On my link building services page, I make a specific claim: (if my client allows it), I track link velocity and anchor text ratios on every campaign, so growth stays natural instead of looking suspicious.
Most link builders never write that claim down because it forces them to actually have a system outside just filling a quota.
Ain’t nobody got time for that extra work…but I do.
Well, here's mine, and the not-especially-flattering story of my own page failing my own rules before I fixed it.
Why anchor text monotony is the real spam signal
Most people assume Google's spam systems flag link volume on its own, or a page having way more links than the top-ranking competitors. That's not really what Google has said.
The Big-G hath spoken
John Mueller has been blunt about this: the total number of links is "completely irrelevant," and Google can just ignore a pile of low-quality ones rather than punish you for having them. He's also called the "toxic link score" concept, something SEO tools invented to keep you paying for their tool.
I can put on my tinfoil hat and say “that’s what Google makes you want to think!” - and that could be warranted given our general distrust of trillion-dollar companies. But that’s what the Big-G said on their official guidelines.
It’s the sudden spike in links that gets you…
What actually raises a flag is a sudden spike of links, from low-quality or irrelevant sources. Two hundred links in a month isn't automatically suspicious. Because yes, your 2-month old local knife-sharpening service could have just gone really, really viral.
But two hundred links in a month from templated directories and spun guest posts is suspicious, and it would be just as suspicious at twenty links.
…and the link monotony
What actually reads as manufactured is anchor text monotony: the same or similar exact-match phrase showing up on link after link. If every backlink to your knife sharpening page uses the words "best knife sharpening service," that's the tell, not the fact that there are twenty links pointed at the same page.
So the fix isn't sending fewer links, or spreading them across more pages "for naturalness" (more on why that specific instinct is wrong further down). It's diversifying what the anchor text actually says, on the page you're trying to rank.
What a bought link profile usually looks like
I've bought bad links before and the pattern is almost always the same: a vendor optimizing purely for the keyword gets you a stack of links that all say some version of "best dog training program" or "dog walker," because that's the fastest way to look like they delivered exactly what you asked for.
Nobody on that vendor's end is tracking a ratio and that's the gap my whole system exists to close.
Link marketplaces make this even worse
Most marketplaces make this worse in a specific way.
They let you type in the exact anchor text yourself, free-text field and all. Nobody on the other end checks whether it fits the sentence it's landing in, or whether you've already used the same phrase on your last nine orders. You ask for "kitchen knife sharpening service," you get the exact anchor every time, if you let it.
That feels like you have control, but without getting too philosophical, it's actually the opposite.
A real guest post editor will sometimes push back if an anchor reads awkwardly in their own copy, because their site has to make sense to their own readers, not just to you. A marketplace order form has no reason to push back on anything. It's not editorial, it's fulfillment - just an assembly line to get your link published somewhere. The anchor field doesn't know or care what the last five anchors on that page already said.
So the ratio discipline does not rest with the marketplace. It has to be with whoever's placing the order.
Leave that field on autopilot with the exact-match keyword because "that's what you're trying to rank for," and the marketplace will happily help you build the exact monotonous profile this whole post is about, one order at a time.
A baseline ratio table, if you don't have bandwidth for more
Not every page gets a full competitor analysis, especially early on or on a tight retainer. Here's a reasonable starting mix for any page getting deliberate links, mine or a client's, until you can do better (more on what "better" means right after this):
Category | Target share (approx.) | Example |
Branded | 30-40% | A name with no domain suffix. "Yesh Quijano," "YQ Digital," a client's brand name |
Naked URL / domain | 10-20% | Any domain-formatted string, bare or full path. "YeshQuijano.com," including brand-as-domain |
Generic | 10-20% | Zero fragment of the target keyword. "Check out his work," "here's how he does it" |
Partial match | 15-20% | Part of the keyword phrase, wrapped in other words. "His link building services," "link building specialist in the Philippines" |
Exact match | 5-10% | The literal keyword phrase, standalone. "Link building services" |
Branded should be the biggest single category on a healthy profile. And if exact match is already your largest category, the page is over-optimized.
Doesn't repeating this same ratio become its own red flag?
Fair question. If every SEO on earth mechanically hit the exact same 30 to 40% branded anchors on every campaign, the ratio itself still wouldn't be what gives it away.
What actually gets flagged as a pattern is uniform execution: the same sites supplying every link, the same author byline on every guest post, the same paragraph position, a flat robotic velocity every month, forever.
For example, if we’re building 10 links a month, "30 to 40% branded anchors" is simply the ratio I’m working towards, because not every anchor I ask for makes the cut. An editor won't always accept an exact-match phrase sitting awkwardly in their own sentence, and a citation listing only ever takes a branded or naked-URL anchor no matter what I'd prefer.
