How to build a link plan that survives a core update

Most link campaigns that collapse after a core update were fragile long before it. The update did not break them; it revealed that everything was resting on one narrow bet.

The shape of a fragile campaign

A fragile campaign usually looks like this: nearly every link points at two or three commercial pages, the anchors cluster tightly around the terms those pages target, and the placements all came from the same kind of site bought in the same way over the same eight weeks.

Each of those on its own is survivable. Together they describe a pattern, and patterns are exactly what an algorithm is good at recognising.

Spread across pages

Point links at the pages that deserve them, which is rarely just the money pages. Supporting articles, comparison pages, and genuinely useful resources all pass authority internally, and they are far easier to earn links to in the first place.

A reasonable split is a minority of your links to commercial pages and the majority to everything that supports them. That also gives you somewhere to send a link when a publisher is not comfortable linking to a sales page, which happens more often than most people expect.

Spread across anchors

Exact-match anchors work, which is why they get overused. The problem is not any single exact-match anchor - it is a profile where most anchors are the phrase you are targeting, because almost nobody links that way naturally.

Vary between your brand name, the bare URL, the page title, a descriptive phrase, and occasionally the target term. The test is simple: read the anchors as a list and ask whether a group of unconnected writers would plausibly have produced it.

Spread across placement types

Guest posts, niche edits and digital PR each leave a different footprint. A profile built entirely from one of them looks procured. A profile that mixes them looks like a business that has been doing several things over time, which is what you want, because it is also true.

Spread across time

The single most common mistake is buying a large batch in one month because budget appeared, then nothing for two quarters. Steady acquisition is both safer and easier to learn from: when everything arrives at once, you cannot tell which placements did anything.

Give people something worth linking to

None of the above helps if the pages receiving the links are thin. Links get a page considered; the page has to earn the position once it is there. This is why content and link building work better planned together than bought separately.

A workable review rhythm

• Every month, check anchor distribution across the whole profile rather than the latest batch
• Every quarter, look at which target pages have moved and which have not
• Before each campaign, confirm the target pages are actually ready to receive traffic
• After a core update, wait for the rollout to finish before changing anything

That last point matters. Most of the damage done after an update is done by people reacting to partial data in week one.