and replacement policy page, in the same plain language as this one.

Our position on how links get sold

One mild opinion, since this is the about page and you may as well know where we stand. The problem with link buying isn’t that money changes hands — publications have taken money since publications existed. The problem is that the industry standardised on metrics that can be bought, and then built a marketplace on top of them.

Domain Rating is the clearest example. It’s a useful comparative signal and a terrible product specification, because it can be inflated with links from sites nobody reads, and the inflation is cheap. Once vendors started quoting by DR band, the incentive changed from finding sites with readers to finding sites with numbers, and those are not the same search. That’s how you end up with a DR 60 domain doing 200 visits a month being sold at a premium to somebody who thinks they’re buying reach.

Marketplaces made it worse in a specific way. When inventory sits in a shared catalogue, everybody buys from the same few hundred domains, and a footprint forms that no individual buyer can see from inside their own account. You can do everything right on your side and still end up in a cluster with four hundred unrelated businesses. The screenshots don’t help either — a traffic graph cropped to its best six weeks, or a metric shown without the traffic figure beside it, is a standard way of selling something that wouldn’t survive being looked at in full. None of this requires anyone to be a villain. It’s what happens when the buyer can’t easily inspect the product and the seller is rewarded for volume. We’re not naming vendors, partly because it would be unfair and mostly because the pattern is structural rather than a handful of bad actors.

How to hold us to this

A published standard is only worth something if somebody checks it. Here are the four checks that take the least time and catch the most, in the order we’d run them on a vendor ourselves.

1. Ask for four things before you pay

   The host page URL, the keyword it ranks for, its traffic figure, and the exact paragraph your link will sit in. You should get all four from us on request. A vendor who can only produce the domain is choosing pages from a spreadsheet rather than from search results.

2. Delete the link in your head

   Open any placement we deliver and read the sentence as though your URL isn’t there. If it stops making sense, the paragraph was written to carry the link, and that’s a failed placement by our own standard. Tell us and we’ll treat it as one.

3. Check what else the host is selling

   Look at the outbound links on the same article and a couple of recent posts. A stack of unrelated commercial links means the page sells paragraphs, and we should have caught it before you did.

4. Audit the anchors against the plan

   Compare the anchors we actually used to the mix we said we’d build — roughly 40-50% branded, 25-35% partial or topic-based, 10-20% naked URL or generic, 5-10% exact-match. If an order has drifted toward exact-match, that’s a drift worth challenging.

Then ask us the harder ones. We wrote out our answers to eight questions worth asking any link vendor on the homepage, including who owns the sites, who writes the outreach, and where we stand on Google’s guidance about paid links. Those answers are the same whether you’re evaluating us or somebody else, and they’re the ones we’d want asked if the positions were reversed.

If any of it doesn’t hold up when you check, say so directly — we reply to everything within 24 hours, including complaints. You can see the whole method on the link insertion service page, what it costs on the pricing page, or start a conversation through contact. And if the honest answer to your situation is that links aren’t the right spend yet, that’s what you’ll get.