So the month's real anchor text ratio lands just close to the target, because some of what I'm aiming for gets vetoed by the site before it goes live. That's why a range works and a fixed number doesn't.
I actually have a list of blacklisted sites since they always, always accept link offers.
What if you just look at your competitors’ anchor text ratios?
Personally, I skip the table above when I have the bandwidth, and instead pull the anchor distribution of the top 5 to 10 ranking pages for the target keyword, then build toward whatever that group actually looks like.
It's a reasonable refinement and more contextually grounded than one table applied to every niche.
But it also has its weaknesses. For example, a sample of 5 to 10 competitors can get skewed hard by one outlier (e.g., a single page with thousands of forum links can wreck the average), and you're often looking at links built up over years under whatever norms applied when those links landed, not necessarily what's safe to build today.
It also assumes the ratio is why those pages rank, when it might just be a byproduct of how they got mentioned naturally, unrelated to rankings at all.
Either way, the honest way to use any ratio is as a guardrail, not a target to hit exactly. The goal was never matching an exact number.
The sequencing rule matters more than the ratio
A 2-month-old page with two links, one of which is an exact-match anchor (so 50% exact-match ratio) looks more suspicious than a mature page that ends up at the exact same eventual ratio.
My rule: don't add a second exact-match anchor to a page until at least two or three branded or partial-match links have landed first.
While a page is young (under roughly 10 total backlinks), front-load branded and generic anchors. Let exact-match accumulate slowly and later, and never let it become the majority category at any point along the way, not just at the end of a plan.
I broke my own guidelines…
Here's the part I don't love admitting, but I'd rather be honest than pretend my own backlink profile was already textbook when I started tracking it.
As of this writing, my own link-building-services page had two backlinks. The anchor texts are "link building services" and “Yesh Quijano”. That's a 50% exact-match ratio and precisely the scenario the ratio guideline exists for: land branded and partial-match links first, dilute that starting ratio, then touch exact-match again.
My reason for the mistake? My team and I are making a big marketing push hoping to land more clients next year and…I just got too excited.
Having a system like this really helped me keep things in check before it got worse. I’m still working on adjusting the ratio.
Target-page concentration isn't the problem you think it is
A common instinct is "I should spread some links to the homepage too, so it doesn't look like I only build links to one page." In my opinion, that’s solving a problem that doesn't exist.
A business sending deliberate links to its own commercial page is exactly what a real business does. The homepage isn't what's being over-optimized here. The anchor text is.
If I genuinely want more link equity spread across pages, I point it somewhere that serves a real purpose: a related service page, a strong supporting blog post, or a case study; not the homepage as a naturalness placebo. Passive links (press mentions, brand mentions, directory listings) already accrue to the homepage on their own, provided you’re actually doing some other type of marketing.
How the anchor rules change by link type
Here’s my quick guide on what anchor texts to use by link type. Again, this is a guide, not rules set in stone:
How I check the current ratio before adding anything new
Before assigning anchors for a new campaign, I pull a backlink report (made through our own internal application) filtered to the specific target URL, to see every anchor already pointing at that page.
[graph]
[anchor text ratio for the recommended page care of our own web app]
It’s better to have these numbers on hand before we start building links.
FAQs
How many backlinks does a page need before this matters?
From link one. The sequencing rule exists specifically because young pages with almost no links are the easiest ones to accidentally over-optimize, not the hardest.
Does this apply to internal links too, or just backlinks from other sites?
Both. Internal links carry less risk, but a page that gets the exact same internal anchor from every post on your own site has created its own smaller version of the same monotony problem.
What if I have a site whose anchor profile is already skewed from a previous agency?
I don't touch the old links. I just make sure every new link I add works toward correcting the ratio, branded and partial-match first, until the whole profile trends back toward healthy. Trying to edit or remove anchor text on links that already exist isn't really an option anyway (more on that below).
Should I disavow links to fix an over-optimized anchor profile?
No. Disavowing is for links you think are actively hurting you, usually spam you didn't build yourself. It doesn't change the anchor text on links that already exist, and it won't fix a skewed ratio. A lot of disavow advice leans on a "toxic link profile", not anchor text overoptimization. Fixing a skewed anchor mix happens forward, with new links, not backward by removing old ones.
Does anchor text on image links count toward the ratio?
Yes. The alt text on a linked image functions as the anchor even with no visible text to read, and it gets classified the same way: branded, generic, partial, or exact match. I don't treat image links as a free pass just because nobody's reading the alt text.
The takeaway
Anchor text ratios aren't about hitting a number for its own sake.
They exist because Google's spam systems are looking for monotony, and the fastest way to create monotony by accident is to keep reaching for the same exact-match phrase because it "feels right" for the keyword you're chasing. Diversify the anchor, sequence it deliberately, and check the running total before you add to it, not after.
If you want to see this running on a real campaign instead of a hypothetical one, that's what I do for clients too.